


































UAFS ABR Journal Vol 4 No 1 2013.pdf


54

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

Persuasion as a Prelude to Proof: Theory Acceptance in Accounting Research 
 

John Sennetti, Nova Southeastern University 

Barri Litt, Nova Southeastern University 
 

 

In his 2013 presidential scholar address to the American Accounting Association, Michael Jensen, an 
early author of agency theory, makes an appeal for a new language and new paradigm to introduce his 

proposed theory of integrity in accounting. In the 1970s, agency theory was promoted over other theories 
because it demonstrated empirical evidence of the accounting procedures actually used in practice, 
compared to those that “should” be used according to established rules (see Jensen and Meckling, 1976). 

Contrarily, in this paper, we show scientific theories (Latour, 1987), and in particular accounting 
theories, are actually accepted by a larger persuasion than just the empirical evidence. The early 

empirical evidence on agency theory was often not even statistically significant. As Jensen now 
demonstrates, consistent with Latour (1987), this persuasion begins with a new language and a new 
world view, and continues with a strategy of properly framing expectations for the data outcomes and the 

empirical evidence.     
 

 

 In his 2013 presidential scholar address to the American Accounting Association, Michael Jensen 

describes how he helped to create the contracting (agency) theory of accounting research. Taking issue 

with Friedman’s (1953) ‘positive economics’ theory of corporate behavior, Jensen claimed that it was in 

fact managers, and not corporations, that generate revenue; and so began the self-interest theory of 

accountancy, referred to as ‘positive accounting theory’ (PAT), and more recently as ‘agency theory’ 

(Watts and Zimmerman, 1978; 1986; 1990). This theory argues that because the costs and benefits of 

accounting decisions (e.g., depreciation methods and other accruals) may apply to managers and not 

necessarily to the firm, managers make these decisions not necessarily based on providing the greatest 

accuracy or clarity to the financial statements, but on the benefits to the creators of these statements. For 

example, compensation interests may motivate managers to create the right accounting numbers (the 

bonus hypothesis); creditor interests limit may dividend payments (the debt-equity hypothesis); and costs 

of regulation on income and tax expenses may reduce manager benefits and thereby limit firm behavior 

(the political cost hypothesis) (Watts and Zimmerman, 1978). By challenging the Statement on Theory 

Acceptance (AAA, 1977), which argues that accountants ought to choose the appropriate procedure to 

improve statement accuracy and clarity, this new positive theory of 1978 argued for an axiomatic intuition 

(everyone is assumed to follow self-interest) that could be observed in "empirical regularities" (Watts and 

Zimmerman, 1990, p. 152).   

 We will show evidence against these “regularities” and conclude that this theory promotion process 

was more consistent with that used in promoting new scientific theories (Latour, 1987), a form of 

methodology itself (McCloskey, 1983). We show that this larger persuasion is not limited to McCloskey’s 

rhetoric, although agency has been described as the “rhetoric of revolt” (Mouck, 1992). Instead, this 

persuasion established by Latour (1987) requires a basis of agreement, a recognizable paradigm, and a 

common language for describing events, in order to establish, as first suggested by Kuhn (1974), “a 

prelude to proof.”  And in 2013, this is now explicitly mentioned by Jensen, the author of agency theory 

himself (Erhard and Jensen, 2013).  

 This paper begins with the necessity of rhetoric for any theory acceptance: the necessary appeal to the 

rationality of specific constituencies. Rhetorical strategies used by scientists and described by Latour 

(1987) are introduced as part of the internal rhetoric used to enroll followers and to frame, fortify, and 

then to stack the evidence for the promoted theory as ‘generally accepted.’ We then show that early 

agency advocates 'fortify' their descriptions with the familiar language of economics and appeals to 

authorities such as the social citation index used by the agency journal, the Journal of Accounting and 

Economics (JAE), the Nobel Prize, and the results obtained from scientific modeling. This ‘stacking of 



55

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

the evidence’ then establishes the agency point of view that managers act out of self-interest. Next, 

statistical empirical 'irregularities' are framed as approximate 'regularities' for this pre-established point of 

view. Finally, agency advocates do not even bother to consider alternative hypotheses to those available 

under agency theory. From this point of view, since there are no financial economics critics, Watts and 

Zimmerman (1990) point to a 'non-engagement' of their own critics. This is consistent with the Latour 

(1987) final procedure of ‘closing the argument.’ We begin with the rhetoric necessary for 'proof.’ 
 

Rhetorical Strategies 
 

 Rhetoric, the language used to persuade, or according to Latour (1987), "the language used to mobilize 

resources" (p. 41), has been elevated from its supplementary or "unnecessary" role (Whittington, 1987, p. 

330) to a form of methodology (see Simons, 1990; Nelson et. al., 1987) necessary for the “conduct of 

[accounting] research" and for the "content and quality of its produce - knowledge" (Arrington and 

Schweiker, 1992, p. 512). Persuasion relies fundamentally on rhetoric (Lavoie, 1987, p. 586). Even 

empirical research begins with a reliance on rhetorical persuasion with statements such as: “In the 

absence of a persuasive theoretical structure on which to build a causal chain, the results cannot be used 

to make inferences about the effect of one variable on another” (Abdel-Khalik and Ajinkya, 1979, p. 23). 

 There are many broad classes of rhetorical strategies found in several research areas (see, for example, 

Simons (1990). Each may be divided into two categories: (1) an internal appeal to a specific constituency 

which requires the language of that constituency and many other strategies and devices, or (2) an external 

appeal to as many followers as possible: 
 

“[T]he ... sciences are becoming increasingly rhetorically self-conscious. They are beginning to 

recognize that their discursive practices, both internal and external, contain an unavoidable 

rhetorical component. Internal here refers to those discursive practices that are internal to the 

specific scientific language community; external refers to the discursive practices of that scientific 

(or non-scientific) language communities and the society in general (Gaonkar, D.P., in Simons, 

1990, p. 352). 
 

 Both internal and external rhetoric build a consensus of belief with an intersubjective appeal for a 

common understanding made by the writer. Such an appeal is a result of the writers' ”having actually 

penetrated (or if you prefer, been penetrated by) another form of life, or having, one way or another truly 

"been there". And that, persuading us that this offstage miracle has occurred, is where the writing comes 

in” (Geertz, 1988, p. 4). 
 

Table 1: Rhetoric as Methodology 
 

Subject Reference 

I. The Use of Language to Persuade ((Rhetoric) 
 

1. McCloskey (1983) 

2. Arrington and Schweiker (1992) 

3. Mouck (1992) 

Intersubjective Agreement 
 

  The rhetoric of economics 

  Rhetoric as methodology 

  Rhetorical strategies and devices; the rhetoric of economics, 

of science and of revolt in agency 

II. The Use of Specific Language For 

Constituencies (Latour, 1987) 
 

1. Preston et al. (1992) and Bloomfield et al. 

(1992) 
2. Chua (1992) 

3.  Watts and Zimmerman (1986; 1990) 

Rhetoric to reduce controversy; arguments to(1) fortify, (2) stack, (3) frame, 

and (4) close rhetorical persuasions, given language specific to a discipline. 
 

  The enrolling of physicians and others for the establishment of a national 
health information system 

  Enrolling hospital administrators toward DRG costing 

  Enrolling allies and reducing controversy for agency research in accounting 

 

 Table 1 attempts to illustrate the application of these two rhetorical strategies. The use of language to 

persuade (in part I) makes broad appeals to fairness, justice, and concepts of morality; it appeals to the 

generic everyman. For example, McCloskey (1983) with a broad external appeal claims that the internal 

rhetoric of economics is an "unofficial" methodology. Generally speaking, both internal and external 

rhetoric use those conditions that would otherwise be considered "in context of discovery" (p. 511): 



56

 

upbringing, politics, introspection, theology, history, case study, and devices such as metaphors, 

allegories, parables, and strategies such as simplicity, symmetry, fashion, and appeals to authority. The 

external and internal rhetorical strategies differ only in the subjects of their appeal.     

 For example, presented next in Table 1 is a reference to Arrington and Schweiker (1992) who, through 

a broad appeal, establish that an internal rhetoric is a necessary condition for accounting research. Next in 

Table 1, Mouck (1992) establishes that the appeal of early agency theory is to a specific internal 

economics constituency and that it uses a rhetoric of science and economics described in part by 

McCloskey (1983).  Mouck presents the rhetoric of science agency as having both strategies and devices. 

The strategies include scientific positioning (objective, independent) and devices include appeals to 

authority, expertise, and the evolution of accounting theory. Although Mouck does consider the 

acceptance of agency theory as socio-historical, the rationality given by Mouck for theory acceptance is 

more sociopolitical: 
  

“... to the extent that the story was successful it was due to the fact that it was read by an audience 

attuned to the rhetoric of science, on one hand, and to the Reagan era rhetoric of revolt on the other 

hand, thus allowing the lack of scientific substance to be overlooked...” (Mouck, 1992, p.55)   
 

 Contrary to Mouck (1992), we propose that (1) the rhetoric of agency theory and the rationality for its 

acceptance (Table 1, part 1) is more historic, hence (2) it will have continued social acceptance by its 

constituencies, but most importantly (3) the advocacy of agency theory (Watts and Zimmerman, 1986; 

1990) is consistent with the rhetorical strategies used by successful scientists as described by Latour 

(1987) and now used by Erhard and Jensen (2013). Through Latour's analysis, this paper attempts to 

answer questions raised by Mouck (1992) and others on the acceptance of accounting research.   

 Latour (1987) presents a model of how scientists and technologists who "discover" facts actually 

enroll others into a particular scientific process of representing what has been observed. Briefly, after 

observing events which are initially described by a weak rhetoric of indecision about what has been 

observed, scientists form opinions, perform laboratory tests, begin to take positions, and then seek to 

defend these positions with evidence that is said to be 'fabricated' and enhanced with rhetoric designed to 

enroll users and reduce controversy. 

 Whether to enroll new users or just to reduce controversy, this rhetoric includes arguments (1) to 

fortify a particular position taken, (2) to stack the evidence incrementally in favor of that position, (3) to 

frame its understanding, and (4) to close arguments against it (see Latour, 1987, pp. 21-57; see 

applications in Preston et al., 1992, pp.574-575, and also Figure 1, later, adopted from p.566).  

 In order to aid in a practical understanding, we provide an example of Latour’s (1987) persuasive 

rhetoric strategy points. This process starts with the fortification of a basis of agreement, established by 

calling upon the familiar. For example, in the development of a hospital information system, it may be 

necessary to enroll users, such as physicians (see Preston et. al., 1992 and Bloomfield et. al., 1992 in 

Table 1, part II). Physicians, not primarily interested in such a computer system, must be persuaded. The 

basis of agreement is patient care: their primary concern. Quality patient care is framed as requiring the 

best possible  diagnostic  information, which can only  be provided by the desired hospital 

information system. 

Further, the system has “ease and efficiency” (as framed) for physicians. This framing can then 

establish the closing argument that, "to ignore such an information system is to ignore patient care." 

Physicians are then "enrolled" and become part of a larger constituency of accountants, managers, 

auditors, and others at the hospital who "need" this system. Similarly, Chua (1992) uses Latour's (1987) 

model to describe the enrolling of hospital administrators to a special costing system.  

 Latour's rhetorical strategies also can be used to defend positions taken with evidence properly framed 

(not falsified but 'fabricated'). One example describes how the DNA molecule was first discovered (1987, 

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

Sequence of Acts 

Descriptions Framed By Positions Taken 



57

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

p.14). First observed as a possible triple helix, the molecule's final description as a double helix resulted 

after the evidence had been properly examined under the many (fortified and stacked) reasons supporting 

the double helix. 
 

Figure 1: Creation of Facts from Scientific Discoveries 

 (Based on Latour, 1987, p.44; Preston et al., 1992, p.566) 
 

Sequence of Acts Act 

1 Events Observed 

2 Weak Descriptions Recorded 

3 Descriptions Framed By Positions Taken 

4 Positions Fortified, Evidence Stacked 

5 Positions Closed, Objections Isolated 

6 Facts Accepted 

 

 Figure 1 presents a model of Latour's description of how facts are 'fabricated' (Latour, 1987, p.44, 

Figure 1.6; Preston et al., 1992, p.566). From the observance of events comes weak and hesitant possible 

descriptions when scientists begin to form theories, build convincing arguments, modify theory and form 

closing arguments, and finally reduce controversy to a point where 'facts' are created and objectors are 

silenced. Along the way, networks of resolved controversies are formed through operational and 

rhetorical strategies.  

 The focus of this essay is limited to the use of rhetorical strategies described by Latour and is not 

intended to be comprehensive enough to cover all of Latour's concepts, such as his rules of method or his 

principles used to describe scientific discoveries. Nor can we do Latour (1987) justice in the omission of 

all his descriptions of how scientists attempt to increase the number of allies, laboratories, and machines 

to increase the acceptance of their theories. We present a limited form of Latour’s analysis here since 

Watts and Zimmerman (1990) use Latour's rhetorical strategies mostly to defend positions taken.   

 We begin with analysis of the first strategy, what Latour calls the process of 'enrolling,' which actually 

is the process of fortifying accepted 'facts,’ found here in the economic language and paradigms of 

financial economics. 
 

Enrolling and Fortifying the Basis of Agreement 
  

 The self-interest economic values, paradigms, and internal language necessary to accept agency theory 

are discussed by many authors, including Tinker et al. (1982) and Mouck (1992). Accounting policy is 

said to serve the firm and management's interests if accounting procedures are chosen to minimize the 

contracting costs to the firm. A bonus plan, for example, is said to direct the self-interests of the manager 

to those of the firm, thereby reducing agency costs to the firm. Since the manager's earnings depend on 

the firm's, ceteris paribus, earnings are less likely to be deferred.       
 

Fortifying 
 

 Watts and Zimmerman (1986), consistent with financial economics of their time (and now challenged 

by behavioral economics), take a cost-benefit approach to theory acceptance, arguing that although the 

predictions of agency may not be perfect, the theory's benefits, such as its "intuitive appeal," far exceed its 

costs of imperfect predictions (p. 12). Following this school of thought, agency advocates had potentially 

many allies, starting with Jensen’s first notable paper entitled ‘The contracting cost theory of the firm’ 

(Jensen and Meckling, 1976), and his Journal of Financial Economics.    
 

Stacking the Evidence
 

 At that time, there were at least three more reasons beyond those mentioned by Whitley (1988) for the 

continued acceptance of agency. First, having financial economics as a home provided these accounting 

researchers with the status of international acceptance and the attendant image of an academic discipline 



58

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

deny agency theory is to deny the authority of well-grounded and accepted economic theory. Finally, to 

refute agency theory is to refute the very authority of Friedman's Nobel Prize in 1953, the beginning of 

the ‘positive theory of economics.’    

 As suggested by Sterling (1990) with regard to the initial acceptance of Friedman’s theory in the 

1950’s, even if “a majority of economists” do not believe in 'positive economics,' the Nobel Prize still has 

a higher authority than a "majority of economists" (p. 122). Indeed by 1978, the majority of financial 

economists from Jensen’s Journal of Financial Economics favored agency theory. Accordingly, Latour 

(1987) explains that "truths" are only verified when an overwhelming number of authorities claim 

acceptance, in this case, accounting research authorities (p. 61). 

 With particular regard to accounting professors, the Nobel Prize may have a special allure. There are 

few occasions for North American accounting professors to be internationally recognized outside of the 

actions of public accountants, such as those from the AICPA or CICA. Accordingly, such actions may 

bring pride or tarnish upon the image of accounting professors. But the possibility that the Nobel Prize 

may be awarded to an accounting professor elevates the entire academic profession of accounting. 

 Other authorities were recruited. Watts and Zimmerman accrued a certain status when their Journal of 

Accounting and Economics (JAE) was ranked above both the University of Chicago's Journal of 

Accounting Research and the American Accounting Association's The Accounting Review by the citation 

index of outside references (Editorial, 1992, pp.3-6).   

 Another authority, this one taken from science, was the practice of economic modeling, seen as more 

scientific than non-mathematical modeling. The self-interest economics theory is supported the 1992 

Nobel Prize of another University of Chicago Economics Professor, Gary Becker, for his research on the 

economic motivation of social choices (such as those made by women in the workforce). Whitley (1988) 

chronicles this kind of modeling and the rise of mathematical sciences in social research, which began in 

the United States prior to Friedman's 'positive economics' (1953):  
 

This belief in scientific knowledge as a crucial resource in maintaining and improving the social 

order has been particularly strong in the U.S.A. since the Progressive era (Bledstein, 1976, pp. 

123-127, 324-331; Wiebe, 1967, Chapter 6) and received considerable reinforcement from its 

evident military utility in World War II (Kevles, 1977, pp. 367- 392, as given by Whitley, 1988, p. 

641). 
 

 Whitley (1988) adds that many other reasons existed for the acceptance of agency theory in North 

America, including (1) the prestige of the 1959 Carnegie and Ford Foundations which supported scientific 

research in reports found in business education publications, (2) the dominance of universities in their 

contribution to professional training in this area, and (3) the related growth of legal interventions, which 

use the same objective functionalist approach in presenting scientific evidence as used in economics. The 

Latour (1987) analysis would consider of all these aspects as comprising the fortifying process for theory 

acceptance.   
 

Acknowledging Controversy and Creating Acceptable Forms of Evidence  
 

 Consistent with Latour's (1987) rules of rhetoric method, which require investigators to study the 

conditions and processes under which facts are created, we will now investigate the conditions under 

which agency theory was created in 1978, and the subsequent controversy and fabrication of research 

evidence.  An earlier North American controversy in financial reporting that arose from the American 

Accounting Association’s Committee on Concepts and Standards for External Financial Reports led to the 

statement that, "until consensus paradigm acceptance occurs, the utility of accounting theories in aiding 

policy decisions is partial" (AAA, 1977, p.51). Further, because no one theory could satisfy all paradigms, 

"defenders of a particular paradigm are forced to rely on persuasion rather than logic and empirics in 

attempting to defend a proposal" (AAA, 1977, p.48).  



59

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

 In search of paradigms, Professor Robert R. Sterling, a member of this Committee, suggested that 

future theories for accounting come from established theories and paradigms in the behavioral and 

economic sciences (Sterling, 1970). With perfect timing, agency theory was introduced by Watts and 

Zimmerman (1978), in part using the size hypothesis (a proxy for the political cost hypothesis), an 

economics paradigm solution for accounting procedure choice: the larger the firm, the larger the firm's 

exposure to the political costs of windfall profits taxes, labor strikes, government regulation, and the like; 

therefore, the firm will be more likely to defer earnings to later periods.  

 Ignoring the Committee's appeal for 'persuasion,' the empirical evidence on this size hypothesis was 

presented and argued as probative (Watts and Zimmerman, 1978). Later, a corresponding theory-driven 

persuasion was given in Watts and Zimmerman (1979), but this exacerbated rather than reduced the 

controversy (Watts and Zimmerman, 1990). Table 2 presents some of the subsequent works that addresses 

this controversy, most of which are summarized in Watts and Zimmerman (1986) or by Watts and 

Zimmerman (1990). These studies directly correlate the size and other agency theory hypotheses with the 

choice of accounting procedures used to defer earnings, and these studies were offered in response to 

design, validity, and statistical questions that arose in the 1978 study. 
 

Table 2: Summary of Results: Size and Choice of Accounting Procedures to Defer Earnings* 
 

Topic Author(s) 

Statistically significant 

size variable 
(p-value<0.05)? Comments 

General price level  

adjustments (GPLA) 

Watts and Zimmerman  

(1978) 

Yes 1. This was the original size hypothesis. No hold 

out sample is used in this or in any “confirming” 
study. 

2. This study and its results were challenged by 

McKee et al. (1984). 
3. Model R2 is 0.20, for in-sample data only 

GPLA McKee et al.  

(1984) 

No 1. Not significant for a holdout sample and Watts 

and Zimmerman (1978). 
2. The Watts and Zimmerman (1978) best model, 

properly stated, was not as successful as the 

naïve alternative. 
3. Model R2 for holdout data is only 0.09. 

Depreciation Hagerman and Zmijewski  

(1979) 

Yes The large firm strategy is not superior to the naive 

strategy (no holdout sample) and is not confirming. 

Oil and Gas Dhaliwal  
(1980) 

No The D/E ratio is significant; size has a p-value of 
0.10. 

Depreciation, ITC, 

Inventory 

Zmijewski and Hagerman  

(1981) 

Yes The poor R2 occurred despite the fact that there was 

no holdout sample and even then, the success rate 

was only 40% (R2 = .09). The result is not superior 
to naive alternative 

Interest capitalization Bowen, Noreen and Lacey  

(1981) 

No The size variable is significant, but only for oil 

industry 

Depreciation Dhaliwal et al.  
(1982) 

No The size variable is significant only at the 0.15 
level. 

Oil and Gas Lilien and Pastena  

(1989) 

No Significance level is 0.06. 

Research and Development Daley and Vigiland  

(1983) 

No Results show that increasing earnings occur for 

smaller firms, but paradoxically, as the firms get 

larger, the importance of size diminishes rather than 
increases. 

Income Tax Zimmerman  

(1983) 

Yes 1. "Size" limited to industry, limited in time. 

2. Smaller, not larger trade firms defer (opposite 

results). 
3. Size is considered a poor proxy for political 

cost. 

Oil and Gas Deakin  
(1979; 1989) 

No Since size is only significant for largest 24 firms, 
considered size not related/poor predictor.                                                                            

* taken in part from Watts and Zimmerman (1986, p. 258) 

 

 The design and validity questions of Ball and Foster (1982) and later Williams (1989) discredited the 

size hypothesis because of its weak construct validity. Christenson (1983) and Mckee et al. (1984) also 



60

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

invalidated the research of Watts and Zimmerman (1978). Christenson's analysis of Watts and 

Zimmerman made an appeal for better scientific research design (1983). McKee et al. (1984) responded 

with a better research design, first in replicating the in-sample Discussion Memorandum (DM) test of 

Watts and Zimmerman (1978), and then in providing additional evidence from another out-of-sample test 

on the 1974 Exposure Draft (ED). McKee et al. (1984) also challenged the methodology, model, and 

'successful' predictions acclaimed by Watts and Zimmerman (1978): 
 

“The theory still fails to classify correctly a disturbingly large number of observations, particularly 

in the ED sample. Moreover, the number of misclassified observations exceeds the expected misses 

from a maximum chance criterion, indicating that a model which predicts that all submissions will 

be unfavorable produces superior classificatory success” (McKee et. al., 1984, p. 650). 
 

 The size hypothesis was not supported. The ignored naive hypotheses and other hypotheses of Ball 

and Foster (1982) still continue to serve as alternative hypotheses for the subsequent appeals by Watts and 

Zimmerman (1986; 1990). Finally, it can be seen in Table 2 that only 3 of the 9 studies on the agency 

hypotheses given by Watts and Zimmerman (1986) yield what is generally accepted as a statistically 

significant size hypothesis, even though these studies were given in response to critics, and were later 

acclaimed for their empirical 'regularities.' 

 In order to fortify their position on their theory with at least the appearance of good science, Watts and 

Zimmerman (1986) seem to agree that the size hypothesis was probably better expressed as the original 

political cost hypothesis. After a discussion of size as a proxy, Watts and Zimmerman conclude that "the 

evidence is consistent with the political cost hypothesis. However the result only appears to hold for the 

largest firms" and "is driven by the oil and gas industry" (1990, p.140). 

  It would appear that Watts and Zimmerman (1986; 1990) do establish some limitations of early 

agency models, and as editors of JAE, they have reported some contrary findings for the size hypothesis 

in their journal. This reporting gives at least an appearance of fairness. Yet, their evidence for the size 

hypothesis still serves as evidence for a pre-established point of view. For example, inconsistent with the 

size hypothesis, Moyer (1990) finds that large commercial banks do not defer earnings. For banks, a 

political incentive to smooth earnings exists by the industry requirement to maintain regulatory capital. 

Although the size hypothesis fails here, indirectly the bonus hypothesis (and agency) is promoted, since 

the board executive salaries are related to regulatory capital and more importantly, the "costs" of 

regulation can be minimized by maintaining regulatory capital (see Scholes, Wilson, and Wolfson, 1990), 

and thus agency is again in part sustained. 

 Watts and Zimmerman had to at least appear to be reducing controversy by seeking "alternative 

hypotheses." Unfortunately, this seeking was done only in the contractual, firm-related, agency context 

and world view (1990, p. 145). Non-agency variables were not considered. For example, in a non-JAE 

publication, Cahan (1992) points to a non-agency, non-firm-specific political cost variable and an 

alternative non-political cost hypothesis. But to seek non-agency variables is in part to deny 'positive 

economics,' to deny the Nobel Prize, and to deny generally accepted "truths."   

 This type of behavior, seeking only allies and qualifying other evidence, is consistent with Latour's 

(1987) description of successful scientists and their references. To deny that alternative hypotheses exist 

outside of the agency world view is to appeal to internal rhetoric, and to frame and close the controversy, 

perhaps unfairly. Latour (1987) adds guidance to critics who ask why Watts and Zimmerman seem to 

limit conversation.  
 

Use of Latour's Rhetoric to Reduce Controversy 
 

 Watts and Zimmerman (1986; 1990) respond to the statistical controversy of their critics with the 

rhetorical fortifying, stacking, framing, and closing strategies described earlier. "When controversies flare 

up, the literature becomes technical" (Latour, 1987, p.30). Fortifying arguments require (1) "bringing 

friends in," (2) "referring to former texts," and (3) "being referred to by later texts." All nine of the 

supporting references for Watts and Zimmerman (1978) given in Table 2 and taken from Watts and 



61

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

2013, Vol. 4, No. 1, 54-67 

Zimmerman (1986) can be identified as "friends" of the 'Rochester School' (see Christenson, 1983). All 

but two references are from their own Journal of Accounting and Economics. In Watts and Zimmerman 

(1990), there are references to Wong (1988), Sutton (1988), Jones (1990), Moyer (1990), and Christie 

(1990), all of whom eventually were published in this journal. This assures that Watts and Zimmerman 

will be "referred to by later texts” (Latour, 1987).   

 Also, fortifying requires going "from numbers of allies and resources to still greater numbers" that go 

beyond the need for the prestige of authorities such as the Nobel Prize (Latour, 1987, p.49). For Watts and 

Zimmerman (1986), 'beyond authority,' means that agency theory has reached an "intuitive appeal" (p.12).  

This is a typical fortifying statement for establishing a generally accepted, beyond-authority 'fact.' More 

fortifying reasons are also given as part of the framing of tests, presented later.   

     Next, the stacking of the evidence is to be done incrementally: 
    

“First rule: never stack two layers exactly one on top of the other; ...Second rule: never go straight 

from the first to the last layer ...Third rule (and most important): prove as much as you can with as 

little as you can, considering the circumstances” (Latour, 1987, p.51). 
 

 An illustration of these rules can be found in Watts and Zimmerman (1986) where they stack the need 

for theory, chapter by chapter, beyond what could logically be expected from any text. Following 

the ’second rule,’ the need for a positive theory different from any general economics theory is not 

presented in the first chapter, as we are to assume all economics theories are positive. After establishing a 

role for theories, the text simply introduces arguments for any economics-based accounting theory. The 

fact that this text is an advocate of only one of two agency theories, the one defined for firm-related, 

economics-based accounting theory, is not disclosed. The beginning of the text, to the uninformed reader, 

could simply be about the efficient market hypothesis and its relationship to accounting research. 

 Next, Watts and Zimmerman (1986) follow Latour's 'third rule.' By establishing as much as they can 

with as little as necessary, they 'evolve' the information and transactional costs of the "costless" efficient 

market hypothesis to a knowledge state that requires a theory for such costs in the next chapters. These 

chapters then form a chronological history of the University of Chicago economics-based financial 

accounting research. Each of the Chapters 3-6 summarize the related research questions and then pose 

some of the Chicago answers available at some point in time. By posing both the questions and answers 

suggested by traditional neoclassical economic analysis, Watts and Zimmerman do not conclude with a 

need for an agency theory until Chapter 8, where the 'last layer' of the stacking process is found. After 

certain market inefficiencies are questioned, "the contracting role of accounting and its relation to the role 

assumed in Chapters 3-6" is finally investigated (p. 180). No mention is made of the "other" 

economics-based agency theory of the Chicago school, or other economics-based solutions to these 

inefficiencies. No mention of the forthcoming contracting role is given earlier in the outlines provided in 

Chapters 1 and 2. This layer was saved for last. Other examples of stacking may be provided, but Watts 

and Zimmerman make more obvious the use of other devices. 

 First, since their position is one of advocacy, Watts and Zimmerman respond to contrary evidence by 

framing the size hypothesis as one evolved from its "early progress" (1986, p. 265). Hence, it can be 

framed that the size hypothesis was just a "demonstration," and that the correct test of agency theory was 

not performed. They continue to refer to the problems of their research as those problems of measurement, 

and those problems of correct modeling the-yet-to-be-recognized alternative, firm-related (agency) 

variables (1990). 

 If an alternative frame exists, it is not given by Watts and Zimmerman (1990). It is interesting to 

compare their comments on the "early progress" of agency theory with those of Lev (1989) on the early 

progress of a related but non-agency area, the financial accounting, market-based research on the 

relationship of earnings to price changes. With success similar to agency theory research (low power, low 

R2
 models), Lev proclaims that success had not been achieved in understanding the underlying earnings 

process.  He instead requests new models and a new approach. In comparison, Watts and Zimmerman do 

none of this.  Alternatively, they continue to frame 'a pre-established point of view.' 



62

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

61 

Second, the framing strategy is to admit no complete failures. Although "poor models" of agency 

theory are admitted (Watts and Zimmerman, 1986; 1990), they are presented as providing new challenges 

to researchers, as the size hypothesis did in 1978, to find better firm-related agency theory variables.  

Limited success is automatically assumed because of the continued "empirical regularities" found (p. 

152).  

Finally, the framing of the correct language and tests for agency theory also requires more fortifying, 

which Latour calls "overwhelming" arguments. For example, in continuing their defense of the size 

hypothesis, Watts and Zimmerman (1990) refer to Christie (1990) who aggregates tests where statistical 

significance for the size hypothesis has been found (ignoring low R2
 levels and the non-statistically 

significant studies). By aggregating the results of these tests, Christie creates the overwhelming Bayesian 

posterior probability statement of 0.99 that size and other agency theory variables have explanatory power.  

Just as any Bayesian analysis requires a subjective prior distribution, this particular analysis is more 

subjective persuasion than recognized evidence. Despite McKee et al. (1984) and much more evidence, 

the size hypothesis of Watts and Zimmerman (1978) continues to be 'confirmed' and framed by Christie 

(1990), and now, because of the related controversy, it is framed as a pioneering effort (Watts and 

Zimmerman, 1990). 

To summarize these framing considerations, Watts and Zimmerman (1986; 1990) would have 

followers (a) consider statistical questions to be just imperfect modeling and measurement questions and 

not ones of validity (1990); (b) consider such irregular results to be somewhat regular given the statistical 

significance of the R2
 and the absence of other established, non-agency theory alternative explanations 

(1990); (c) consider the "usefulness" of the prediction in terms of the users' welfare (which may be 

depend less on statistical success) (1986, p.14); and (d) consider that since size and other alternative 

agency theory-related variables may be proxies for political costs, agency costs, and other "contracting" 

costs, better ones soon will be found, and these new findings will yield research benefits to those who 

find them (1986, p.358-62; 1990, p.151-3).   
 

Closing and Summary Persuasion for Reducing Controversy  
 

The framing of agency theory research is just part of the rhetorical persuasion needed. The reader may 

still disagree or have doubts, and “may still reach different conclusions” (Latour, 1987, p.56). The 

solution is to lay out the text so that the reader has only one way to go.  One does not truly engage the 

critics directly (Latour, 1987).   
 

“The nature of the game is exactly like that of building a dam. It would be foolish for a dam 

engineer to suppose that water will always obey his wishes…any engineer should start with the 

principle that if water can leak away it will. Similarly with readers, if you leave the smallest outlet 

open to them they will rush out...If the digging and damming is well set up, the reader, although 

taken in, will feel entirely free” (Latour, 1987, p.57). 
 

All objectors' moves are to be controlled through closing (damming) arguments. Doubters are to be 

made to feel alone and "without support and ally, alone in [his/her] profession, or, even worse, isolated 

from the [research] community or maybe, still more awful, sent to an asylum!" (Latour, 1987, p.59). 

For Watts and Zimmerman, this persuasion also includes closing arguments for (a) the paradigm, (b) 

the forms of logic, (c) the evidence, and (d) the plausible demonstrations of success. For (a) the paradigm, 

they state that their paradigm is accepted, one that "has been successful in accounting", and if "the alleged 

errors [were] important...the referees would adopt the suggestions" (1990, pp. 149-150). That is, Watts 

and Zimmerman would have us believe that the objections are not important, not mainstream. Since the 

objections are not important, they seem to rightfully have no responsibility for other (objector) paradigms. 

For (b) the forms of logic, to those who "place unreasonable demands" on their methodology (such as 

Hines, 1988), Watts and Zimmerman defend the right of new ideas to evolve through "the evolutionary 

nature of accounting research" (1986, p. 362; 1990, p.149). That is, they would have us believe that 

objectors are unreasonable and have not 'evolved'. For (c) the evidence, Watts and Zimmerman (1990) 



63

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

argue that their forms of evidence have been accepted and that "criticisms [of agency theory] have failed 

the market test" [i.e., the objectors have failed] (p. 149), even though "the fundamental questions of the 

relation between the observed empirical regularities and this theory remain unexamined" (Leftwich, 1990, 

p. 37).   
 

Conclusion 
 

We find Latour’s (1987) rhetorical strategies to create a persuasion in the research and widespread 

acceptance of agency theory (Watts and Zimmermans, 1978; 1986; 1990; Jensen and Meckling, 1976). 

Our findings add to the current growing body of literature now reinforced by Erhard and Jensen (2013) 

that recognizes rhetoric as methodology necessary for theory acceptance and ‘proof.’ The limited 

short-term interest component of agency theory is now rejected by Erhard and Jensen (2013) in a new 

attempt at a persuasion. The extent of this new persuasion, i.e., framing, fortifying, stacking, and closing 

of arguments, is now just beginning, but as before; it must first be reinforced by a rising world view. We 

see this new world view in the increasingly popular acceptance of behavioral economics over financial 

economics (Kahneman, 2011). 
 

REFERENCES 
 

Abdel-Khalik, A., & Ajinkya, B. 1979. Empirical research in accounting: A methodological viewpoint. 

 Accounting Education Series No. 4, American Accounting Association. 
 

American Accounting Association. 1977. Committee on concepts and standards for external financial 

 reports. Statements on Accounting Theory and Theory Acceptance. Sarasota, Florida: AAA. 
 

Arrington, C., & Francis, J. 1989. Letting the chat out of the bag: Deconstruction, privilege and 

accounting research. Accounting Organizations and Society, 14(1-2): 1-28. 
 

Arrington, C., & Schweiker, W. 1992. The rhetoric and rationality of accounting research. Accounting 

 Organizations and Society, 17(6): 511-533. 
 

Ball, R., & Foster, G. 1982. Corporate financial reporting: A methodological review of empirical research. 

Studies in current research methodologies in accounting: A critical evaluation. Journal of Accounting 

 Research, 20(5): 161-234.   
 

Bloomfield, B., Combs, R., Cooper, D., & Rea, D. 1992. Machines and maneuvers: Responsibility 

accounting and the construction of hospital information systems. Accounting, Management, and 

Information Technologies, 2(4): 197-219. 
 

Bowen, R., Noreen, E., & Lacey, J. 1981. Determinants of the corporate decision to capitalize interest. 

Journal of Accounting and Economics, 4(3): 151-179. 
 

Cahan, S. 1992. The effect of antitrust investigations on discretionary accruals: A refined test of the 

political cost hypothesis. Accounting Review, 67(1): 77-96. 
 

Christenson, C. 1983. The methodology of positive accounting. Accounting Review, 58(1): 1-22. 
 

Christie, A. 1990. Aggregation of test statistics: An evaluation of the evidence on contracting and size 

hypotheses. Journal of Accounting and Economics, 12(1): 15-36. 
 

Chua, W. 1993. Following the fabrication of product costs the case of three public hospitals. Working 

Paper, University of New South Wales. Presented at the Third Critical Perspectives on Accounting 

Symposium. 
 

 



64

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

63 

Daley, L., & Vigeland, R. 1983. The effects of debt covenants and political costs on the choice of 

accounting methods: The case of accounting for R & D costs. Journal of Accounting and Economics, 

6(4): 195-211. 
 

Deakin, E. 1979. An analysis of differences between non-major oil firms using successful efforts and full 

cost methods. Accounting Review, 54(4): 722-734. 

Deakin, E. 1982. Rational economic behavior and lobbying on accounting issues: Evidence from the oil 

and gas industry. Accounting Review, 64(1): 137-151. 
 

Dhaliwal, D., Salomon, G., & Smith, E. 1982. The effort of owner versus management control on the 

choice of accounting methods. Journal of Accounting and Economics, 5(3): 41-53. 
 

Editorial. 1992. Five year report. Journal of Accounting and Economics, 15(1): 3-6. 
 

Erhard, W. & Jensen, M. 2013. Putting integrity into finance: A purely positive approach. Harvard 

Business School Negotiation, Organizations and Markets Research Papers, NO. 12-074. http://hbswk. 

hbs.edu/item/6995.html & http://papers.ssrn.com/sol3/papers.cfm?abstract_id= 1985594. 
 

Friedman, M. 1953. Essays in positive economics. Chicago, IL: University of Chicago Press. 
 

Gadamer, H. 1987. Economics in perspective. Boston, MA: Houghton-Mifflin. 
 

Geertz, C. 1988. Works and lives: The anthropologist as an author. Stanford, CA: Stanford University 

Press. 
 

Habermas, J. 1987. The philosophical discourse of modernity. Cambridge, MA: MIT Press. 
 

Hagerman, R., & Zmijewski, M. 1979. Some economic determinants of accounting policy choice. 

Journal of Accounting and Economics, 2(3): 141-161. 

Hines, R. 1988. Popper's methodology of falsification and accounting research. Accounting Review, 

63(4): 651-662. 
 

Holthausen, R., & Leftwich, R. 1983. The economic consequences of accounting choice: Implications of 

costly contracting and monitoring. Journal of Accounting and Economics, 6(3): 77-117. 
 

Jensen, D. 1989. Agency theoryhologies of science, precognition, and modern psychophysics. Skeptical 

Inquirer, 13(4): 142-160. 
 

Jensen, M. 2013. Future of academic research. Plenary Session, American Accounting Association 

Annual Meetings, Anaheim, California, August 6, 2013. http://commons.aaahq.org/posts/0696eab8b b 

& http://commons.aaahq.org/posts/c6e882eda6. 
 

Jensen, M., & Erhard, W. 2011. A 'value-free' approach to values, Harvard  Business School NOM 

Unit. Working Paper No. 11-010; Barbados Group Working Paper No. 10-06; Simon School Working 

Paper No. FR10-26. Available at SSRN: http://ssrn.com/ abstract=1640302. 
 

Jensen, M., & Meckling, W. 1976. Theory of the firm: Managerial behavior, agency costs and ownership 

structure. Journal of Financial Economics, 3(4): 305–360. 
 

Jensen, M., Murphy, K., & Wruck, E. 2004. Remuneration: Where we've been, how we got to here, 

what are the problems, and how to fix them. HBS NOM Paper No 04-28; ECGI - Finance Working 

Paper No. 44/2004. Available at SSRN: http://ssrn.com/abstract=561305 
 

Jevons, W. 1888. Theory of political economy. London: Macmillan. 
 



65

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

Jones, J. 1988. The effect of foreign trade regulation on accounting choices, and production and 

investment decisions. Working Paper, University of Michigan. 
 

Kahneman, D. 2011. Thinking, fast and slow. London: Penguin. 
 

Kuhn, T. 1974. Reflections on my critics. in I. Lakatos & A. Musgrove, (Eds.), Criticism and growth of 

knowledge (3). London: CambridgeUniversity Press: 238-265.  
 

Kuhn, T. 1996. The structure of scientific revolutions. Chicago, IL: University of Chicago Press. 
 

Latour, B. 1987. Science in action. Cambridge, MA: Harvard University Press. 
 

Lavoie, D. 1987. The accounting of interpretations and the interpretation of accounts: The communicative 

function  of 'The language of business.' Accounting, Organizations and Society, 12(6): 579-604. 
 

Leftwich, R. 1990. Aggregation of test statistics: Statistics vs economics. Journal of Accounting and 

Economics, 12(1): 37-44. 
 

Lehman, C., & Tinker, T. 1987. The real cultural significance of accounts. Accounting Organizations 

and Society, 12(5): 503-22. 
 

Lev, B. 1989. On the usefulness of earnings and earnings research: Lessons and directions from the 

decades of empirical research. Journal of Accounting Research, 27(5): 153-201. 
 

Lilien, S., & Pastena, V. 1982. Determinants of intramethod choice in the oil and gas industry. Journal of 

Accounting and Economics, 4(4): 145-170. 

McCloskey, D. 1983. The Rhetoric of economics. Journal of Economic Literature, 21: 481-517. 
 

McKee, A., Bell, T., & Boatsman, J. 1984. Management preferences over accounting standards: A 

replication and additional tests. Accounting Review, 59(4): 647-659. 
 

Mouck, T. 1993. The rhetoric of science and the rhetoric of revolt in the story of positive accounting 

theory. Accounting, Auditing, and Accountability Journal, 5(4): 35-56. 
 

Moyer, S. 1990. Capital adequacy, ratio regulations, and accounting choices in commercial banks. 

Journal of Accounting and Economics, 12(3): 123-154. 
 

Nelson, J., Megill, A., & McClosky, D. 1987. The rhetoric of the human sciences. Madison, Wisconsin:  

University of Wisconsin Press. 
 

Okcabol, F., & Tinker, A. 1990. The market for positive theory: Deconstructing the theory for excuses. 

Advances in Public Interest Accounting, 3: 71-95. 
 

O'Neil, C., Cathey, J., & Flesher, T. 1988. An analysis of Ph.D. dissertations in taxation: 1977-1985. 

Issues in Accounting Education, 3(1): 120-130. 

 

Preston, A., Cooper, D., & Coombs, R. 1992. Fabricating budgets: A study of the production of 

management budgeting in the national health service. Accounting Organizations and Society, 17(6): 

561-593. 
 

Reiter, S. 1988. Economic imperialism and the crises in financial accounting research. Critical 

Perspectives on Accounting, 9: 143-171. 
 

Robertson, T. 1971. Innovative behavior and communication. Philadelphia, PA: Holt Rinehart and 

Winston, Inc. 



66

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

Ronen, J., & Aharoni, A. 1989. The choice among accounting alternatives and management compensation: 

Effects of corporate tax. Accounting Review, 64(1): 69-86. 
 

Scholes, M., Wilson, G., & Wolfson, M. 1990. Tax planning, regulatory capital planning, and financial 

reporting strategy for commercial banking. Review of Financial Studies, 1(4): 625-650. 
 

Simons, H. 1990. The rhetorical turn: Invention and persuasion in the conduct of inquiry. Chicago, 

IL: University of Chicago Press. 
 

Sterling, R. 1970. On theory construction and verification. Accounting Review, 45(3): 444-457. 
 

Sterling, R. 1990. Positive accounting: An assessment. ABACUS, 26(2): 97-135. 
 

Sutton, T. 1988. The proposed introduction of current cost accounting in the U.K.: Determinants of 

corporate preference. Journal of Accounting and Economics, 10(3): 127-149. 
 

The Economist. 1987. The philosophy of science in schools brief. The Economist: 70-71. 
 

Tinker, A., Merino, B., & Meimark, M. 1982. The normative origins on positive theories: Ideology and 

accounting thought. Accounting Organizations and Society, 7: 167-200. 
 

Viator, R. Accounting information systems doctoral dissertations: 1997. Journal of Information 

Systems, 12(2): 109-111. 
 

Watts, R., & Zimmerman, J. 1978. Towards a positive theory of the determination of accounting 

standards. Accounting Review, 53(1): 112-134. 
 

Watts, R., & Zimmerman, J. 1979. The demand and supply of accounting theories: The market for 

excuses.  Accounting Review, 54(2): 273-305. 
 

Watts, R., & Zimmerman, J. 1986. Positive accounting theory. Englewood Cliffs, N.J.: Prentice-Hall. 
 

Watts, R., & Zimmerman, J. 1990. Positive accounting theory: A ten year perspective. Accounting 

Review,  65(1): 131-156. 
 

Whitley, R. 1988. The possibility and utility of positive accounting theory. Accounting, Organizations, 

and Society, 13(6): 631-645.  
 

Whittington, G. 1987. Positive accounting: A review article. Accounting and Business Research, 17(68): 

327-336. 
 

Williams, P. 1989. The logic of positive accounting research. Accounting Organizations and Society, 

14(5-6): 455-468.  
 

Wong, J. 1988. Political costs and intraperiod accounting choice for export tax credits. Journal of 

Accounting and Economics, 10(1), 37-51. 
 

Zeff, S. 1966. American Accounting Association: Its first fifty years, 1916-1966. Sarasota, FL: 

American Accounting Association. 
 

Zmijewski, M. & Hagerman, R. An income strategy approach to the positive theory of accounting 

standard  setting/choice. Journal of Accounting and Economics, 4(3):129-149. 
 

 

John Sennetti is a professor of graduate statistics and accounting at Nova Southeastern University, whose 

Ph.D. is from Virginia Tech, supports the research of his graduate students, and has also published in 



67

Sennetti and Litt                                                                              Advances in Business Research 

2013, Vol. 4, No. 1, 54-67 

Auditing: A Journal of Practice and Theory, Journal of Finance, Journal of Business Ethics, Advances in 

Accounting Behavioral Research, and many other journals. 
 

Barri Litt is an assistant professor of accounting at Nova Southeastern University. She received her Ph.D. 

in accounting from Florida International University. She is also an Elijah Watt Sells Award recipient, 

having scored in the top ten in the nation on the 2007 Certified Public Accountant Examination. Her current 

research interests include auditing, accounting education, and corporate social responsibility.  


