Professional Documents
Culture Documents
Ethical Behavior and Bureaucratic Structure in Marketing Research Organizations
Ethical Behavior and Bureaucratic Structure in Marketing Research Organizations
SKINNER*
The authors examine the relationship between bureaucratic structure and ethical
behavior in research firms, data subcontractor organizations, and corporate re-
search departments. The results suggest that the bureaucratic structure of the or-
ganization is related to ethical behavior; the nature of this relationship varies across
the three types of research organizations investigated.
•
Ethical Behavior and Bureaucratic Structure In
Marketing Research Organizations
The most extensive concern about ethical problems that bureaucratic structures in research organizations have
within the field of marketing has been in the area of mar- an impact on decision making.
keting research (Murphy and Laczniak 1981). As the The purpose of our study is to determine whether per-
utilization of marketing research continues to increase, ceived ethical behavior is influenced by the bureaucratic
opportunities for the misuse of marketing research also structures and policies (codes of ethics) of the diverse
increase. Deception, invasion of privacy, dissemination organizations that participate in the marketing research
of faulty conclusions, and disguising sales efforts as process. More specifically, we investigate the relation-
marketing research are some common examples of ship between three facets of bureaucracy-formaliza-
unethical actions associated with marketing research tion, centralization, and controls-and perceived ethical
practices. It is not surprising that Gallup (1987, p. 10) behavior. This relationship is examined in the three types
reported "forty-seven percent of the executives [sur- of organizations involved in marketing research activi-
veyed] said it is either very or fairly likely that some ties-data subcontractor organizations, research firms
form of regulation [of marketing research] will be im- (marketing research agencies), and corporate research
posed. " departments.
Hunt, Chonko, and Wilcox (1984) found that actions
of top managers are the single best predictor of ethical CONCEPTUAL FRAMEWORK
problems for marketing researchers. Almost half of the Ethics has been defined as " . . . inquiry into the na-
marketing researchers in their sample believe their man- ture and grounds of morality," where the term "moral-
agers have an opportunity to engage in unethical behav- ity" is taken to mean "moral judgments, standards, and
ior and 17% believe managers in their companies fre- rules of conduct" (Taylor 1975, p. 1). Ethics in the con-
quently engage in unethical behavior. They also report text of marketing research warrants special analysis be-
that research integrity, fair treatment of outside clients, cause the participants (data subcontractors, research firms,
and research confidentiality are the three most important and corporate research departments) have different ob-
ethical concerns ofresearchers. Additionally, several re- jectives.
search studies that do not specifically address ethical French and Ebner (1986) suggest that data subcon-
problems (Deshpande and Zaltman 1982, 1984) indicate tractors and their clients have different objectives that
create conflict and each use different philosophical
frameworks for ethical decision making. Corporate re-
search departments are often results-oriented, with their
*0. C. Ferrell is Associate Professor, Department of Marketing, evaluation depending on the success of their recommen-
Texas A&M University. Steven J. Skinner is Associate Professor, De- dations. Many decision makers in corporate research de-
partment of Marketing, University of Kentucky. partments may be consequentialists or users of teleolog-
The authors are grateful for suggestions made by Paul Busch, James ical moral philosophies (French and Ebner 1986).
H. Donnelly, Jr., Joseph P. Guiltinan, A. Parasuraman, the editor,
and anonymous JMR reviewers. Teleological philosophies pertain to the moral worth of
a behavior determined totally by the consequences of the
103
behavior. Utilitarianism, a teleological philosophy, states of bureaucratic structure to predict ethical behavior in
that if a particular technique works for the greatest bal- marketing research organizations. The models include
ance of value for all persons, it could be considered eth- the dependent variable ethical behavior and predictor
ical. variables representing bureaucratic structure.
In contrast, data subcontractors often are evaluated on Bureaucracy was selected because it is related directly
"how well they have abided by a set of standards and to the development and enforcement of codes of ethics.
rules" (French and Ebner 1986, p. 50). Abiding by stan- A bureaucracy is a "continuous organization of official
dards or rules such as the Marketing Research Associ- functions bound by rules" (Weber 1947, p. 330). Rules,
ation (MRA) Code of Ethics implies moral obligations standards, and systematic procedures enable organiza-
or commitments. Philosophies that focus on principles tional activities to be oriented toward objectives. Also,
of justice, basic rights, or rules designed for necessary each lower department or function is assumed to be un-
and proper conduct are deontological. der the control and supervision of a higher one. "Ad-
Research firms are in a double bind because they fre- ministrative acts, decisions, and rules are formulated and
quently are the clients of data subcontractors and the recorded in writing" (Weber 1947, p. 332).
research suppliers to corporate research departments. Beginning with Weber's studies at the end ofthe nine-
Research firms may be evaluated on the basis of teenth century, numerous researchers have examined or-
deontological philosophies when providing objective data ganizational structure and have disaggregated bureau-
for corporate research departments, but may be evalu- cracy into its component parts, using them as variables
ated teleologically when providing recommendations that in analyzing organizational structure (Hall 1963). These
are used by management. Because different marketing variables include formalization of operating procedures,
research organizations may use both types of moral phi- centralization of authority, and controls (Aldrich 1979).
losophy, ethical behavior may vary with their role or re-
lationship in the marketing research process. Moreover, Formalization
it is very likely that individuals in the same organization Formalization is the standardization and recording of
use different moral philosophies reflecting their personal ". . . statements of procedures, rules, roles, . . . an
moral development. operation of procedures" (Pugh et al. 1963, p. 303-4).
Ethical behavior of individuals also may differ across More formalized marketing research organizations should
organizations. Deshpande and Zaltman (1982, 1984) have have more control over the ethical behavior of their em-
examined the marketing manager/marketing researcher ployees. Codes of ethics in organizations are part of the
interface. Their findings indicate that the unique social written rules and regulations associated with the for-
environment surrounding marketing research is a two- malization of a bureaucracy. Several studies (e.g., Fer-
communities metaphor. In a two-communities metaphor, rell and Weaver 1978; Fritzsche and Becker 1983; Hunt,
decision makers operate in different worlds with differ- Chonko, and Wilcox 1984) link corporate policy to im-
ent reward systems, different languages, and different, proved ethical conduct. Weaver and Ferrell (1977) found
often conflicting, values. Mutual mistrust is commonly that enforcement of company policy was necessary to
a part of the two-communities metaphor (Deshpande and establish desired ethical conduct of marketing managers.
Zaltman 1984). Corporate policy related to ethical behavior is often in
Clearly the two-communities metaphor sets the stage the form of codes of ethical conduct.
for ethical conflict, but this difference is not necessarily
dysfunctional. Though Deshpande and Zaltman (1984) Centralization
did not address ethical issues specifically, their findings Centralization refers to the hierarchical level that has
suggest the potential for ethical conflicts across different authority to make a decision. If decisions are delegated
types of marketing research organizations. Possibly all to lower levels the organization is decentralized and if
parties to the marketing research process think they are decision-making authority is kept at the top level it is
ethical but use different moral philosophies to justify de- centralized (Pugh 1973). Centralized power structures
cisions. Krugman and Ferrell (1981) support the two- are vertical in direction with power in the hands of a
communities metaphor in advertising ethics; they found small group at the top of the organization. Decentralized
that corporate advertising managers believe their agency power structures result in "decisions on various orga-
counterparts have lower ethical standards. nizational issues distributed laterally and vertically
throughout the organization" (Zey-FerreIl1979, p. 156).
MODEL DEVELOPMENT Highly centralized organizations should have more op-
No empirical studies have been conducted to deter- portunity to control ethical decisions.
mine whether bureaucratic structure is related to ethical Authority is located in the formal hierarchy and re-
behavior. Because marketing research organizations dif- porting relationships of the bureaucracy. Two properties
fer significantly in their task environments, any rela- that identify authority are (I) position in the organiza-
tionship between bureaucratic structure and ethical be- tion, rather than personal characteristics, and (2) the
havior would be expected to vary across organizations. compliance of subordinates because they perceive that
Therefore we developed models incorporating concepts superiors have a legitimate right to exercise authority
ETHICAL BEHAVIOR AND BUREAUCRATIC STRUCTURE 105
(Grimes 1978). Subordinates obey authority because it was consistent with procedures suggested by Churchill
is something they respect and they often go along whether (1979). A pool of 70 items for the ethical behavior scale
they agree with a superior or not. The acceptance of au- was generated via prestudy interviews with marketing
thority should improve ethical behavior within the or- researchers representing each of the three organization
ganization. types. Hence these items were developed specifically for
our study. Items were judged for face validity by a panel
Controls of 11 marketing researchers to detect any problems in
Organizational control is a cycle that includes the three wording or construction. An item was eliminated if a
stages of target setting, measuring or monitoring, and single judge felt it lacked face validity.
feedback (Ouchi 1977). Obviously, one strategy for con- The scales were assessed by confirmatory factor anal-
trol could be market control whereby exchange relations ysis (Joreskog and Sorbom 1981). Constructs were pu-
and ethical behavior are monitored. Controls in the bu- rified by deleting items that did not load in excess of .30
reaucracy can consist of rules, standards, and internal (Green 1978). The scale items retained for analyses are
procedures that could influence ethical behavior. The or- listed in the Appendix. All scale items were significant
ganizational culture also can exert control by influencing indicators (t > 2.326) of their respective constructs. The
shared values and philosophical orientations toward eth- reliability for the constructs ranges from .69 to .82 and
ical behavior. is acceptable for research purposes (Nunnally 1978).
Respondents were asked whether their organization had
METHOD a code of ethical conduct; responses were yes, no, or
don't know. The fact that all respondents answered either
Sample yes or no suggeststhat they understood the question. Those
Data were collected by a self-administered question- responding yes were asked whether they perceived that
naire mailed to a sample of 1500 researchers. The frame the code was enforced; again, all respondents answered
used was the membership list of the Marketing Research either yes or no.
Association. The corporate classification was supple-
mented by the American Marketing Association (AMA) RESULTS
membership list because the Marketing Research Asso- The models were examined by multiple regression
ciation has less than 500 members in that category. All analysis. The criterion variable was ethical behavior; the
respondents were chosen by systematic sample selection. predictor variables were formalization, centralization,
The sample was inspected to ensure no duplication. The controls, acceptance of authority, existence of an ethical
final mailing included the questionnaire, a cover letter, code, and enforcement of an ethical code. One addi-
and a stamped, pre-addressed return envelope. From a tional predictor variable, sex of respondent, was in-
single mailing, 550 usable questionnaires were returned. cluded in the models because of the differences in per-
Additionally, 52 questionnaires were returned by the post centages of women and men across the organizations.
office, resulting in a 37.9% response rate. This response Additionally, several studies indicate that women have
compares favorably with that of previous surveys on this more ethical concerns than men (Beltramini, Peterson,
topic (cf. Hunt, Chonko, and Wilcox 1984). and Kozmetsky 1984; Ricklets 1983).
Thirty percent of the researchers responding were data The results are summarized in Table 1. The existence
subcontractors. As expected, more women (86%) than of an ethical code explains 28.1 % of the variance in the
men (14%) represented this type of organization. About ethical behavior of data subcontractors. Additionally,
45% of the respondents were employed by research firms. formalization explains 8.9% of the variance, respon-
Respondents in this type of organization were more bal- dents' sex 8%, and the enforcement of the ethical code
anced in terms of women (54%) and men (46%). Fi- 1.9%. Controls, acceptance of authority, and centraliza-
nally, 25% of the repsondents worked in corporate re- tion explain very little variance in this particular model.
search departments; 53% were women and 47% were All variables combine to explain 47.2% of the variance
men. in ethical behavior in the data subcontractor model.
The results of the research finn model also are re-
Measures ported in Table 1. Formalization explains the most vari-
Formalization, centralization, and controls were mea- ance (11.3%) in ethical behavior. Existence of an ethical
sured by scales adapted by John (1984). These scales code explains 6.1%, centralization 3.8%, enforcement
indicate the extent to which the operating procedures of of the ethical code 2.3%, and respondents' sex 3.1 %.
a finn are structured (formalization), responsibility is Controls and acceptance of authority have little predic-
delegated (centralization), and rules are enforced (con- tive power in this model. The seven predictors combine
trols). Acceptance of authority was measured by a scale to explain 26.8% of the variance in the ethical behavior
developed by Withey (1965) that indicates the extent to of researchers in marketing research agencies.
which an individual tends to obey orders and respect au- For corporate research departments, the existence of
thority. an ethical code explains 6% of the variance in ethical
Development of a scale to measure ethical behavior behavior. Formalization and centralization explain 3.2%
106 JOURNAL OF MARKETING RESEARCH, FEBRUARY 1988
Table 1
STEPWISE REGRESSION RESULTS
Type of R 2 for
organization Independent variables the model
Has an Enforces Acceptance
ethical Formal- ethical of Central-
code' ization Sex· code' Controls authority ization
Data subcontractors .472
Standardized beta - .361 d .230 d -.280d -.171' .041 -.037 .003
Incremental R 2 .281 .089 .080 .019 .001 .001 .001
and 2.6% of the variance, respectively. Additional vari- because of the differences in the distribution of this vari-
ance is explained by respondents' sex (2%) and accep- able across the three types of research organizations. The
tance of authority (1.7%); enforcement of the ethical code results suggest that respondents' sex explains the most
and controls do not explain much additional variance. variance (8%) in the data subcontractor model; nearly
About 16% of the variance in ethical behavior is ex- 86% of the data subcontractors responding were women.
plained in this model. In comparison with men, women report significantly
Comparison of the three models shows several differ- higher levels of ethical behavior in this model. Sex of
ences. The most variance is explained for data subcon- the respondent also is related significantly to ethical be-
tractors, 47.2%. For the research firm model, 26.8% of havior for research firms. Again, women report higher
the variance is explained. The least variance, 16.1%, is levels of ethical behavior. The repondents' sex is not
explained in the corporate research department model. related significantly to ethical behavior in the model for
Hence the measures employed in our study appear to be corporate researchers.
most useful in predicting perceived ethical behavior of
individuals in data subcontractor organizations. DISCUSSION AND IMPLICATIONS
The existence of an ethical code and higher levels of A significant finding of our study is that bureaucratic
formalization are related significantly to greater per- structure explains different amounts of variance in eth-
ceived ethical behavior in all three types of organiza- ical behavior across the three major types of marketing
tions. Centralization is related to higher perceived ethi- research organizations. Additionally, respondents' sex is
cal behavior in research firms, but is not related a significant predictor of ethical behavior in two of the
significantly to ethical behavior in the other two models. research organizations, data subcontractors and research
Enforcement of ethical codes is related to higher ethical firms. The predictors we used explain the most variance
behavior of data subcontractors and research firms, but in the data subcontractor model. Bureaucratic structure
is not related significantly to ethical behavior for cor- and respondents' sex explain the least variance in the
porate researchers. Acceptance of authority and controls corporate research department model; respondents' sex
are not related significantly to perceived ethical behavior is not related significantly to ethical behavior of re-
in any of the models. searchers in corporate research departments. The amount
The sex of the respondent was included in the model of variance explained for research firms is between these
ETHICAL BEHAVIOR AND BUREAUCRATIC STRUCTURE 107
two extremes, with respondents' sex significantly related vised their codes. The efforts of these organizations should
to ethical behavior in this model. For all three organi- help members develop codes that can be implemented at
zations, the more respondents perceived bureaucratic the organizational level. In addition, the AMA and MRA
structuring in their organizations, the higher the level of should examine the success of the National Advertising
reported ethical behavior. Review Board (NARB), as well as other trade groups,
Our findings are similar to the two-communities met- in developing and implementing meaningful administra-
aphor phenomenon that Deshpande and Zaltman (1984) tive procedures to enforce ethical behavior.
identified in marketing manager/marketing researcher Top managers have the opportunity to structure the
interactions. The three types of research organizations organization's bureaucracy and to enforce ethical poli-
we investigated probably operate in different organiza- cies within the company. If organizations implement
tional environments, each with different reward systems ethical codes and develop more understanding about the
and conflicting ethical philosophies. Possibly marketing impact of bureaucratic structure on ethical behavior, the
researchers in corporate research departments are re- environment for ethical practice in marketing research
sults-oriented, using teleological moral philosophies with should improve.
more freedom to make decisions; therefore, the bureau-
cratic model does not explain ethical decision making as APPENDIX
well for corporate research departments as it does for the ASSESSMENT OF SCALES
other two types of organizations. Research firms have
some of the characteristics of both corporate research de- Standardized t- Scale
partments and data subcontractors because of their unique Scale/items' loading value reliability
relationship with both of these types of organizations. Formalization .75
Hence it is not surprising that the proportion of variance If a written rule does not
explained for marketing research firms is between the cover some situation, we
proportions found for the other two types of research or- make up informal rules
for doing things as we go
ganizations. Data subcontractors abide by a set of rules along" .56 10.95
or standards and follow specific instructions in perform- There are many things in
ing assigned tasks; therefore, they probably use deon- my business that are not
tological moral philosophies in their work environment. covered by some formal
procedure for doing it'' .58 11.59
The findings indicate that women report more ethical Usually, my contact with
behavior than men for two types of organizations in our my company and its rep-
study. A Gallup poll found that women consistently re- resentatives involves
port having higher levels of ethical behavior than men doing things "by the rule
(Ricklets 1983). An interpretation in one study that found book" .74 15.60
Contact with my company
women students more ethically concerned than men was and its representatives are
that "the increasing participation of females in the work- on a formal preplanned
force will have a significant impact on what we consider basis .61 12.38
ethical business practices" (Beltramini, Peterson, and I ignore the rules and reach
informal agreements to
Kozmetsky 1984). As the ratio of women in marketing handle some situations" .51 9.98
research organizations increases, ethical behavior should When rules and procedures
improve. exist in my company,
The relationships found between formalization and they are usually written
ethical behavior in all three types of organizations pro- agreements .44 8.42
vide some insights to improving marketing research eth- Centralization .82
ics. Marketing research organizations can encourage eth- Any major decision that I
ical behavior through formal statements of policies, make has to have this
rewards, and punishments. In the absence of formalized company's approval .45 9.06
In my dealings with this
standards and rules, the acceptability of various activi- company, even quite
ties and procedures across the marketing research in- small matters have to be
dustry is ambiguous. All parties may indeed believe they referred to someone
are making ethical decisions, but use different ethical higher up for a final an-
frameworks for decision making. This situation is con- swer .78 18.15
My dealings with this com-
ducive to mistrust and ethical conflict between organi- pany are subject to a lot
zations. of rules and procedures
The findings suggest that the existence and enforce- stating how various as-
ment of codes of ethics are associated with higher levels pects of my job are to be
done .60 12.64
of ethical behavior. Possibly codes of ethics, the most I have to ask company reps
obvious device to improve ethical behavior, have not been before I do almost any-
fully utilized. The AMA and the MRA recently have re- thing in my business .90 22.52
108 JOURNAL OF MARKETING RESEARCH, FEBRUARY 1988
APPENDIX-(Continued) APPENDIX-(Continued)
Krugman, Dean M. and O. C. Ferrell (1981), "The Organi- Ethical Behavior is Declining in U.S.," Wall Street Journal
zational Ethics of Advertising," Journal of Advertising, 10 (October 31), 33.
(January), 21-30, 48. Taylor, Paul W. (1975), Principles ofEthics: An Introduction.
Murphy, Patrick E. and Gene R. Laczniak (1981), "Marketing Encino, CA: Dickensen Publishing Company, Inc.
Ethics: A Review with Implications," in Review of Mar- Weaver, K. Mark and O. C. Ferrell (1977), "The Impact of
keting, Ben M. Enis and Kenneth J. Roering, eds. Chicago: Corporate Policy on Reported Ethical Beliefs and Behavior
American Marketing Association, 251-66. of Marketing Practitioners," in Contemporary Marketing
Nunnally, Jum C. (1978), Psychometric Theory, 2nd ed. New Thought, Barnett Greenberg and Danny N. Bellenger, eds.
York: McGraw-Hill Book Company. Chicago: American Marketing Association, 477-80.
Ouchi, William G. (1977), "The Relationship Between Or- Weber, Max (1947), The Theory of Social and Economic Or-
ganizational Structure and Organizational Control," Admin- ganizations, translated by M. A. Henderson and T. Parsons.
istrative Science Quarterly, 22 (March), 95-113. New York: The Free Press.
Pugh, D. S. (1973), "The Measurement of Organizational Withey, Stephen (1965), "The US and the USSR: A Report
Structure," Organizational Dynamics, 1 (Spring), 19-34. of the Public's Perspective on the United States-Russian .
- - - , D. H. Hickson, C. R. Hinings, K. N. Macdonald, Relations in Late 1961," in Components of Defense Policy,
C. Turner, and T. Lupton (1963), "A Conceptual Scheme D. Bobrow, ed. Chicago: Rand McNally, 164-74.
for Organizational Analysis," Administrative Science Quar- Zey-Ferrell, Mary (1979), Dimensions of Organizations. Santa
terly, 8 (December), 289-316. Monica, CA: Goodyear Publishing Company.
Ricklets, Roger (1983), "Executives and General Public Say
Exciting-
(')
~
Informative - S·
~
CQ
•
g'
en
S·
~
en
en