PRECARIOUS VALUES AND MUNDANE …...structure in a sample of 515 private liberal arts colleges over the period 1987 through 2006. THEORY: PRECARIOUS VALUES AND MUNDANE INNOVATIONS

PRECARIOUS VALUES AND MUNDANE INNOVATIONS:ENROLLMENT MANAGEMENT IN AMERICAN

LIBERAL ARTS COLLEGES

MATTHEW S. KRAATZUniversity of Illinois

MARC J. VENTRESCAUniversity of Oxford

LINA DENGUniversity of Illinois

Drawing primarily from Selznick’s institutionalism, we make a general case for re-newed attention to the “mundane administrative arrangements” that underlie theorganizational capacity for value realization and a particular case for the study ofvalue-subverting management innovations. An empirical study of “enrollment man-agement” in liberal arts colleges reveals this ostensibly innocuous innovation’s value-undermining effects and identifies the organizational and environmental factors thathave made these venerable organizations more or less susceptible to its adoption.

The lesson is this: Those who deal with the moreobvious ideals—such as education, science, creativ-ity, or freedom—should more fully recognize thedependence of these ideals on congenial though of-ten mundane administrative arrangements.

-Selznick, 1957: 14

In the middle decades of the last century, PhilipSelznick and his colleagues developed their insti-tutional theory of the organization. This theorydrew attention to the previously undocumentedphenomenon of “value infusion” and portrayed for-mal organizations as the primary structural “vehi-cles” through which various social values werepursued (Selznick, 1943, 1949, 1957, 1996). Recog-nizing the critical role that organizations played inthe promotion and protection of many societal val-ues (e.g., “education, science, creativity, and free-dom”), Selznick and his peers built a research pro-gram in which they aimed to address two distinctbut integrally related questions. Specifically, theyset out to identify the factors that tended to makeorganizational values either precarious (and thussubject to attenuation and displacement), or elsesecure (and therefore capable of being sustainedover time) (Clark, 1956, 1960; Gusfield, 1955; Mess-

inger, 1955; Selznick, 1949, 1957; Zald & Denton,1963; see Perrow [1986] for a review).1

This early institutional research produced anumber of enduring insights and also laid the em-pirical and theoretical foundation for the entirebody of work that Selznick would develop over thecourse of his long career. In the preface to his 1992magnum opus, The Moral Commonwealth, Sel-znick observed:

Most of my specialized writings in the sociology oforganizations and the sociology of law have beenpreoccupied with the conditions and processes thatfrustrate ideals or, instead, give them life and hope.

Both Selznick (1992, 1996, 2000, 2008) and hiscontemporary interpreters (Heclo, 2002; Kraatz,2009; Krygier, 2002, 2008; Lacey, 2002) have notedthat this highly distinctive and dual-edged mastertheme first emerged in his early organizationalstudies, particularly TVA and the Grass Roots(1949) and Leadership in Administration (1957). Inthe former, Selznick revealed formal organizationsto be “recalcitrant tools” that often fell victim togoal displacement and failed to realize their initial

We thank Geoff Love, Huseyin Leblebici, Eric Neu-man, and the anonymous AMJ reviewers for their helpfulcomments and other assistance in the development ofthis research.

1 Selznick explained his conception of values thusly:“‘Value’ is not used here in a very general sense, such as‘object of interest,’ but rather to denote something whichin the given organization is taken as an end in itself”(1957: 57). This is consistent with other definitions in theorganizational literature (Amis, Slack, & Hinings, 2002;Enz, 1988).

� Academy of Management Journal2010, Vol. 53, No. 6, 1521–1545.

1521

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download or email articles for individual use only.

purposes. In the latter, he put forth a more prescrip-tive statement of the same theory, one that high-lighted the positive aspects of institutionalizationand also identified various structures and tacticsthat could be used to forestall goal displacementand value attenuation.

Although Selznick’s theory fell out of favor for anumber of years as the more cognitive and field-focused neoinstitutional perspective rose to prom-inence, Selznick has subsequently enjoyed a signif-icant resurgence (Greenwood & Hinings, 1996;Hirsch & Lounsbury, 1997; Stinchcombe, 1997). In-stitutional scholars have evinced a renewed inter-est in organizational values, and a parallel concernwith processes of value change and subversion(Amis et al., 2002; Brint & Karabel, 1991; D’Aunno,Succi, & Alexander, 2000; Hinings, Thibault, Stack,& Kikulus, 1996; Kraatz & Zajac, 1996; Oliver,1992). Contemporary institutionalists have also de-voted a great deal of recent attention to intraorgan-izational conflict and to the strategic work that isinvolved in building, sustaining, and changing in-stitutions (D’Aunno, Sutton, & Price, 1991; Heimer,1999; Kraatz & Block, 2008; Kraatz & Moore, 2002;Lawrence & Suddaby, 2006; Oliver, 1991; Zilber,2002). Selznick—who cast the “institutional leader”as the central actor in his value-centered, change-focused, and thoroughly political theory of organi-zations—has been cited as an important influencein much of this work.

The overarching purpose of the present study isto contribute to this ongoing revival and to furtherthe reconciliation between Selznick’s institutional-ism and the now-dominant neoinstitutional per-spective. We aim to do this in two related ways.First, this study recovers a critical aspect of Sel-znick’s theory that appears to have been largelyoverlooked in most recent efforts to revisit the per-spective. Building from Selznick’s opening quote,we explain why organizational values do, in fact,frequently depend on the existence of “congenialthough often mundane administrative arrange-ments.” We then go on to show how these samevalues can be subverted as organizations adopt newand ostensibly innocuous innovations that promiseto solve their technical and administrative prob-lems. This demonstration represents a significantaddition to recent scholarship, which has tended tofocus on agency, politics, and interinstitutionalconflict as the main mechanisms through whichorganizational values change, while saying muchless about adaptive and evolutionary processes(which were also a major focus of Selznick’s the-ory). Second, the study develops an integrative the-oretical and empirical model that explains organi-zations’ propensity to initially adopt (or, instead,

presciently reject) value-threatening administrativeinnovations. This model draws both from Sel-znick’s theory (with its insights about the organiza-tional and local environmental factors that tend tomake values more precarious or secure) and fromthe contemporary neoinstitutional perspective(with its core insights about innovation diffusionprocesses and the causal role of the broaderenvironment).

The empirical context for our research is Ameri-can higher education, a sector comprising organi-zations that serve important and readily apparentsocial values. We focus more specifically on liberalarts colleges, a well-established subgroup of organ-izations in the higher education sector, with anaverage age of more than 100 years. These collegestend to be particularly self-conscious about theirvalues and missions and have previously been por-trayed as archetypal Selznickian institutions (Clark,1970; Kraatz & Zajac, 1996).2

The administrative innovation of concern is en-rollment management. This innovation (which isfrequently abbreviated as EM) is both an adminis-trative structure and a set of accompanying prac-tices that have come into wide use over the past twodecades (Dolence, 1998; Hossler, 1984, 2004). Thestructural aspect of EM involves the consolidationof various administrative functions that affect en-rollments (most typically the admissions and finan-cial aid offices). These merged units are usuallyplaced under the control of a new administrativeofficer who is given a title such as “vice presidentfor enrollment management.” Though the practiceof EM is varied, the “strategic” use of financial aidappears to be its most central element. Collegesengaged in EM purposively allocate financial aid inorder to achieve and maintain desired enrollmentlevels, attract top students, and fulfill financial ob-jectives. This work (often called “tuition discount-ing”) is made possible both by the removal of thestructural barrier separating the admissions and fi-nancial aid offices and by the creation of an inte-grating administrative position (Hossler, 2004;Russo & Coomes, 2000).

Drawing from Selznick’s theory, we will explainwhy enrollment management poses a predictable

2 Liberal arts colleges are small, privately governedinstitutions that typically provide only undergraduateeducation and grant most of their degrees in the liberalarts and sciences. They are distinct from state-fundedinstitutions, and from other private institutions that offera broader array of graduate and undergraduate degreeprograms. The higher education literature indicates thatenrollment management has had outsize consequencesfor these organizations (Redd, 2000).

1522 DecemberAcademy of Management Journal

(if incipient and often unrecognized) threat to es-tablished organizational values. Specifically, weemphasize that the EM structure undermines theautonomy of key internal elites and thereby lessenstheir ability and inclination to “defend the valuesentrusted to them” (Selznick, 1957: 94). We alsoargue that EM allows market values to unobtru-sively penetrate a college and provides these valuesand their advocates with a structural and politicalfoothold inside the organization. We buttress thesetheoretical arguments about EM’s prima facie sig-nificance for organizational values with evidencefrom previous higher education research docu-menting its longer-term consequences for organiza-tional value attainment. Studies have shown that“strategic” financial aid policies strongly favormiddle-class students with a greater ability to pay,thereby undermining equal access to higher educa-tion (Baum & Lapovsky, 2006; Davis, 2003). Theyhave also shown that rapidly growing enrollmentmanagement expenditures have consumed a grow-ing percentage of colleges’ budgets and divertedscarce institutional resources away from core aca-demic areas such as instruction and academic sup-port (Goral, 2003; Redd, 2000).

Having established the institutional significanceof this outwardly mundane innovation on both log-ical and empirical grounds, we return to our coretask of identifying the factors that have made indi-vidual colleges more or less susceptible to its initialadoption. In developing our hypotheses, we locatecausal influences that lie both within an organiza-tion (e.g., in its history, internal power structure,and leadership) and outside it (e.g., in its compet-itive relationship with other organizations and itsbroader cultural environment). In so doing, we ef-fectively “situate” Selznick’s organizational insti-tution in its broader environmental context andshow how larger field-level trends can penetrate anorganization, disrupt its internal normative and po-litical order, and thereby affect its capacity forvalue realization. We test our integrative theoreti-cal model in an event history analysis tracing theinitial adoption of the enrollment managementstructure in a sample of 515 private liberal artscolleges over the period 1987 through 2006.

THEORY: PRECARIOUS VALUES ANDMUNDANE INNOVATIONS

Institutions embody values, but they can do so onlyas operative systems or going concerns. The troubleis that what is good for the operative system doesnot necessarily serve the standards or ideals that theinstitution is supposed to uphold. Therefore insti-tutional values are always at risk. Insofar as organi-

zational, technological and short-run imperativesdominate decision-making, goals and standards arevulnerable. They are subject to displacement, atten-uation, and corruption. (Selznick, 1992: 244)

Our study adopts an essentially Selznickian the-oretical frame. Our core concern is with “under-standing the fate of ideals in the course of socialpractice” and shedding additional light on the or-ganizational “conditions and processes” that affectthis fate for better and worse (Selznick, 1992: x).We have suggested that other scholars who areinterested in these same general questions mightprofit from renewed attention to the prosaic struc-tures and practices that serve to either support orundermine values (Clark, 1956; Selznick, 1957).We have also identified a particular need for re-search examining the factors that make organiza-tions more or less impervious to the adoption ofvalue-threatening administrative innovations. Therationale behind these core arguments rests on twoSelznickian premises.

The first premise is that values need congenial,sustaining social structures if they are to be real-ized, even partially. As earlier explained, this is themost central idea in Selznick’s institutionalism andthe cornerstone of his larger social theory. In Lead-ership in Administration, he suggested that organ-izations can (and should) “embody” institutionalvalues and repeatedly emphasized the need tobuild values “into the social structure of the enter-prise” (1957: 90). The book elaborates many of theways in which this can be done (e.g., through so-cialization, selective hiring, creation of participa-tive decision-making structures, and formulation ofsocially integrative mission statements). It alsogives detailed attention to the value-sustaining roleof formal organizational structure and places a par-ticular emphasis on the role played by “elites”therein.3 Selznick portrayed these internal elites asinstitutional “guardians” and stressed that theyneed to be granted sufficient power to “defend thevalues entrusted to them” (1957: 94). He also ex-pounded upon the need for elite autonomy, a par-tially distinct concern that we discuss in more de-tail below.4

3 In the context of his theory, Selznick defined an eliteas “any group that is responsible for the protection of asocial value.” He noted this: “There would be no greatharm in substituting the term ‘profession’ or ‘profes-sional group’ for ‘elite’” (1957: 120). As the definitionmakes clear, these “elites” do not necessarily occupyhigh-status positions.

4 Knudsen (1995) insightfully characterized Selznick’sperspective as a “constitutional” theory of the organiza-

2010 1523Kraatz, Ventresca, and Deng

The second premise, also well reflected in thequote from Selznick that begins this section, is thatorganizations are not just embodiments of institu-tional purpose or vehicles for the pursuit of highideals. They are also “operative systems”—bureau-cratic entities with their own prosaic and recurrentproblems that can be expected to follow their owndistinct (and often narrowly technical) rationalityin attempting to solve these problems. Thoughproblem-solving efforts may take many differentforms, changes in the design of an organizationitself are high on the list of available options.

Taken together, these two premises reveal an im-portant and abiding paradox: Values are manifestlydependent on the existence of sustaining struc-tures, yet these values do not inevitably come tobear upon organizational design (and redesign) de-cisions. To the contrary, there is a demonstrabletendency for organizations to alter their structuresand practices in response to “organizational, tech-nological, and short-run imperatives” emanatingfrom their operative systems. As such changes oc-cur, values that had been effectively “built into” thesocial structure of an organization are at risk ofbeing removed, and previously granted autonomiesare in danger of being revoked. New values andcommitments—ones that are not necessarily desir-able for the organization as a whole—may also finda place in its internal social structure as a result ofthese ordinary and ongoing adaptation processes.

It is because of this fundamental paradox thatSelznick emphasized the need to construct formalorganizations that thoroughly embodied institu-tional values, from their core governance structuresall the way down to their “mundane administrativearrangements.” The same paradox is at the root ofour expressed concern with prosaic managementinnovations. Colleges, hospitals, charities, govern-ment agencies, and other organizations that aspireto achieve important social goals routinely adopt“rational” and “progressive” administrative inno-vations designed to improve efficiency and achieveother operative system goals (Birnbaum, 2000;Burns & Wholey, 1993; Fiol & O’Connor, 2003). Webelieve that these innovations may often have un-foreseen and undesired consequences that frustratethe realization of organizational values. We alsobelieve that the ostensibly innocuous and purelytechnical nature of these organizational changesmay substantially exacerbate the risk that they

pose. Some changes in organizational design (e.g.,changes in core governance structures) have obvi-ous implications for values and can be reliablyexpected to activate internal, value-based resis-tance for this reason. We cannot, in contrast, safelyassume that would-be value “guardians” will al-ways provide this same degree of scrutiny and at-tention to banal administrative innovations. Forthese reasons, we believe that it is especially im-portant to document these innovations’ conse-quences and to develop knowledge about the forcesand processes that make an organization more orless vulnerable to their adoption.

ENROLLMENT MANAGEMENT AND ITSIMPLICATIONS FOR INSTITUTIONAL VALUES

What Is EM?

As noted above, enrollment management is bothan organizational structure and a set of accompa-nying practices (Hossler, 1984, 2004; Ingersoll,1988). The structural aspect of this managementinnovation involves the consolidation of variousadministrative functions that have the potential toaffect enrollments and tuition revenues. The ad-missions and financial aid offices are the two func-tions that have the strongest and most obvious in-fluence on these outcomes, and the integration oftheir activities is at the core of most EM initiatives.This integration is facilitated by both the dissolu-tion of the structural barrier separating these func-tions and the creation of a new administrative po-sition charged with overseeing them both (typicallya “vice president for enrollment management”)(Hossler, 1984; Russo & Coomes, 2000).

Because the purpose of these structural changesis to increase coordination among admissions, fi-nancial aid, and other relevant functions, EM prac-tices go hand in hand with the structure itself(Hossler, 1984, 2004). The most common enroll-ment management practice—now so widespread asto be virtually synonymous with the concept—is“tuition discounting.” Colleges engaged in EM stra-tegically allocate financial aid in an effort to: (1)maintain desired enrollment levels, (2) meet finan-cial goals, and (3) manipulate the academic anddemographic profiles of their incoming classes(Baum & Lapovsky, 2006; Dolence, 1998). Severalobservers have noted that the practice of tuitiondiscounting is strikingly similar to airline pricing(Bronner, 1998; McPherson & Schapiro, 1998). Col-leges using this practice adjust the “prices” chargedto would-be students, just as airlines adjust priceson certain routes in an effort to match utilizationwith capacity. The key difference is that colleges

tion. In keeping with this characterization, it stressesboth the need for the “vesting of powers” in elite groupsand the simultaneous need to maintain an appropriate“separation of powers” among them.


make this price adjustment by granting institution-ally funded financial aid, rather than by reducingtheir “sticker prices” (i.e., their published tuitionrates).5 Another important distinction is that col-leges using enrollment management are typicallyconcerned with attracting particular students,rather than just paying customers in general. Thus,they generally offer larger discounts to studentswho are more academically qualified. This partic-ular form of discounting, known as “merit aid,” is akey part of most contemporary enrollment manage-ment programs (Heller, 2004; Heller & NelsonLaird, 1999).

Research has shown that enrollment manage-ment has spread widely throughout Americanhigher education over the last two decades and thatit is particularly prevalent among the small, privateinstitutions that make up our sample (Baum &Lapovsky, 2006). Our own data reveal that 46 per-cent (239 of 515) of all private liberal arts collegeshad adopted an EM structure by 2006. This figure isup from only 3 percent in 1987. Higher educationresearchers have documented a corresponding in-crease in the mean “tuition discount rate” withinthe larger population of private colleges and uni-versities, from 21 percent in 1990 to 39 percent in2004 (Baum & Lapovsky, 2006; Redd, 2000).6 Al-though the macrolevel forces underlying thesesharp increases are complex and multifarious, ob-servers have previously explained EM’s spreadwith reference to a number of larger trends. Theseinclude: (1) escalating competition for students andtuition dollars, (2) the emergence of college-rankingschemes and the accompanying obsession with in-stitutional status, (3) declining government supportfor higher education (particularly in the area ofstudent financial aid), and (4) the larger trend to-ward marketization that has manifested itself inmany different areas of American higher education(Baum & Lapovsky, 2006; Geiger, 2004; Hossler,2004; Kirp, 2003).

How Does EM Threaten Institutional Values?

As the preceding discussion reveals, enrollmentmanagement is a multifaceted innovation. It is astructure, a set of accompanying practices and, tosome extent, a manifestation of a larger culturaltrend. Liberal arts colleges, for their part, also havemany different values, and these vary appreciablyamong members of the category. In light of thesecomplexities, we restrict our focus to a few coreelements of EM and discuss these elements primar-ily as they bear on the normative and political orderof a typical college. Both our past research on thiscontext (Kraatz & Moore, 2002; Kraatz & Zajac,1996) and Selznick’s theory of the organizationalinstitution shape our understanding of this internalsocial order. Throughout our discussion, we but-tress our arguments about enrollment manage-ment’s prima facie relevance for organizational val-ues with secondary evidence of its indirect andlonger-term consequences.

Internal elites lose autonomy. As earlier noted,Selznick’s theory of the organizational institutionencourages an analyst to see organizational sub-units not only as manipulable parts of an organiza-tion’s bureaucratic machinery, but also (and indeedprimarily) as internal “elites” who often serve asthe guardians of important organizational values.The theory emphasizes that these elites often re-quire substantial autonomy if they are to effectivelyperform their value-protecting function (Clark,1956; Selznick, 1957: 94, 120–133). Selznickstressed that these groups’ “distinctive aims andstandards” are apt to be undermined if they aresubjected to strong pressures from their organiza-tion’s operative system, or from each another(1957: 120).

These arguments shed much light on the insti-tutional threat posed by EM, and particularly onthe problems created by the consolidation of fi-nancial aid and admissions units. The autonomytraditionally granted to the former allowed finan-cial aid personnel to perform a set of highly cir-cumscribed duties (evaluating students’ financialneed and allocating commensurate levels of aid)in the pursuit of a well-defined and essentiallycharitable purpose: promoting equal access tohigher education. The latter unit’s autonomylikewise enabled admissions professionals to fo-cus on upholding their own “distinctive aims andstandards”: evaluating the merits of applicantsand making fair and independent admissionsdecisions.

The consolidation of the two units clearly putsthese values and competencies at risk. The intent ofthe enrollment management structure is to enable

5 Institutionally funded financial aid comes from acollege’s own funds. It is distinct from other, more tradi-tional forms of financial aid, including governmentgrants, foundation-supported scholarships, and subsi-dized loans.

6 The tuition discount rate is the percentage of totaltuition revenues that a college gives back to students inthe form of aid. The specific calculation is total institu-tional aid/total tuition revenues. This is a key metric inthe enrollment management field.


financial aid decision makers to allocate scarce in-stitutional resources in the service of enrollmentand revenue goals. However, its frequent effect is tovenerate these technical goals and to require thatthey be taken into primary account. This require-ment is anathema to the financial aid function’saforementioned charitable purpose. Higher educa-tion economist Michael McPherson has succinctlysummarized this transformative effect: “It used tobe, providing [financial] aid was a charitable oper-ation. Now it’s an investment, like brand manage-ment” (quoted in Kirp & Holman, 2002: 30). Thethreat to established values appears to run in theopposite direction as well. Under EM, admissionsprofessionals also lose autonomy and are oftencompelled to heavily weight financial consider-ations (such as a student’s ability to pay full tu-ition) as they evaluate applications. These new,structurally induced pressures have led to a num-ber of apparent abuses that have subjected bothenrollment management practitioners and Ameri-can higher education more generally to significantpublic criticism in recent years (Baum, 2000; Gau-diani, 2000; Green, 2005; Hossler, 2004; Kirp, 2003;Quirk, 2005).7

It would be naıve to think that financial consid-erations never affected admissions decisions beforeEM came into wide use, and equally wrong toblame it for all current ills. Nevertheless, it is clearthat the new structure takes necessary autonomyaway from the admissions and financial aid func-tions and thus greatly decreases the probability thattheir respective values will be realized.8 It is alsoclear that many of the threats latent in the EM

structure itself have become manifest in the wakeof its diffusion. Among these various manifesta-tions, the most notable (and widely noted) is anepic shift in the pattern of financial aid allocationthat has occurred over the past two decades(Ehrenberg, Zhang, & Levin, 2006; Heller, 2004;McPherson & Schapiro, 1998). Over this period,colleges and universities have, in the aggregate,diverted billions of dollars in financial aid awayfrom the neediest students and toward those fromhigher socioeconomic strata. Researchers, critics,and administrators have directly attributed theseshifts to enrollment management. Lapovsky andHubbell conducted a thorough review of the ex-tensive research on this topic and provided anapt summary of its collective findings: “The en-rollment management approach has reduced ed-ucational access to the economically disadvan-taged while providing financial subsidies tothose with the ability to pay” (2000: 19).

Power shifts to the operative system. In addi-tion to undermining the autonomy of a college’svalue-bearing elites, enrollment management alsoalters the balance of power between its normativeand operative systems. Selznick identifies a deep-rooted and persistent tension between these mutu-ally dependent but often antagonistic subsystemsand actually uses universities as an example inmaking this tension clear. He specifically warnsagainst confusing “organizational achievement”(operative) with “institutional success” (normative)and notes that a university should not allow “re-sources, stability, or reputation to become the cri-terion of [its] success. . . . A university led by ad-ministrators without a clear sense of the values tobe achieved,” he explains, “may fail dismally whilesteadily growing larger and more secure”(1957: 27).

Viewed against this theoretical backdrop, EM ap-pears as a clear accommodation to the goals andinterests of the operative system of a college.Though the innovation is “mundane” in the sensethat it is intended to solve prosaic, administrativeproblems and is largely indifferent to establishedpurposes and values, it is significant in the sensethat it greatly increases the power and resourcesthat are made available for the pursuit of operativesystem goals. The creation of a cabinet-level enroll-ment management vice presidency represents animportant commitment to the practice (and practi-tioners) of enrollment management. This commit-ment has both political and financial implications.

7 Although many enrollment management practiceshave been criticized, some of the more egregious onesare: (1) financial aid “leveraging,” whereby schools in-tentionally allocate scarce institutional aid resources tostudents with a higher ability to pay and thereby maxi-mize their “net tuition revenues,” (2) so-called “admitdeny” decisions wherein schools accept qualified high-need students but offer financial aid packages so low asto deliberately discourage matriculation, and (3) the com-mon practice of taking a student’s expressed intention toenroll at a college into account in making financial aidoffers (so as to avoid “wasting” financial aid on an ap-plicant who is sure to enroll anyway) (Kirp, 2003;McPherson & Schapiro, 1998; Quirk, 2005). All of thesepractices are made possible by the breakdown of thebarrier between the admissions and financial aidfunctions.

8 We hope that the significance of this structuralchange is by now evident; still, one can acquire a greaterappreciation for its gravity by imagining the conse-quences if a college altered its internal “constitution” soas to make its faculty directly responsible for meeting

revenue goals, in addition to fulfilling their academicduties.


Although these may not be fully apparent to a col-lege’s faculty or top administrators at the time ofadoption, they are precisely the sort of institutionalthreat that Selznick urged institutional analysts(and would-be institutional leaders) to pay atten-tion to.9

These theory-based concerns about the institu-tional threat posed by this innovation also appearto be borne out by studies of its longer-term finan-cial and political consequences. Higher educationscholars have documented dramatic increases inenrollment management spending over the last 20years. By 2005, the average private college devotedover 26 percent of its total annual budget to insti-tutional aid expenses (the primary component ofenrollment management spending), up from only11 percent in 1987 (Thomas, Tan, & Hoffman, 2006;Wellman, Desrochers, & Lenihan, 2008). Thesenumbers are telling for at least two reasons. First,they suggest that colleges have grown their budgetsfor this practice by taking resources away fromother core areas. As Hossler has noted, “Providingmerit aid and discounting tuition make it hard forinstitutions to invest enough money in their corebusiness: educating students” (2004: B4). Redd(2000) provided direct, organization-level evidenceof this value-undermining effect, showing that mis-sion-critical budget categories such as instructionand academic support tend to be shorted as EMspending increases. Second, the numbers also pro-vide strong (if indirect) evidence of enrollmentmanagers’ growing strategic influence. Decades oforganizational research and theory on the sourcesof intraorganizational power suggest that EM prac-titioners’ ability to influence organizational policyis likely to grow in direct proportion to their shareof an institutional budget (Hickson, Hinings, Lee,Schneck, & Pennings, 1971; March & Simon, 1958;Pfeffer & Salancik, 1978).10

Market values penetrate the organization. Thepreceding paragraphs suggest that prosaic adminis-trative innovations can subvert values by disrupt-ing an organization’s internal normative and polit-

ical order. However, these same innovations mayalso threaten institutional values through more ex-ogenous and ideational processes. Enrollment man-agement is, in the final analysis, a structural changethat takes place in a particular organization. But itis also a partial manifestation of a larger and morepervasive market logic, one that has affected col-leges and universities in many different ways. Theongoing spread of this logic throughout higher ed-ucation has been widely observed (Bok, 2004; Gei-ger, 2004; Kirp, 2003; Washburn, 2005), and it par-tially parallels similar field-level change processesthat have been uncovered in previous neoinstitu-tional research (Lounsbury, 2007; Scott, Ruef, Men-del, & Caronna, 2000; Thornton & Ocasio, 1999;Washington & Ventresca, 2004).

The connection of EM to the market logic is nei-ther incidental nor trivial. Marketing conceptsdeeply influenced the pioneers of the innovation,and they conceived enrollment management in anexplicit attempt to apply these ideas and tech-niques to higher education (Hossler, 1984; Inger-soll, 1988; Kemerer, Baldridge, & Green, 1982). Thecontemporary tendency of EM practitioners to ap-ply business metaphors to the academic contexthas also been widely noted and criticized. Studentsare often recast as “customers,” financial aid as a“marketing expense,” organizational identity as a“brand,” and education itself as a “product” (Kirp,2003; Quirk, 2005). Don Hossler, who was one ofthe architects and early advocates of EM, recentlysummarized the larger ideational shift: “In somerespects, enrollment management is simply onesmall indicator of the ascendancy of capitalism andthe extent to which the market metaphor has takenhold throughout the United States and the rest ofthe world” (2004: B3). The depth and impact of thiscultural change is perhaps even more succinctlycaptured in the words of one prominent enrollmentmanager, who matter of factly observed: “The ob-jective of the enrollment process is to improve yourmarket position” (quoted in Kirp, 2003: 12).

It is important (and perhaps encouraging) to notethat these new metaphorical understandings arenot uniformly embraced throughout higher educa-tion, or even in all the individual colleges that haveadopted enrollment management. In so much as amarket logic has penetrated American colleges anduniversities, its influence is often localized withintheir operative systems (administrations). Manyfaculty members have never even heard of thisinnovation, much less fallen under the spell of itsmaster metaphor. Nevertheless, EM’s strong con-nection to the market logic clearly indicates that itshould be understood as a culture-bearing innova-tion, as well as a politically disruptive one. Enroll-

9 “Institutional analysis asks the question: What is thebearing of an existing or proposed procedure on thedistinctive role and character of the enterprise? . . . Thequestion of whether a set of proposed administrativereforms endangers the maintenance of desired values isalways legitimate and necessary” (Selznick, 1957: 140).

10 A recent New York Times article suggests this stra-tegic influence. The article, “To Keep Students, CollegesCut Anything But Aid,” suggests that enrollment man-agement expenditures have remained constant or grown,even as colleges have made deep budget cuts in responseto the recent economic crisis (Zernike, 2009).


ment management matters both because it alters theintricate constitutional order of a college and be-cause it brings new ideas and values into that order.We attempted to bear this duality in mind in de-veloping our hypotheses and interpreting ourfindings.

HYPOTHESES

Having explained how enrollment managementthreatens organizational values, we now turn to ourcore task of identifying the “conditions and pro-cesses” that made colleges more or less vulnerableto this threat. Because we have characterized EM asan incipient threat, we attempt to identify factorsthat are likely to have affected colleges’ interpreta-tion of the innovation, as well as those affecting theinternal political process surrounding its ultimateadoption/rejection. Because we have characterizedEM as both a distinctly organizational change and amanifestation of a larger field-level trend, we locatecausal explanations both inside an organizationand in its broader cultural and economic contexts.

Organizational and Local Environmental Factors

Power and professionalism of internal elites.As we have already seen, power and professional-ism are the two main ingredients in Selznick’s rec-ipe for avoiding goal displacement and value atten-uation. He portrayed professionalism as “the basicanswer society has evolved for the protection ofinstitutional integrity” (1957: 40). In his “constitu-tional” theory, he posited that organizations caneffectively offset the short-term, technical pres-sures that continually emanate from their operativesystems by granting power and autonomy to inter-nal professional groups (Knudsen, 1995). In so do-ing, Selznick argued, an organization creates “irre-versible commitments” to these groups’ respectivevalues and effectively builds them into its socialstructure.11

Building on this general argument, we predictthat organizations are at lower risk of adoptingvalue-subverting management innovations whenthe internal elites who stand to be affected by themare relatively powerful and highly professional-ized. As the preceding discussion revealed, facultyand admissions personnel are two internal groups

whose values and identities are threatened by theenrollment management structure. We propose thatcolleges are less vulnerable to its adoption to theextent that these groups are empowered and de-monstrably committed to the values of their respec-tive professions.

Hypothesis 1. Colleges that have more power-ful and professionalized faculties are lesslikely to adopt the enrollment managementstructure.

Hypothesis 2. Colleges that have more power-ful and professionalized admissions personnelare less likely to adopt the enrollment manage-ment structure.12

Autonomy from external resource pressures.An organization should also be less likely to adoptvalue-subverting innovations when it is relativelyindependent from external resource pressures. Sev-eral of the classic studies in the early institutionaltradition identified dependence on external re-source providers as the prime cause of value atten-uation (Clark, 1956; Messinger, 1955; Zald & Den-ton, 1963). Clark’s (1956) study of goaldisplacement in adult education organizations(which is discussed at length in Selznick’s 1957book) is perhaps the most noteworthy example. Hefound that these organizations’ extreme depen-dence on students and their tuition set in motion anadaptation process that ultimately led them towholly abandon their missions and to devolve into“educational service stations” offering classes onany topic that would fill their classrooms (includ-ing “cake decorating, rug making, and square-danc-ing”[Clark, 1956: 334]). Scholars working in other,more contemporary research traditions have alsorecognized the practical need for organizational au-tonomy and the potentially dire consequences ofextreme external dependence (Mitchell et al., 1997;

11 “A wise management will readily limit its own free-dom, accepting irreversible commitments, when the ba-sic values of the organization and its direction are atstake” (1957: 40). Faculty tenure is one obvious exampleof this type of irreversible commitment.

12 Although these predictions are well-grounded inSelznick’s theory, it is necessary to qualify them in twoways. First, we should note that a strictly power-orientedview might lead to the prediction that admissions per-sonnel would favor the EM structure because it appearsto provide them with increased power and centrality,even as it undermines their profession’s traditional val-ues (which are preeminent in Selznick’s normative the-ory). Second, we should also note that faculty power andprofessionalism are likely to exert an indirect effect onadoption. Because EM is a largely technical, administra-tive structure, faculty may not be directly involved in theadoption decision. Nonetheless, we expect that facultyvalues and concerns will be more salient to administra-tors when faculty power and professionalism are higher(Mitchell, Agle, & Wood, 1997).


Oliver, 1991; Pfeffer & Salancik, 1978; Thompson,1967).

We posit that autonomy from external pressuresis particularly important in understanding colleges’ability to resist enrollment management. Becausethe innovation is specifically designed to providecolleges with more control over their enrollmentsand revenues, it is likely to be very alluring toschools with an obvious deficiency in this area. Incontrast, colleges that occupy a stronger positionvis-a-vis their would-be students might be expectedto take a more principled and long-term view incontemplating the structure’s adoption. Two tellingindicators of a college’s relative autonomy fromenrollment pressures are: (1) lower dependence ontuition as source of revenue and (2) geographicdispersion of its student body. Colleges with signif-icant nontuition revenue streams (e.g., from en-dowment and grant income) and national “drawingpower” are thus predicted to be at a lower risk ofEM adoption.

Hypothesis 3. Colleges that are less depen-dent on tuition as a source of revenue are lesslikely to adopt the enrollment managementstructure.

Hypothesis 4. Colleges that have greater stu-dent drawing power are less likely to adopt theenrollment management structure.

Organizational history and character. An or-ganization’s historically accreted “character” is athird factor likely to affect its vulnerability to theadoption of value-threatening management innova-tions. Selznick’s theory emphasizes the need tostudy organizations through a “developmental”and diachronic lens (Knudsen, 1995; Kraatz &Block, 2008; Selznick, 1957). Seen through thislens, current organizational actions and decisionsappear in the context of a stream of prior choicesand actions made throughout the organization’s lifecourse. Selznick hypothesized that an organizationdevelops a discernible character as the result of thecritical and identity-forming decisions made overtime (1957: 38–56). He also emphasized that thishistorically accreted character can be an importantresource for fending off value subversion (or a keyliability, depending on its content).

Because our study employs neoinstitutionalmethods rather than the case study approach of theearly institutionalists, we lack the ability to assesscharacter in the deep and multifaceted sense inwhich Selznick conceptualized it. We are, how-ever, able to observe some specific and plausibly“character-defining” choices that our sample col-leges made at earlier in their developmental histo-

ries. During the 1970s and 1980s, a majority ofliberal arts colleges adopted professional degreeprograms in response to changing student degreepreferences and enrollment and budgetary pres-sures (Kraatz & Zajac, 1996). These changes werecritical ones because they were manifestly incon-sistent with these colleges’ traditionally espousedvalues (and very identities). We propose that col-leges that adopted professional programs in re-sponse to this earlier crisis were more vulnerable tothe adoption of enrollment management during oursubsequent study period. We posit two causalmechanisms underlying this proposed effect. Thefirst is explicitly Selznickian and character-based.Professional program adoption is similar to enroll-ment management in that it reflects a defectionfrom principle and an apparent capitulation to ex-ternal market pressures. It is plausible to argue that“going along” with the former set of pressurespaved the way for a college to conform to the cur-rent ones. The second mechanism is more ide-ational and cultural. Specifically, it is possible thatprior adoption of professional programs created afoothold for the market logic and the ideology ofstudent consumerism—sensibilities that are alsomuch apparent in the structure and practice ofenrollment management (Hossler, 2004; Kirp,2003). Following this logic, the EM innovationwould likely be less culturally offensive to collegesthat had previously adopted professional programs.

Hypothesis 5. Colleges that adopted profes-sional programs in the 1970s and early 1980sare more likely to have adopted the enrollmentmanagement structure in our subsequent studyperiod.

Leadership. Leadership is the final organization-level factor in our theoretical model of vulnerabil-ity to enrollment management adoption. We havefollowed Selznick in portraying value subversionas a primarily sociological phenomenon, one that isboth impelled and impeded by social-structuralconditions in and around an organization. How-ever, we also follow him (and the other “old” insti-tutionalists) in reserving a prominent place for hu-man agency and error and in seeing individualpeople as the ultimate “guts of institutions.”13

13 “The guts of institutions is that somebody some-where really cares to hold an organization to the stan-dards and is often paid to do that. Sometimes that some-body is inside the organization, maintaining itscompetence. Sometimes it is in an accrediting body,sending out volunteers to see if there is really any algebrain the algebra course. And sometimes that somebody, or


Our particular concern is with organizations’ topadministrators and their ability to discern and re-sist the incipient threat posed by enrollment man-agement and other like innovations. In Selznick’saccount, institutional leaders are the ultimate guar-antors of values and the last defense against threatsto institutional integrity. Selznick argued that insti-tutional leaders needed to assume a posture of“statesmanship,” using their authority to protecttheir organizations’ core values and long-term in-terests. He also emphasized that this work requireddeep commitment, foresight, and knowledge of anorganization’s values, commitments, and internalsocial structure.14 Although our methodology lim-ited our ability to assess college presidents’ values,competencies, and commitments in depth, we wereable to assess some important aspects of their insti-tutional and personal experience that serve as re-vealing proxies thereof. We propose that colleges’vulnerability to EM varies systematically as a func-tion of these prior social experiences.

Our first leadership hypothesis centers on a col-lege president’s experience within her or his insti-tution itself. Specifically, we posit that collegeswith longer-tenured presidents are less likely toadopt the enrollment management structure. Weexpect that longer experience in a college (and,more specifically, as its president) makes a leadermore attuned to institutional values and commit-ments, more aware of organizational complexitiesand interdependencies, and more wary of newmanagement innovations that offer easy solutionsto perennial problems. We also expect longer-ten-ured leaders to have a greater inclination and abil-ity to resist change, in general (Miller, 1991;Wiersema, & Bantel, 1992). This entrenchment andconservatism are clearly pathological in somecases, but they may often serve a more useful func-tion. Selznick had much to say about the leaders’role in building institutions and enacting change,but he also observed that institutional leadershiprequired “an essentially conservative posture”(1957: 81).

Hypothesis 6. Colleges with longer-tenuredpresidents are less likely to adopt the enroll-ment management structure.

We also posit that institutional leaders’ extraor-ganizational experience bears on a decision toadopt EM. Specifically, we argue that schools aremore likely to adopt enrollment management whenthey are led by presidents who have worked atanother institution that has previously adopted thestructure. These leaders should be less normativelyopposed to the change and more likely to see it asan “appropriate” administrative device. They mayalso function as institutional carriers who createinterorganizational linkages and thus promote thediffusion of enrollment management and other ad-ministrative practices (Scott, 2008). Kraatz andMoore (2002) showed that presidential migrationplayed exactly this role in promoting professionalprogram adoption. They drew both from Selznick’stheory and neoinstitutionalism in explaining theirfindings.

Hypothesis 7. Colleges are more likely to adoptthe enrollment management structure whenthey are led by presidents who have mi-grated from previous adopters of enrollmentmanagement.

Macroenvironmental Factors

To develop a full understanding of organizations’vulnerability to value-subverting administrative in-novations, it is also necessary to look beyond theirboundaries and to see them in the context of theirbroader institutional and competitive environment.Although we follow Selznick in viewing an organ-ization as a self-governing entity that is the ultimateauthor of its own decisions, it is clear that “condi-tions and processes” operating at the interorganiza-tional and field levels of analysis often stronglyshape these choices. Our remaining hypotheses fo-cus on these macroenvironmental risk factors.

Field-level theorization and diffusion pro-cesses. We have already explained that enrollmentmanagement is, in some important sense, a mani-festation of a larger, field-level change process. Itmay appear to an individual college as a value-neutral administrative device, and it is probablymost often adopted in an effort to solve prosaic,operative system problems. Nevertheless, it is im-portant to recognize that this particular technicalsolution has been “theorized” and socially con-structed as an appropriate and efficacious one(Strang & Meyer, 1993). We see at least two causallysignificant field-level processes that are responsi-ble for its field-level legitimation. First, readily

his or her commitment, is lacking, in which case thecenter cannot hold, and mere anarchy is loosed upon theworld” (Stinchcombe, 1997: 17).

14 “The institutional leader cannot permit any partialviewpoint to dominate decisions regarding the organiza-tion as a whole . . . this control will not be not be possibleunless a true conception of the nature of the enterprise—its long-run aims as shaped by long-run commitments isgrasped and held” (Selznick, 1957: 81).


identifiable institutional actors or “carriers”(Greenwood & Suddaby, 2002; Scott, 2008) havestrongly advocated EM. A large number of consult-ing firms, for instance, now operate lucrative prac-tices that provide colleges with prepackaged EM“solutions” (Gose, 1999). A nascent EM professionhas also sprung up, further contributing to thestructure’s cultural standing (Henderson, 2001).Second, enrollment management’s rapid spreadduring our study period also suggests the strongpossibility of mimetic diffusion (DiMaggio & Pow-ell, 1983). The neoinstitutional literature suggeststhat the mere prevalence of a structure or practice isstrong evidence of its cultural acceptability andthat this “taken-for-grantedness” is often sufficientto further its continued spread (Suchman, 1995;Westphal, Gulati, & Shortell, 1997). The literatureon fads and “bandwagons,” which has shown thatmany management innovations spread throughprocesses of social contagion, also supports thisinsight (Abrahamson, 1991; Abrahamson & Rosen-kopf, 1993; Burns & Wholey, 1993; Fiol &O’Connor, 2003).

It is important to note that by hypothesizing afield-level diffusion process, we are not suggestingthat EM has gained uniform acceptance throughouthigher education. Neither are we arguing that col-leges have adopted it in an effort to display sym-bolic conformity with social norms and therebygarner public approval. As we have explained, EMhas been widely portrayed as a practice that actu-ally undermines public trust in higher education,and it has generated growing negative publicity asits consequences have become better understood. Itis also clear that market forces have played a majorrole in its spread (a topic we turn to next). What weare positing, instead, is only that this innovationhas become more culturally acceptable within thenarrower community of college administrators(who are typically responsible for making decisionsabout the appropriateness of various “mundane ad-ministrative arrangements”). We propose that thechanging norms of this narrower community willinfluence an individual college’s susceptibility toadoption of EM, irrespective of the structure’s(un)acceptability to other constituencies in the widerhigher education arena.

Hypothesis 8. Colleges are more likely to adoptenrollment management as the structure be-comes more prevalent at the field level.

Competitive interaction. Diffusion processes areknown to be driven by competitive forces as well asby institutional ones (Abrahamson & Rosenkopf,1993; Lieberman & Asaba, 2006). Organizations re-spond not only to changing social norms and pre-

scriptions, but also directly to one another (Porac,Thomas, Wilson, Paton, & Kanfer, 1995; Smith,Grimm, Gannon, & Chen, 1991). These responses,although still imitative in the broadest sense of theterm, are sometimes more accurately understood asdefensive or even retaliatory. It is likely that thesecompetitive interactional dynamics also play a keyrole in driving organizations to adopt value-sub-verting innovations.

Competitive interaction appears to be particu-larly important in explaining college’s relative vul-nerability to enrollment management adoption.This is the case because the structure is, amongother things, a tool that schools use in the deliber-ate effort to attract students and their tuition dollarsaway from directly competing institutions. Theterm “arms race” is very frequently used in describ-ing the contemporary landscape of college admis-sions, and enrollment management is widely citedas the major weapon in this continually escalatingcompetition (Davis, 2003; Goral, 2003; McPherson& Schapiro, 1998; Quirk, 2005). Many administra-tors have suggested that their colleges have beenforced into EM as a direct result of the aggressiveactions undertaken by their competitors. Some ad-ministrators have also complained that they areunable to “unilaterally disarm,” given their com-petitors’ continued use of EM weaponry (Quirk,2005). These external attributions may be some-what self-serving, but they nonetheless point to astraightforward hypothesis. Specifically, a collegeshould be more vulnerable to the adoption of EMwhen its direct competitors have adopted the struc-ture and are heavily engaged in its practice.

Hypothesis 9. A focal college is more likely toadopt the enrollment management structurewhen more of its direct competitors haveadopted the structure and when its directcompetitors are heavily engaged in tuitiondiscounting.

Middle status. An organization’s position in theoverall status hierarchy of its field is a final con-textual factor likely to affect its vulnerability to theadoption of value-subverting management innova-tions. Organizational sociologists have noted thatmany industries and fields have clear status hier-archies that exert important effects on organization-al actions and outcomes (Jensen, 2003; Lounsbury,2002; Podolny, 1993, 2005; Washington & Zajac,2005). They have also noted a strong tendency to-ward “middle-status conformity” that appears to bedriven by the social insecurity and “status anxiety”of organizations that occupy positions in the mid-dle of a status hierarchy (Jensen, 2006; Phillips &Zuckerman, 2001). Because these middle-status or-


ganizations face both the prospect of status ad-vancement and the perpetual risk of marginaliza-tion, they are thought to “work feverishly tosolidify their social standing” (Phillips & Zucker-man, 2001: 382).

Building on these arguments, we propose thatmiddle-status colleges are the most vulnerable toenrollment management adoption. Higher educa-tion is well known to be a highly stratified field,and status concerns have become increasingly sa-lient as the result of the college-ranking schemesthat emerged at the outset of our study period. Wepredict that EM is particularly alluring to middle-status schools because it provides tools and tech-niques that promise status protection and/or ad-vancement (e.g., by allowing a school to increasethe average test scores of incoming students andthus enhance its rankings). Our argument differssignificantly from the classic middle-status confor-mity hypothesis, in that enrollment management isa means to the end of status enhancement, ratherthan a status marker in its own right. (As previouslynoted, EM’s symbolic connotations do not appearto be positive.) Nevertheless, the predicted patternof adoption is the same.

Hypothesis 10. Middle-status colleges are morelikely to adopt the enrollment managementstructure.

Diminishing effects of macroenvironmentalpressures. Because value-subverting innovations(by definition) exert undesired consequences foradopting organizations, we expect their diffusion tofollow an atypical temporal trajectory. Specifically,we propose that the macrocontextual drivers ofthese innovations become less influential over timeas an innovation’s full effects become increasinglywell understood and widely publicized. This pro-posed pattern, which is essentially a process offield-level learning (Miner & Haunschild, 1995),contrasts sharply with the more typical trajectory,wherein macroinstitutional forces gain causal pre-cedence as an innovation spreads widely and be-comes fully taken-for-granted at the field level (Tol-bert & Zucker, 1983; Westphal et al., 1997). Thisalternative trajectory appears particularly likely tounfold in our study context. Although relativelylittle was written about enrollment managementduring the early part of its diffusion, it has receivedmuch more scrutiny and criticism in recent years.Notably, all of the critical research and commen-tary that we have reviewed above was published inthe second (post-1996) half of our 20-year studyperiod. Thus:

Hypothesis 11. The macrocontextual drivers ofenrollment management identified in Hypoth-eses 8–10 had less effect on enrollment man-agement adoption after 1996.

METHODS

Sample and Data

Our sample is the complete population of privateliberal arts colleges as defined by the CarnegieFoundation for the Advancement of Teaching(1987). We studied these 515 schools over the pe-riod 1987–2006. Liberal arts colleges are value-in-fused organizations that espouse and self-con-sciously pursue social missions. They have beenpreviously described as archetypal Selznickian in-stitutions (Clark, 1970; Kraatz & Zajac, 1996) andare known to have been deeply affected by enroll-ment management (McPherson & Schapiro, 1998).They are also very well established, with an aver-age age of over 100 years at the outset of our studyperiod. Our data come from three separate sources.The main source is the Higher Education Directory,which annually publishes the names and titles ofvirtually all administrators at American colleges.The second is the federal government’s IntegratedPost-Secondary Education Data System (IPEDS)and its predecessor, the Higher Education GeneralInformation Survey. These provide annual data onthe finances, enrollments, faculty, curricula, andother institutional characteristics of our sampleschools dating back to the early 1970s. The third isPeterson’s Guide to Four-Year Colleges (Peterson’sGuides, 1987), from which we obtained our statusmeasure.

Measures

Dependent variable. Our study focuses on ex-plaining colleges’ vulnerability to the initial adop-tion of the enrollment management structure. Weinferred this adoption when a college first reportedthe existence of an administrative position with thetitle “vice president for enrollment management”or some close approximation thereof. Focusing onthe initial adoption of the structure was theoreti-cally appropriate, given our interest in understand-ing organizations’ ability to recognize and rejectincipient threats to their values. It was also appro-priate in light of our strong theoretical interest in“administrative arrangements” (i.e., structures).Previous research has shown that colleges with EMstructures are, on average, much more deeply en-gaged in the practice than those without the struc-


tures, and our own data strongly support thisconclusion.15

Independent variables. We assessed the powerand professionalism of a college’s faculty (Hypoth-esis 1) using the following three measures: Averagefaculty salary (standardized within years to controlfor inflation), tenure-track faculty to student ratio,and research budget.16 Faculties that are relativelybetter paid, more permanent, larger, and better sup-ported in their academic pursuits are, ceteris pari-bus, expected to be more powerful and professionalones. We gauged the power and professionalism ofschools’ admissions directors (Hypothesis 2) usingdata about their titles and lengths of tenure. Asignificant percentage of admissions directors(20%) had the title “dean of admissions.” We tookthis as evidence of greater authority and profession-alization and thus included a has admissions deanvariable. We also included admissions officer ten-ure in testing Hypothesis 2, with the expectationthat a director’s influence would increase with timein office.

We assessed a school’s autonomy from enroll-ment pressures (Hypotheses 3 and 4) using twomeasures. Tuition independence (Hypothesis 3)captured the percentage of a school’s total reve-nues that come from sources other than tuition(calculated as [total revenues � tuition reve-nues]/total revenues). We measured the school’sstudent drawing power (Hypothesis 4) withgeographic dispersion of student body. We calcu-lated this variable using data from the IPEDS“residency survey,” which requires schools toreport the home states of their incoming students.Specifically, we created a Herfindahl index that

measured the geographic concentration of aschool’s student body (across all 50 states). Wethen reverse-coded this variable so that highervalues would reflect greater dispersion (which isthe obverse of concentration).

We included the variable professional pro-grams previously adopted to assess the effects oforganizational character and history (Hypothesis5). This variable indicates the number of pro-grams that a school adopted between 1971 and1985. To test Hypothesis 6 (about presidentialtenure) and Hypothesis 7 (about the effects ofpresidential migration), we included the self-ex-planatory measures presidential tenure and pres-ident from school with enrollment managementstructure, respectively.

We tested Hypothesis 8 by including cumulativeprior adoption of enrollment management at thefield level. This variable, which specifically cap-tures the overall percentage of liberal arts collegesthat had adopted the structure as of time t � 1, is astandard measure in institutional diffusion studies.

Testing Hypothesis 9, which focused on the ef-fects of competitive interaction, required us toidentify a unique set of direct competitors for eachindividual college. We identified these schools us-ing three criteria: geographic overlap, tuition simi-larity, and similarity in selectivity. Colleges liter-ally compete in geographic space (by vying for thesame students in the same locations). Schools thatare highly similar on the other two dimensions arenot competitors if they get their students from op-posite ends of the country. Thus, we defined aschool’s group of direct competitors as the fiveother schools that had the highest degree of geo-graphic overlap and were also within one standarddeviation of a focal school on either of the othertwo dimensions. We limited the number to fivebecause of our interest in the effects of direct com-petition, and so that each school would have acompetitor group of the same size. Other cutoffsproduced similar results. The actual measures usedto test Hypothesis 9 were the percentage of directcompetitors previously adopting the enrollmentmanagement structure, and mean tuition discountrate of direct competitors. As explained, tuitiondiscounting is the core practice of EM.17

15 The consulting firm Noel-Levitz conducts a biennialsurvey of enrollment management practices in a repre-sentative sample of American colleges and universities.Their 2002 survey showed that colleges with enrollmentmanagement vice presidents had an average tuition dis-count rate of 31 percent, institutional aid expenditures of$1,174 per full-time-equivalent (FTE) student, and totalrecruiting costs of roughly $600 per FTE student. Thenumbers for schools without the structure were 22 per-cent, $571, and $403, respectively (Noel-Levitz, 2002).Our 2006 data show that liberal arts colleges with theenrollment management structure spent 29 percent oftheir total annual budgets on institutional aid, comparedto 23 percent for those without.

16 We used an indicator variable (0, 1) to capture theeffects of research budget. Because roughly two-thirds ofour sample colleges provided no research funding, mostof the variance in this measure is between the “haves”and “have nots.” Results are highly similar to those ob-tained using a continuous measure of research spending.

17 Geographic overlap was assessed using data fromthe IPEDS residency survey. For every pair of schools (iand j), we first created a state-level measure of overlap bycomparing the percentages of i’s and j’s enrollments thatcame from that state and retaining the lesser of the twovalues. (Examples: If i gets 100 percent from Alabamaand j gets 0, their overlap in Alabama is 0. If i gets 100


To test Hypothesis 10 (regarding the effects ofstatus position), we relied on a five-point measureof a college’s overall selectivity in admissions,which was obtained from Peterson’s Guide to Col-leges. Because status is relatively stable, we usedthe 1987 measure throughout. We included asquared term (selectivity squared) to test the morespecific argument that middle-status colleges arethe most vulnerable to enrollment managementadoption. We tested Hypothesis 11 (which positsthat contextual influences have a declining influ-ence in later years) by first creating a late perioddummy variable and then interacting it with thevariables used to test Hypotheses 8–10. The lateperiod was defined as the latter half of our 20 yearstudy period (1997–2006).

Control variables. We included control vari-ables that captured the size (undergraduate enroll-ment) and tuition levels of our sample colleges.Because tuition rates greatly increased over thestudy period, we standardized the latter measureby year.

Treatment of missing data. Because the federalgovernment did not release finance data from 1997through 2001, data for three of our key variables(discount rate of direct competitors, tuition inde-pendence, and research budget) were missing forthese years. We filled in these missing values usinglinear interpolation between the 1996 and 2002values. Other methods of treating these missingvalues produced similar results. Data on the tenureof admissions officers were also missing in approx-imately 20 percent of all cases. We treated thesemissing values by substituting zeroes and includ-ing a dummy variable (admissions tenure missing)to denote these observations. This allowed us toexamine tenure effects in cases in which data wereavailable, without dropping the entire observationwhen they were not. We were unable to imputethese missing values because they were spread ran-domly over years and observations. Models exclud-

ing the admissions tenure variable produced simi-lar results.

Analyses

We modeled enrollment management adoptionusing discrete time event history analysis (Allison,1984; Yamaguchi, 1991). Discrete-time analysis es-timates logit models of dichotomous outcomes forpooled time series data in which the same units areobserved at multiple intervals. These logit modelsallow one to estimate the hazard of an event (e.g.,enrollment management adoption) occurring inany one of the discrete time periods. Covariates areallowed, but not required, to vary between timeperiods. The discrete time model has the followinggeneral form:

log[P(t)/(1 � P(t)] � a(t) � b1 x1 � b2 x2(t),

where log[P(t)/(1 � P(t)] represents the log-odds ofthe event occurring for a particular unit at any timet, a represents the baseline hazard of the eventoccurring at any time t, b1 represents the change inthe log-odds for each one-unit increase in a time-invariant covariate x1, and b2 represents thechange in the log-odds for each one-unit increase ina time-varying covariate x2(t).

RESULTS

Descriptive statistics and correlations appearin Table 1. The results of tests of Hypotheses1–10 are presented in Table 2. Hypotheses 1 and2, which invoke arguments about the power andprofessionalism of internal value “guardians,”were both supported. Colleges were less likely toadopt enrollment management when they hadhigher average faculty salaries, higher tenuretrack faculty to student ratios, research budgets(Hypothesis 1), and admissions deans (Hypothe-sis 2). The effect of the admissions dean variablewas marginal. This may be because admissionsprofessionals have mixed motives regarding en-rollment management (an issue that we discussin a footnote to Hypothesis 2).

Table 3 also reveals support for Hypotheses 3and 4, both of which emphasize that organiza-tional autonomy from enrollment pressuresmakes colleges less susceptible to EM adoption.We found that colleges were less likely to adoptEM when they were more tuition-independent(Hypothesis 3) and had greater student drawingpower, as evidenced by the geographic disper-sion of their student bodies (Hypothesis 4).

and j gets 50, their overlap is 50.) We then summed thestate-level overlaps for i and j across all 50 states inwhich the two schools could hypothetically compete.The resultant measure of overall geographic overlapranged from 0 to 1. Importantly, high overlap values werefound both in dyads in which the two schools got alltheir students from the same 1 or 2 state(s) and in thosein which the two schools competed nationally (e.g., eachgets 5 percent from the same 20 states). This measure is,in itself, very effective in identifying plausible competi-tor groups for both regional and national colleges. Theaddition of our tuition similarity and selectivity similar-ity variables further increased its validity and predictivepower.


TA

BL

E1

Des

crip

tive

Sta

tist

ics

and

Cor

rela

tion

s

Var

iabl

eM

ean

s.d

.1

23

45

67

89

1011

1213

1415

1617

1819

1.A

dop

tion

ofen

roll

men

tm

anag

emen

tst

ruct

ure

0.03

0.18

2.A

vera

gefa

cult

ysa

lary

a�

0.03

0.98

�.0

33.

Ten

ure

-tra

ckfa

cult

yto

stu

den

tra

tio

0.12

0.07

�.0

5.1

04.

Res

earc

hbu

dge

t(0

.1)

0.30

0.46

�.0

4.5

1.1

15.

Has

adm

issi

ons

dea

n0.

190.

39�

.02

.29

.02

.20

6.A

dm

issi

ons

offi

cer

ten

ure

3.76

4.59

.00

.23

�.0

3.1

6.2

57.

Tu

itio

nin

dep

end

ence

40.3

816

.64

�.1

0�

.10

.20

.02

�.0

5�

.07

8.G

eogr

aph

icd

isp

ersi

onof

stu

den

tbo

dy

�0.

410.

28�

.04

.40

.18

.35

.18

.13

.12

9.P

rofe

ssio

nal

pro

gram

sp

revi

ousl

yad

opte

d1.

431.

19.0

5�

.19

�.1

8�

.09

�.0

9�

.06

�.0

7�

.21

10.

Pre

sid

enti

alte

nu

re6.

075.

40�

.04

.07

�.0

8.0

1.0

0.1

3�

.10

�.0

6�

.04

11.

Pre

sid

ent

from

sch

ool

wit

hen

roll

men

tm

anag

emen

tst

ruct

ure

0.02

0.13

.06

�.0

3�

.01

�.0

2�

.03

.00

�.0

3�

.02

.02

�.1

4

12.

Cu

mu

lati

vep

rior

adop

tion

atfi

eld

leve

l21

.03

13.9

0.0

9.0

5.0

0.0

7.1

3.0

9�

.35

.05

�.0

6.0

7.0

9

13.

Per

cen

tage

ofd

irec

tco

mp

etit

ors

adop

tin

g22

.11

23.9

7.0

9�

.14

�.0

3�

.09

�.0

2�

.02

�.2

6�

.14

.10

.08

.05

.60

14.

Mea

ntu

itio

nd

isco

un

tof

dir

ect

com

pet

itor

s26

.05

7.42

.06

.14

�.0

2.1

5.1

8.1

4�

.23

.21

�.0

2.0

3.0

5.6

3.4

0

15.

Sel

ecti

vity

1.70

0.90

�.0

1.6

6.1

0.3

5.2

5.1

5�

.07

.34

�.1

2.0

0�

.03

.00

�.1

3.1

516

.S

elec

tivi

tysq

uar

ed3.

712.

94�

.03

.72

.13

.40

.26

.17

.01

.39

�.2

1�

.01

�.0

2.0

2�

.15

.17

.92

17.

Un

der

grad

uat

een

roll

men

t1,

089.

6064

8.68

.01

.54

�.1

8.4

4.1

3.1

9�

.28

.17

�.0

2.0

8�

.02

.16

.04

.18

.27

.34

18.

Tu

itio

na

�.0

031.

00�

.01

.78

.10

.50

.32

.20

�.2

3.5

0�

.22

.02

�.0

2.0

2�

.18

.20

.69

.72

.41

19.

Ad

mis

sion

ste

nu

rem

issi

ng

0.23

0.42

.01

�.1

3�

.06

�.1

3�

.23

�.4

0.0

0�

.09

.06

�.0

6.0

1.0

2.0

7�

.01

�.1

0�

.10

�.0

7�

.13

20.

Lat

ep

erio

d0.

500.

50.0

6.0

4.0

0.0

8.1

0.0

7�

.33

.05

�.0

6.0

5.0

8.8

8.5

3.5

5.0

0.0

2.1

4.0

2.0

4

aS

tan

dar

diz

edw

ith

inye

ar.

Colleges that had previously adopted profes-sional programs were found to be at greater riskof enrollment management adoption, supportingHypothesis 5’s arguments about organizationalhistory and character. We also found support forHypotheses 6 and 7, both of which drew fromSelznick’s leadership arguments. Colleges wereless likely to adopt enrollment managementwhen they had longer-tenured presidents, andthey were more likely to do so when their leadershad migrated from an earlier adopter of thestructure.

Moving beyond the boundaries of individual or-ganizations, Hypothesis 8 posits that EM’s growingprevalence within the larger higher education field

positively affects an individual college’s propen-sity to adopt the structure. This hypothesis wassupported. Hypothesis 9 predicts that local, com-petitive pressures also increase vulnerability. Re-sults showed that colleges were more likely toadopt when more of their direct competitors hadthe EM structure, but the effect of our second mea-sure (direct competitor’s mean discount rate) didnot exert an independent effect. We also foundsupport for Hypothesis 10, which predicts thatmiddle-status schools will be particularly suscep-tible to EM adoption. This support is evident in thepositive coefficient for the selectivity variable andthe negative coefficient for the selectivity-squaredterm. Figure 1 presents visual evidence of the mid-dle-status effect, showing the percentage of schoolsin each status category that had adopted enrollmentmanagement by the end of the study period. Figure1 also reveals that no colleges in the highest statuscategory adopted EM during the study period.

Figure 2 shows the effect sizes for all variableswith significant effects in Table 3. Except wherenoted, the bars show the change in the odds ofadoption associated with a one standard deviationincrease in each variable, holding all others con-stant at their mean value. For dichotomous vari-ables, the bars show the effect of a one-unit increase(from 0 to 1).18

We now turn to Hypothesis 11, which posits thatthe contextual pressures identified in Hypotheses8–10 will weaken over time as EM’s consequencesbecame better understood and more widely publi-cized. Table 3 shows that this hypothesis was sup-ported for three of the four variables (the exceptionbeing the middle-status effect). An examination ofthe coefficients in Table 3 shows that two of thevariables merely declined in magnitude, but thethird (discount rate of direct competitors) actuallyreversed direction in the latter period. This reversallikely explains the variable’s insignificant effect inTable 2’s overall model. It may also reflect evidenceof vicarious learning from competitors. Althoughcompetitors’ increasing use of tuition discountinginitially spurred colleges to adopt EM, their contin-ually escalating financial aid expenses may have

18 An exponential term governs the odds of adoptionin a logistic regression equation. Accordingly, we calcu-lated the displayed change in the odds by multiplyingthe change in a variable by its coefficient from Table 3,then taking the exponential of the resultant product, andfinally subtracting 1 (i.e., �odds � eb(�x) � 1).

TABLE 2Antecedents of Enrollment Management Adoptiona

Hypothesis Variable Coefficient (s.e.)

1 Average faculty salarystandardized within year

�0.240* (0.135)

1 Tenure-track faculty to studentratio

�4.053** (1.699)

1 Research budget (0, 1) �0.456* (0.201)2 Has admissions dean �0.266† (0.184)2 Admissions officer tenure 0.008 (0.015)3 Tuition independence �0.011* (0.006)4 Geographic dispersion of

student body�0.641* (0.329)

5 Professional programspreviously adopted

0.104* (0.060)

6 Presidential tenure �0.054*** (0.014)7 President from school with

enrollment managementstructure

0.733* (0.328)

8 Cumulative prior adoption atfield level

0.021** (0.008)

9 Percentage of directcompetitors adoptingenrollment management

0.007* (0.004)

9 Mean tuition discount ofdirect competitors

0.005 (0.013)

10 Selectivity 0.677** (0.276)10 Selectivity squared �0.231** (0.097)

Undergraduate enrollment 0.0001 (0.0001)Tuition standardized within

year0.414** (0.114)

Admissions tenure missing �0.067 (0.193)Constant �4.672*** (0.552)

Chi-square 157.05***

a Observations � 6,493.† p � .10* p � .05

** p � .01*** p � .001Tests were two-tailed for control variables and one-tailed for

hypothesized effects.


been interpreted as more of a cautionary signal inlater years.19

19 In the interest of completeness, we also looked forevidence of time dependence in the effects of all of theother variables in our model, using the same early/late

approach. These analyses, which are available on re-quest, revealed no significant effects, with one notableexception. Specifically, the positive effect of the presi-dential migration variable dissipated between the earlyand late periods. This effect appears to cohere with the

TABLE 3Time-Dependent Effects of Contextual Variables on Enrollment Management Adoption

Hypothesis Variable Model 1 Model 2 Model 3 Model 4

Average faculty salarystandardized within year

�0.247* (0.135) �0.243* (0.135) �0.260* (0.135) �0.245* (0.136)

Tenure-track faculty to studentratio

�3.877* (1.697) �4.032** (1.702) �4.136** (1.700) �4.020** (1.701)

Research budget (0, 1) �0.437* (0.202) �0.423* (0.202) �0.436* (0.201) �0.418* (0.202)Has admissions dean �0.291† (0.184) �0.269† (0.184) �0.272† (0.183) �0.258† (0.184)Admissions officer tenure 0.009 (0.015) 0.007 (0.015) 0.013 (0.015) 0.009 (0.015)Tuition independence �0.012* (0.006) �0.013* (0.006) �0.012* (0.006) �0.013* (0.006)Geographic dispersion of student

body�0.607* (0.331) �0.605* (0.329) �0.593* (0.331) �0.623* (0.330)

Professional programspreviously adopted

0.108* (0.060) 0.106* (0.060) 0.104* (0.060) 0.102* (0.060)

Presidential tenure �0.056*** (0.014) �0.054*** (0.014) �0.055*** (0.014) �0.054*** (0.014)President from school with

enrollment managementstructure

0.723* (0.326) 0.740* (0.327) 0.727* (0.328) 0.741* (0.328)

Cumulative prior adoption atfield level

0.090*** (0.019) 0.038*** (0.012) 0.039*** (0.012) 0.041*** (0.012)

Percentage of direct competitorsadopting enrollmentmanagement

0.007* (0.004) 0.018** (0.006) 0.07* (0.004) 0.007* (0.004)

Mean tuition discount of directcompetitors

0.000 (0.013) 0.004 (0.013) 0.043** (0.017) 0.006 (0.013)

Selectivity 0.660** (0.276) 0.634* (0.276) 0.708** (0.276) 0.971* (0.436)Selectivity squared �0.227** (0.097) �0.217* (0.097) �0.238** (0.097) �0.317* (0.146)Undergraduate enrollment 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000)Tuition standardized within year 0.425** (0.146) 0.400** (0.145) 0.435** (0.145) 0.401** (0.145)Admissions tenure missing �0.060 (0.194) �0.047 (0.193) �0.025 (0.194) �0.033 (0.194)Late period 0.817† (0.507) �0.276 (0.333) 1.403 (0.678) �0.311 (0.507)

11 Late period � cumulative prioradoption at field level

�0.075*** (0.022)

11 Late period � percentage ofdirect competitors adoptingenrollment management

�0.016* (0.007)

11 Late period � mean tuitiondiscount of direct competitors

�0.076*** (0.023)

11 Late period � selectivity �0.525 (0.533)11 Late period � selectivity

squared0.159 (0.176)

Constant �5.249*** (0.581) �4.965*** (0.564) �5.743*** (0.623) �5.121*** (0.631)

Chi-square 174.11*** 167.35*** 173.42*** 163.23***Observations 6,493 6,493 6,110 3,757

† p � .10* p � .05

** p � .01*** p � .001Tests were two-tailed for control variables and one-tailed for hypothesized effects.


DISCUSSION

Master trends limit choice, but no outcome is inex-orable. They tell us what will happen if history isallowed to follow a line of least resistance. The mostimportant master trends are those that affect thehigh excellences of human life: success or failure inchild rearing, the arts, education, or government.The values that inform these activities are precari-ous, subject to attenuation or distortion by ambienttemptations. Precarious values need sustained nur-ture and special support. (Selznick, 2008: 26)

We began by noting the recent resurgence of in-terest in Selznick’s institutionalism and briefly re-viewing that theory’s distinctive premises, foci,and aims. Taking these assumptions and purposesas our own, we set out with two more specific goalsin mind. The first was to draw renewed attention to“mundane administrative arrangements” and theircontemporary role in enabling (or frustrating) or-ganizational value attainment. The second was tobegin to develop knowledge about the conditionsand processes that make organizations more or lessimpervious to the adoption of value-subverting ad-

ministrative innovations. We suggested that the lat-ter goal followed directly from the former and wasparticularly important given that colleges, hospi-tals, charities, government agencies, and othervalue-infused organizations routinely adopt pro-gressive management innovations that have largelyunexamined consequences for value realization.

We pursued these goals through a study of en-rollment management in liberal arts colleges. Draw-ing deeply from Selznick’s theory, we explainedhow the EM structure disrupts the normative andpolitical order of a college and thereby threatensestablished values (i.e., by undermining the auton-omy of value-bearing elites and altering the balanceof power between the college’s operative and nor-mative systems). We buttressed these theoreticalarguments about EM’s prima facie importance byreviewing higher-education research documentingits indirect and longer-term consequences—for in-stance, its negative effect on equal access to highereducation and its tendency to divert resourcesaway from core educational purposes. We thenwent on to build a theoretical and empirical modelof enrollment management adoption. This modeldrew both from Selznick’s theory (with its insightsabout the organizational factors that tend to makevalues more precarious or secure) and from thecontemporary neoinstitutional perspective (withits insights about diffusion processes and thecausal role of the broader environment).

other early/late differences revealed in Table 3. Thoughwe presented migration under the heading of leadership,we also emphasized that it is a mechanism throughwhich contextual forces make themselves felt in an or-ganization. Its declining significance over time may alsoreveal evidence of vicarious learning from earlieradopters.

FIGURE 1Percentage of Colleges in Sample with Enrollment Management Structure in 2006, by Statusa

a The t-value for comparison between middle-status group and all others � 5.83 (p � .001).


FIG

UR

E2

Rel

ativ

eM

agn

itu

de

ofS

ign

ific

ant

Eff

ects

from

Tab

le2a

aA

llva

lues

ind

icat

eef

fect

sof

aon

est

and

ard

dev

iati

onch

ange

inva

riab

les

un

less

oth

erw

ise

not

ed.

Ast

eris

ks(*

)in

dic

ate

effe

cts

ofa

one

un

itin

crea

sein

vari

able

s.

Results provided robust support for this model.Colleges were less vulnerable to EM adoption whenthey had more powerful and professionalized fac-ulty and admissions personnel, were less depen-dent on students and their tuition dollars, and hadlonger-tenured presidents. They were more suscep-tible to the adoption of the structure when theywere led by presidents who had migrated fromearlier EM adopters and when they had previouslyaltered (and perhaps “corrupted”) their internalcharacters by adopting professional degree pro-grams. Contextual factors also exerted their hypoth-esized effects on enrollment management adoption.Colleges were more likely to adopt when they facedmiddle-status pressures and direct competitionfrom earlier EM adopters. The structure’s growingprevalence in the larger field also positively af-fected adoption, though this effect abated over timeas EM consequences became better understood andmore widely publicized.

Taken together, these findings present a pictureof organizations that were simultaneously drawn toand, in some instances, repelled from enrollmentmanagement adoption. The structure proved mostalluring to those colleges facing the various opera-tive system problems that EM promises to solve,and its growing cultural prevalence appeared tomark it as an increasingly appropriate solution tothese technical problems. However, these internaland external drivers of adoption were mitigated tothe extent that a college had powerful internalvalue guardians, experienced leadership, and sig-nificant autonomy from its resource environment.The results also suggest some field-level learning,as evidenced by the declining rate of EM adoptionin later years.

Our study has a number of empirical and theo-retical implications. We begin with the former andmove toward the latter in concluding this article.Most directly, the study implies that scholars whoare interested in understanding processes of valueattenuation and change should give more attentionto the role that management innovations (and mun-dane administrative changes more generally) playtherein. EM may represent an extreme example, butthe basic organizational processes underlying itsobserved effects appear to be highly generalizable.Any administrative change that alters the internalnormative and political order of an organizationhas the ability to generate theoretically similar (ifsubstantively distinct) consequences (Hallett &Ventresca, 2006). Although this risk would appearto be most formidable for nonprofit organizationswith some social value at the very core of theirexistence, the process may be no less common infor-profit contexts. A growing number of contem-

porary businesses are founded, for instance, withthe expressed purpose of achieving various socialand environmental values in addition to makingprofits. When these organizations adopt ration-alizing management innovations in an effort to im-prove operating effectiveness, they may (quite un-intentionally) tip the internal balance of poweraway from the former set of goals. Similar processesmay also occur within the (perhaps smaller) popu-lation of large, established corporations that canmake credible claims to serving values beyondmere profit making. A notable current example is3M. That much-celebrated institution’s ability torealize its most prominent, self-defining value—innovation—appears to have been deeply undercutby its recent adoption of process management prac-tices (Hindo, 2007).20 Evidence suggests that thisprototypically rational management innovation hashad similar consequences in many other corpora-tions (Benner & Tushman, 2002).

Our research focused on understanding organ-izations’ ability to recognize and reject value-subverting administrative innovations. This is aparticularly important concern, given that suchinnovations promise to solve organizational prob-lems and often have a positive cultural valence—and do not therefore appear to an individual organ-ization as things that obviously need to be resisted.(They are, as we have noted, incipient threats). Ourexclusive focus on the adoption decision also en-abled us to engage deeply with neoinstitutionalismand to employ much of its theoretical and analyti-cal apparatus alongside Selznick’s theory. Never-theless, the latter perspective reminds us that theinitial adoption of the enrollment managementstructure is but one consequential moment in eachindividual organization’s ongoing evolution.Though such events threaten the values of an or-ganization, these events do not determine its ulti-mate fate. In light of this, future research might alsoproductively examine postadoption dynamics.Studies could, for instance, explore organizations’efforts to contain the damage caused by value-threatening innovations, or focus on explaining theultimate causes of their abandonment.

Our article also has a number of farther-reachingand perhaps ultimately more important implica-

20 It is worth noting that this foreign practice (specifi-cally, Six Sigma) was introduced to 3M by a migrantchief executive (James McNerney), who imported it fromhis previous employer (General Electric). Robert Nardelliexported the same practice from GE to Home Depot, withsimilarly negative effects on that organization’s vauntedcustomer service culture (Grow, Brady, & Arndt, 2006).


tions that go beyond the future study of value-subverting innovations. These are to be found inour rearticulation of Selznick’s institutionalismand in our effort to more faithfully reconcile histheory with the neoinstitutional perspective. Al-though our study employed many of the latter per-spective’s theoretical arguments and analyticaltools, it takes its core premises, questions, and pur-poses directly from Selznick. We have conceptual-ized organizations as recalcitrant tools that exist forthe purpose of achieving social values and haveconducted our study of enrollment management inan effort to shed further light on the conditions andprocesses that frustrate and enable this all-impor-tant organizational competency. In our efforts atintegration, we have portrayed our sample collegesfirst as autonomous entities with an inherent ca-pacity for self-determination, and only thereafter asmembers of a larger institutional field. We have,likewise, presented enrollment management pri-marily as a disruption to an individual organiza-tion’s intricate constitutional order, and only sec-ondarily as a manifestation of a larger, field-levelcultural trend. In short, although we have clearlystressed that the institutional environment matters,we have suggested that it matters primarily becauseof its effects on each individual organization and itscapacity for value realization.

This value-focused and organization-centric ap-proach is clearly not the only one that can be used inreconciling these two distinct theories, and we recog-nize that it is likely to be less appealing to scholarswho have a primary interest in institutional logicsand field-level processes of diffusion and change.Nevertheless, it does have a number of clear advan-tages to recommend it. First, and most obviously, itencourages a basic respect for values and the individ-ual organizations that exist to serve them. This re-spectful and empathetic orientation, which is ubiqui-tous in Selznick’s writings, is very difficult tomaintain if one chooses to portray organizations (orpeople) as units of larger institutional systems and toreinterpret their espoused values as arbitrary socialconstructions, covers for narrow self-interest, or mereshadows of hidden power relationships (Heclo, 2008;Selznick, 2008). Integrity, leadership, responsibility,commitment, character, mission, and most of theother humanist concepts at the very core of Selznick’sthinking dissolve into the solution of the latter per-spective(s). Second, our effort to bring Selznick intothe foreground highlights the tremendous internalcomplexity of organizations and compels institu-tional scholars to recognize the persistent and inerad-icable threats to value realization that exist withinthem. In Selznick’s view, institutional values are in-herently “precarious,” and never to be “taken–for-

granted”—even in very well established organiza-tions such as those we have studied. The realizationof values requires sustaining social structures, as wellas commitment, foresight, and ongoing “institutionalwork” (what he called leadership) (Kraatz, 2009; Law-rence & Suddaby, 2006). Third, the perspective’s pre-occupation with the local and immediate causes ofvalue subversion encourages us to look for local so-lutions, in addition to distant, field-level ones. To theextent that organizational design, unchecked statuscompetition, and leadership failure are the proximatecauses of value subversion in higher education orelsewhere, remedies are not far to seek. Field-level“master trends,” although obviously consequential,are by no means determinative or inexorable.

Finally, and perhaps most importantly, we be-lieve that our neo-Selznickian approach is advan-tageous because it enables researchers to assume aposture that is realistic, critical and empirical, butalso idealistic, affirmative, and forward-looking.No one can seriously read Selznick without acquir-ing a deep and lasting awareness of organizationalrecalcitrance and human frailty. But his theory alsoreminds readers that there are genuine values to beachieved in organizational life and mandates equalattention to the organizational structures and pro-cesses that provide these values with their best(and perhaps only) chance for “life and hope.” Wethink that this unique outlook, which Krygier(2002) aptly labeled “Hobbesian idealism,” may beSelznick’s most important legacy for contemporaryinstitutional scholarship. We hope that our studyprovides some small evidence of its continuingpromise. Much more is needed.

REFERENCES

Abrahamson, E. 1991. Managerial fads and fashions: Thediffusion and rejection of innovations. Academy ofManagement Review, 16: 586–612.

Abrahamson, E., & Rosenkopf, L. 1993. Institutional andcompetitive bandwagons: Using mathematical mod-eling as a tool to explore innovation diffusion. Acad-emy of Management Review, 18: 487–517.

Allison, P. D. 1984. Event history analysis: Regressionfor longitudinal event data. Beverly Hills, CA: Sage.

Amis, J., Slack, T., & Hinings, C. R. 2002. Values andorganizational change. Journal of Applied Behav-ioral Science, 38: 436–465.

Baum, S. 2000. Equity and enrollment management:Conflicting goals? Paper presented at the 2000 Na-tional Association of College and University Busi-ness Officers Forum on Tuition Discounting, Dallas.

Baum, S., & Lapovsky, L. 2006. Tuition discounting: Notjust a private college practice. New York: CollegeBoard.


Benner, M. J., & Tushman, M. 2002. Process managementand technological innovation: A longitudinal studyof the photography and paint industries. Adminis-trative Science Quarterly, 47: 676–706.

Birnbaum, R. 2000. Management fads in higher educa-tion: Where they come from, what they do, whythey fail. San Francisco: Jossey-Bass.

Bok, D. C. 2004. Universities in the marketplace: Thecommercialization of higher education. Princeton,NJ: Princeton University Press.

Brint, S., & Karabel, J. 1991. Institutional origins andtransformations: The case of American communitycolleges. In W. W. Powell & P. J. DiMaggio (Eds.),The new institutionalism in organizational analy-sis: 337–359. Chicago: University of Chicago Press.

Bronner, E. 1998. College efforts to lure the best set othersback. New York Times, June 21.

Burns, L. R., & Wholey, D. R. 1993. Adoption and aban-donment of matrix management programs: Effects oforganizational characteristics and interorganization-al networks. Academy of Management Journal, 36:106–138.

Carnegie Foundation for the Advancement of Teaching.1987. A classification of institutions of higher ed-ucation. Princeton, NJ: Carnegie Foundation for theAdvancement of Teaching.

Clark, B. R. 1956. Organizational adaptation and precar-ious values: A case study. American SociologicalReview, 21: 327–336.

Clark, B. R. 1960. The open door university: A casestudy. New York: McGraw-Hill.

Clark, B. R. 1970. The distinctive college: Reed, Anti-och, and Swarthmore. Chicago: Aldine.

D’Aunno, T., Succi, M., & Alexander, J. A. 2000. The roleof institutional and market forces in divergent organ-izational change. Administrative Science Quar-terly, 45: 679–703.

D’Aunno, T., Sutton, R. I., & Price, R. H. 1991. Isomor-phism and external support in conflicting institu-tional environments: A study of drug abuse treat-ment units. Academy of Management Journal, 34:636–661.

Davis, J. S. 2003. Unintended consequences of tuitiondiscounting. Indianapolis: Lumina Foundation forEducation.

DiMaggio, P. J., & Powell, W. 1983. The iron cage revis-ited: Institutional isomorphism and collective ration-ality in organizational fields. American Sociologi-cal Review, 48: 147–160.

Dolence, M. G. 1998. Strategic enrollment management.In C. C. Swann & S. E. Henderson (Eds.), Handbookfor the college admission profession: 71–91. West-port, CT: Greenwood.

Ehrenberg, R. G., Zhang, L., & Levin, J. M. 2006. Craftinga class: The trade-off between merit scholarships and

enrolling lower-income students. Review of HigherEducation, 29(2): 195–211.

Enz, C. A. 1988. The role of value congruity in intraor-ganizational power. Administrative Science Quar-terly, 33: 284–304.

Fiol, C. M., & O’Connor, E. J. 2003. Waking up! Mindful-ness in the face of bandwagons. Academy of Man-agement Review, 28: 54–70.

Gaudiani, C. L. 2000. The hidden costs of merit aid.Change, 32(4): 19.

Geiger, R. L. 2004. Knowledge and money: Researchuniversities and the paradox of the marketplace.Palo Alto, CA: Stanford University Press.

Goral, T. 2003. Is discounting dangerous? Caught in an“arms race,” higher education needs to find a wayout of the tuition discounting spiral. UniversityBusiness, 6(8): 22–26.

Gose, B. 1999. Colleges turn to consultants to shape thefreshman class. Chronicle of Higher Education,45(35): A49–A51.

Green, T. C. 2005. Financial aid, access, and America’ssocial contract with higher education. College andUniversity, 80(3): 9–13.

Greenwood, R., & Hinings, C. R. 1996. Understandingradical organizational change: Bringing together theold and the new institutionalism. Academy of Man-agement Review, 21: 1022–1054.

Greenwood, R., & Suddaby, R. 2002. The role of profes-sional associations in the transformation of institu-tionalized fields. Academy of Management Jour-nal, 45: 58–80.

Grow, B., Brady, D., & Arndt, M. 2006. Renovating HomeDepot: Skip the touchy-feely stuff. The big-box storeis thriving under CEO Bob Nardelli’s military-stylerule. BusinessWeek, March 6: 50–58.

Gusfield, J. R. 1955. Social structure and moral reform: Astudy of the Women’s Christian Temperance Union.American Journal of Sociology, 61: 221–232.

Hallett, T. P., & Ventresca, M. J. 2006. Inhabited institu-tions: Interaction and organization in Gouldner’sPatterns of Industrial Bureaucracy. Theory & Soci-ety, 35: 213–236.

Heclo, H. 2002. The statesman: Revisiting Leadership inAdministration. In R. A. Kagan, M. Krygier, & K.Winston (Eds.), Legality and community: On theintellectual legacy of Philip Selznick: 295–310.Lanham, MD: Rowman & Littlefield.

Heclo, H. 2008. On thinking institutionally. Boulder,CO: Paradigm.

Heimer, C. A. 1999. Competing institutions: Law, medi-cine, and family in neonatal intensive care. Law andSociety Review, 33: 17–66.

Heller, D. E. 2004. The changing nature of financial aid.Academe, 90(4): 36–38.


Heller, D. E., & Nelson Laird, T. F. 1999. Institutionalneed-based and non-need grants: Trends and differ-ences among college and university sectors. Journalof Student Financial Aid, 29(3): 7–24.

Henderson, S. E. 2001. On the brink of a profession. In J.Black (Ed.), The strategic enrollment managementrevolution: 3–36. Washington, DC: American Asso-ciation of Registrars and Admissions Officers.

Hickson, D. J., Hinings, C. R., Lee, C. A., Schneck, R. E.,& Pennings, J. M. 1971. A strategic contingencies’theory of intraorganizational power. AdministrativeScience Quarterly, 16: 216–229.

Hindo, B. 2007. At 3M, a struggle between efficiencyand creativity: How CEO George Buckley is man-aging the yin and yang of discipline and imagina-tion. BusinessWeek, June 11(special section):8 –14.

Hinings, C. R., Thibault, L., Stack, T., & Kikulus, L. 1996.Values and organization structure. Human Rela-tions, 49: 885–916.

Hirsch, P. M., & Lounsbury, M. 1997. Ending the familyquarrel: Toward a reconciliation of “old” and “new”institutionalism. American Behavioral Scientist,40: 406–418.

Hossler, D. R. 1984. Enrollment management: An inte-grated approach. New York: College Entrance Ex-amination Board.

Hossler, D. R. 2004. How enrollment management hastransformed—or ruined—higher education. Chron-icle of Higher Education, 50(34): B3–B5.

Ingersoll, R. J. 1988. The enrollment problem: Provenmanagement techniques. New York: Macmillan.

Jensen, M. 2003. The role of network resources in marketentry: Commercial banks’ entry into investmentbanking, 1991–1997. Administrative Science Quar-terly, 48: 466–497.

Jensen, M. 2006. Should we stay or should we go?Accountability, status anxiety, and client defec-tions. Administrative Science Quarterly, 51: 97–128.

Kemerer, F. R., Baldridge, J. V., & Green, K. C. 1982.Strategies for effective enrollment management.Washington, DC: American Association of State Col-leges and Universities.

Kirp, D. L. 2003. Shakespeare, Einstein, and the bottomline: The marketing of higher education. Cam-bridge, MA: Harvard University Press.

Kirp, D. L., & Holman, J. T. 2002. This little student wentto market. American Prospect, 13(18): 29–31.

Knudsen, C. 1995. The competence view of the firm:What can modern economists learn from Philip Sel-znick’s sociological theory of leadership? In W. RScott & S. Christensen (Eds.), The institutional con-struction of organizations: 135–163. ThousandOaks, CA: Sage.

Kraatz, M. S. 2009. Leadership as institutional work: Abridge to the other side. In T. B. Lawrence, R. Sud-daby, & B. Leca (Eds.), Institutional work: Actorsand agency in institutional studies of organiza-tions: 59–91. Cambridge, U.K.: Cambridge Univer-sity Press.

Kraatz, M. S., & Block, E. S. 2008. Organizational impli-cations of institutional pluralism. In R. Greenwood,C. Oliver, R. Suddaby, & K. Sahlin-Andersson (Eds.),Handbook of organizational institutionalism: 243–275. London: Sage.

Kraatz, M. S., & Moore, J. D. 2002. Executive migrationand institutional change. Academy of ManagementJournal, 45: 120–143.

Kraatz, M. S., & Zajac, E. J. 1996. Exploring the limits ofthe new institutionalism: The causes and conse-quences of illegitimate organizational change. Amer-ican Sociological Review, 61: 812–836.

Krygier, M. 2002. Philip Selznick, normative theory, andthe rule of law. In R. A. Kagan, M. Krygier, & K.Winston (Eds.), Legality and community: On theintellectual legacy of Philip Selznick: 19–48. Lan-ham, MD: Rowman & Littlefield.

Krygier, M. 2008. Foreword to P. Selznick, A humanistscience: Values and ideals in social inquiry: vii–xv. Stanford, CA: Stanford University Press.

Lacey, M. 2002. Taking ideals seriously: Philip Sel-znick and the natural-law tradition. In R. A. Kagan,M. Krygier, & K. Winston (Eds.), Legality andcommunity: On the intellectual legacy of PhilipSelznick: 67– 81. Lanham, MD: Rowman & Little-field.

Lapovsky, L., & Hubbell, L. L. 2000. Positioning for com-petition: Results from NACUBO’s Annual TuitionDiscounting Survey. In Proceedings from theNACUBO forum on tuition discounting: 22–30.Washington, DC: National Association of Collegeand University Business Officers.

Lawrence, T. B., & Suddaby, R. 2006. Institutions andinstitutional work. In S. R. Clegg, C. Hardy, T. B.Lawrence, & W. R. Nord (Eds.), Handbook of or-ganization studies (2nd ed.): 215–254. London:Sage.

Lieberman, M. B., & Asaba, S. 2006. Why do firms imitateeach other? Academy of Management Review, 31:366–385.

Lounsbury, M. 2002. Institutional transformation andstatus mobility: The professionalization of the fieldof finance. Academy of Management Journal, 45:255–266.

Lounsbury, M. 2007. A tale of two cities: Competinglogics and practice variation in the professionalizingof mutual funds. Academy of Management Journal,50: 289–307.

March, J. G., & Simon, H. A. 1958. Organizations. NewYork: Wiley.


McPherson, M. S., & Schapiro, M. O. 1998. The studentaid game: Meeting need and rewarding talent inAmerican higher education. Princeton, NJ: Prince-ton University Press.

Messinger, S. L. 1955. Organizational transformation: Acase study of a declining social movement. Ameri-can Sociological Review, 30: 3–10.

Miller, D. 1991. Stale in the saddle: CEO tenure and thematch between organization and environment. Man-agement Science, 37: 34–52.

Miner, A. S., & Haunschild. P. R. 1995. Population levellearning. In L. L. Cummings & B. M. Staw (Eds.),Research in organizational behavior, vol. 17: 115–166. Greenwich, CT: JAI Press.

Mitchell, R. K., Agle, B. R., & Wood, D. J. 1997. Towarda theory of stakeholder identification and salience:Defining the principle of who and what reallycounts. Academy of Management Journal, 22:853– 886.

Noel-Levitz. 2002. National Enrollment ManagementSurvey findings for fall 2001. Littleton, CO: Noel-Levitz.

Oliver, C. 1991. Strategic responses to institutional pro-cesses. Academy of Management Review, 16:145–179.

Oliver, C. 1992. The antecedents of deinstitutionaliza-tion. Organization Studies, 13: 563–588.

Perrow, C. 1986. Complex organizations: A critical es-say (3rd ed.). New York: Random House.

Peterson’s Guides. 1987. Guide to four-year colleges1988. Princeton, NJ: Peterson’s Guides.

Pfeffer, J., & Salancik, G. R. 1978. The external control oforganizations. New York: Harper.

Phillips, D. J., & Zuckerman, E. W. 2001. Middle-statusconformity: Theoretical restatement and empiricaldemonstration in two markets. American Journal ofSociology, 107: 379–429.

Podolny, J. M. 1993. A status-based model of marketcompetition. American Journal of Sociology, 98:829–872.

Podolny, J. M. 2005. Status signals: A sociological studyof market competition. Princeton, NJ: PrincetonUniversity Press.

Porac, J. F., Thomas, H., Wilson, F., Paton, D., & Kanfer,A. 1995. Rivalry and the industry model of Scottishknitwear producers. Administrative Science Quar-terly, 40: 203–227.

Quirk, M. 2005. The best class money can buy. AtlanticMonthly, 296(November): 128–134.

Redd, K. E. 2000. Discounting toward disaster: Tuitiondiscounting, college finances, and enrollment oflow-income undergraduates. Indianapolis: USAGroup Foundation.

Russo, J. A., & Coomes, M. D. 2000. Enrollment manage-

ment, institutional resources, and the private col-leges. New Directions for Student Services, 89:33–46.

Scott, R. W. 2008. Institutions and organizations (3rded.). Los Angeles: Sage.

Scott, W. R., Ruef, M., Mendel, P., & Caronna, C. 2000.Institutional change and healthcare organiza-tions: From professional dominance to managedcare. Chicago: University of Chicago Press.

Selznick, P. 1943. An approach to a theory of bureau-cracy. American Sociological Review, 8: 47–54.

Selznick, P. 1949. TVA and the grass roots. Berkeley:University of California Press.

Selznick, P. 1957. Leadership in administration. NewYork: Harper & Row.

Selznick, P. 1992. The moral commonwealth: Socialtheory and the promise of community. Berkeley:University of California Press.

Selznick, P. 1996. Institutionalism “old” and “new.” Ad-ministrative Science Quarterly, 41: 270–277.

Selznick, P. 2000. On sustaining research agendas: Theirmoral and scientific basis. Journal of ManagementInquiry, 9: 277–282.

Selznick, P. 2008. A humanist science: Values and ide-als in social inquiry. Stanford, CA: Stanford Univer-sity Press.

Smith, K. G., Grimm, C., Gannon, M., & Chen, M. 1991.Organizational information processing, competitiveresponses, and performance in the U.S. domesticairline industry. Academy of Management Journal,34: 60–85.

Stinchcombe, A. L. 1997. On the virtues of the old insti-tutionalism. In J. Hagan & K. Cook (Eds.), Annualreview of sociology, vol. 23: 1–18. Palo Alto, CA:Annual Reviews.

Strang, D., & Meyer, J. W. 1993. Institutional conditionsfor diffusion. Theory and Society, 22: 487–511.

Suchman, M. C. 1995. Managing legitimacy: Strategicand institutional approaches. Academy of Manage-ment Review, 20: 571–610.

Thomas, S. D., Tan, A. G., & Hoffman, C. M. 2006.Digest of education statistics 2005, NCES 2006 –030. Washington, DC: U.S. Government PrintingOffice.

Thompson, J. D. 1967. Organizations in action. NewYork: McGraw-Hill.

Thornton, P. H., & Ocasio, W. 1999. Institutional logicsand the historical contingency of power in organiza-tions: Executive succession in the higher educationpublishing industry, 1958–1990. American Journalof Sociology, 105: 801–844.

Tolbert, P. S., & Zucker, L. G. 1983. Institutionalsources of change in the formal structure of organ-izations: The diffusion of civil service reform,


1880 –1935. Administrative Science Quarterly,28: 22–39.

Washburn, J. 2005. University, Inc.: The corporate cor-ruption of higher education. New York: BasicBooks.

Washington, M., & Zajac, E. J. 2005. Status evolution andcompetition: Theory and evidence. Academy ofManagement Journal, 48: 282–296

Washington, M., & Ventresca, M. J. 2004. How organi-zations change: Three mechanisms supporting theincorporation of emerging strategies in U.S. highereducation. Organization Science, 15(1):82–97.

Wellman, J. V., Desrochers, D. M., & Lenihan, C. M.2008. The growing imbalance: Recent trends inU.S. postsecondary education finance. Indianap-olis: Lumina Foundation for Education.

Westphal, J. D., Gulati, R., & Shortell, S. M. 1997.Customization or conformity? An institutional andnetwork perspective on the content and conse-quences of TQM adoption. Administrative Sci-ence Quarterly, 42: 366 –394.

Wiersema, M. F., & Bantel, K. A. 1992. Top manage-ment team demography and corporate strategicchange. Academy of Management Journal, 35:91–121.

Yamaguchi, K. 1991. Event history analysis. NewburyPark, CA: Sage.

Zald, M. N., & Denton, P. 1963. From Evangelism togeneral service: The transformation of the YMCA.Administrative Science Quarterly, 8: 214–234.

Zernike, K. 2009. To keep students, colleges cut anythingbut aid. New York Times, February 28.

Zilber, T. B. 2002. Institutionalization as an interplaybetween actions, meanings and actors: The case of arape crisis center in Israel. Academy of Manage-ment Journal, 45: 234–254.

Matthew S. Kraatz ([email protected]) is an associateprofessor in the Department of Business Administrationat the University of Illinois. His scholary interests in-clude organizational adaptation, learning, governance,identity, reputation, leadership, and other institutionalprocesses. He received his Ph.D. from the Kellogg Grad-uate School of Management at Northwestern University.

Marc J. Ventresca ([email protected]) is onthe faculty at the University of Oxford, where he is alecturer in strategy at the Saïd Business School, a Fellowof Wolfson College, and Faculty Fellow at the Institutefor Science, Innovation and Society. He is also researchassociate professor of Global Public Policy at the NavalPostgraduate School and a senior scholar at the Center forInnovation and Communication at Stanford University.His research focus is the economic sociology of strategyand innovation. Current projects investigate governanceinnovations in knowledge-intensive organizations andindustries, including global financial markets, online in-formation services, the “ancient” universities, ecosys-tems services markets in Peru and Ghana, and multisec-tor complex operations. He received his Ph.D. inorganizational sociology at Stanford University.

Lina Deng ([email protected]) is a Ph.D. student inorganizational behavior and theory in the Department ofBusiness Administration at the University of Illinois. Herresearch interests include organizational status, institu-tional change, and organizations’ strategic responses tovarious social pressures. Her dissertation research ex-amines how status concerns affect presidential hiring inAmerican higher education. She received her master’sdegree in management from China Petroleum Universityin Beijing.


Copyright of Academy of Management Journal is the property of Academy of Management and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

Documents

PRECARIOUS VALUES AND MUNDANE …...structure in a sample of 515 private liberal arts colleges over the period 1987 through 2006. THEORY: PRECARIOUS VALUES AND MUNDANE INNOVATIONS