Statement on Accounting Theory and Theory Acceptance

I!#?

T%

ftneri

canAc

counti

n8Ass

ociatio

nU

/AitC

ommi

tteeo

nConc

eptsand

Standa

rds

F; ccep

tan

ce

/

STATEMENT ON ACCOUNTING THEORYAND THEORY ACCEPTANCE

Committee on Concepts and Standardsfor Externa] Financial Reports

AMERICAN ACCOUNTING ASSOCIATION

Copyright, American Accounting Association, 1977, All rights reservedLibrary of Congress Catalog Card Number 77-747-36

Printed in the United States of America

CONTENTS

PREFACE ix

CHAPTER 1Introduction l

CHAPTER2

AlternativeTheory Approaches 5

CHAPTER3

Criticisms of Present Theory Approaches 31

CHAPTER4Difficulties in Achieving Consensus: A General View 41

CHAPTER5Implications 49

1976-77

EXECUTIVE COMMITTEE

CHARLES T. HORNGRENWILTON T. ANDERSON

R.LEEBRUMMET

DONT.DcCOSTER

THOMAS R.DYCKMAN

ROBERT L. GRINAKER

VII

ALFRED RAPPAPORT

ROSS SKINNER

DAVID SOLOMONS

FLOYD W.WINDAL

Preface

This project was initially commissioned by the Executive Committee of the AmericanAccounting Association in 1973. The charge was to write a statement that would providethe same type of survey and distillation of current thinking on accounting theory as AStatement of Basic Accounting Theory' (ASOBAT) provided in an earlier decade. Thisdocument represents our response to that assignment.

A periodic reappraisal was deemed necessary since accounting is an important andconceptually rich area and has naturally attracted a wide range of thoughtful and talented researchers. The past ten years have been unusually active in this regard. Fundamental changes have occurred since the publication of ASOBAT. The basic disciplinestraditionally utilized by accounting theory have been altered considerably, and accounting researchers have enthusiastically employed their new tools, perspectives, and analytical techniques toexplore a wide range of accounting issues from new directions.

In our view, these changes are exciting and indicative of intellectual growth. Yet, theyare also frustrating since the new knowledge gathered during the past decade amplytestifies to the view that there are no easy theoretical answers to many of the urgentproblems faced by the profession.

For this reason, ours is not a statement of accounting theory; instead we have prepared a statement about accounting theory and theory acceptance. This distinction is anaturalconsequenceof the recent advances in knowledge. Just as a synthesis of leadingthought ten years ago led to an ASOBAT-type document, an equally conscientious synthesis of today's research findings generates this current statement. While different interms of scope, mode of analysis, and conclusions, this document—like ASOBAT—is intended to be a faithful rendition of the frontiers of accounting theory at the time it is written.

The committee has worked on this document for two years. We have met as an entiregroup eight times and have had several subcommittee meetings. Numerous drafts havebeen written and rewritten, edited, and re-edited. It is our hope that the final product notonlysynthesizes the predominant theory perspectives in the contemporary literature butalso serves as an educational document for students of accounting thought.

This report benefited from the editorial suggestions of Lanny G. Chasteen and JeanEros. Much of the groundwork for this document was developed by a predecessor committee in 1973-74 under the chairmanship of Kermit Larson. In addition to most of themembers of the current group, that committee also included Enrico Petri, Gary L. Sun-dem and Jan R. Williams. Their contributions, as well as the excellent guidance providedby Kermit Larson, are gratefully acknowledged.

Lawrence Revsine, ChairmanCommittee on Concepts andStandards for External Financial Reports

* Committee to Prepare A Statement of Basic Accounting Theory. A Statement of Basic Accounting TheoryIAmerican Accounting Association. 1966>

IX

CommitteeonConcepts andStandards forExternal Financial Reports

Statement on Accounting TheoryandTheory Acceptance

Members: James R. BoatsmanJoel Demski

JohnW.KennellyKermit D.LarsonLawrence Revsine,

ChairmanGeorge J. StaubusRobert R. SterlingJerryJ.WeygandtStephen A. Zeff

Oklahoma State UniversityStanford UniversityDePaul UniversityUniversity of Texas, Austin

Northwestern UniversityUniversity ofCalifornia, BerkeleyRice UniversityUniversity of Wisconsin, MadisonTulane University

The views expressed in this report are those ofa majority ofthe committeemembers. Publication in this volume does not imply agreement orofficial endorsement on the partof theAmerican Accounting Association or its Executive Committee.

Chapter 1

Introduction

There has been a persistent, widely heldbelief among accountants that the accumulation of accounting theory literaturewould eventually lead to a compellingbasis for specifying the content of externalfinancial reports. Indeed, a primary objective implicit in the charge to this committee was to ascertain the extent to which existing accounting theories do in fact provide a basis for determining the content ofexternal financial reports and resolvingaccounting controversies.

If a compelling conceptual basis couldbe discerned, the committee would thenidentify the basic concepts that shouldunderlie the design of external reports. Alternatively, if the theoretical literature ofaccounting has not provided a sufficientbasis for guiding the design of external reports, the committee would then assess theplausible reasons for this state of affairs.

In the view of this committee, a singleuniversally accepted basic accountingtheory does not exist at this time. Instead,a multiplicity of theories has been—andcontinues to be—proposed. Therefore, thisstatement cannot provide accounting withan unequivocally acceptable conceptualsuperstructure when the underlying foundation has not yet settled. Accordingly, wehave defined our task to be somewhat different from past theory statements. Hereinwe seek to explain why the accountingcommunity has been unable to achieve theoretical closure. That is, we attempt to explore the problem that characterizes theoretical debate at this stage of accountingdevelopment: virtually endless argumentation and inability to resolve issues thatare raised. Once the nature of the problemis better and more widely understood, webelieve that unrealistic expectations regarding "authoritative theory pronouncements" will be reduced. While we most emphatically do not call for a moratorium ontheory building and conceptual modeling,we do call for a better understanding of thereasons why such efforts in the past seemto have persuaded only a small proportionof the intended audience.

We believe that theory acceptancewould not be facilitated by this committee's attempting to impose theory closure.Numerous theories already exist. Selectionof one such theory by this committee mightbe persuasive to some. However, becauseof the complex nature of the process oftheory acceptance (which is discussed inChapter 4), any imposed selection of onefrom among competing theories would notresolve existing debates and would not provide rigorously defensible theoretical closure. Accordingly, this report does not attempt to develop a statement of universally accepted accounting theory; insteadours is a statement about accountingtheory and theory acceptance. Specifically, we explore certain theoretical approaches to accounting and explain thereasons why achieving consensus is an arduous task.

Basis for Different AccountingTheories

There is currently an abundance of theories of external reporting. At a very general level, accounting writers appear toagree that the central purpose of financialaccounting is the systematic provision ofeconomic data about reporting entities.The data are provided to individuals andgroups external to the reporting entity, andit is generally acknowledged that profit-seeking as well as not-for-profit entitiesmay be involved.

However, when one attempts to applythis doctrine to resolve actual accountingissues, divergent theories arise. Such divergence results from differences in theway people specify both the "users" of accounting data and the "environment" inwhich preparers and users of accountingdata are thought to behave.

In this section we survey alternativecharacterizations of "user" and "environment." The differences in specificationsthat we observe among writers and the variety of analyses induced by these divergences lead us to conclude that:

1. no single governing theory of finan-

Stat

emen

tonA

ccou

ntin

gT

heor

yand

The

oryA

ccep

tanc

e

cial

acco

unti

ngis

rich

enou

ghto

enco

mpa

ssth

efu

llra

ng

eo

fus

er-e

nvi

ronm

ent

spec

ific

atio

nsef

fect

ivel

y;he

nce,

2.th

ere

exis

tsin

the

fina

ncia

lac

coun

tin

gli

tera

ture

nota

theo

ryof

finan

cial

acco

untin

g,bu

ta

colle

ctio

nof

the

orie

sw

hich

can

bear

raye

dov

erth

edi

ffere

nces

inus

er-e

nvir

onm

ent

spe

cif

ica

tio

ns.

Use

rs The

seto

fuse

rsof

finan

cial

acco

untin

gre

port

sso

met

imes

isde

fined

narr

owly

toin

clud

eonl

yth

epre

sent

owne

rsof

are

port

ing

enti

ty;

atot

her

tim

es,

itis

defin

edbr

oadl

yto

incl

ude

cred

itors

,em

ploy

ees,

regu

lato

ryan

dta

xing

auth

oriti

es,

char

ita

ble

inst

itutio

ns,

and

the

"gen

eral

pub

lic." Fu

rthe

r,th

ere

isso

me

ques

tion

abou

tw

hatc

onst

itute

s"u

se"o

facc

ount

ing

data

.It

isap

pare

ntth

atin

divi

dual

sw

hodo

not

actu

ally

rece

ive

the

acco

untin

gda

tain

ques

tion

can

beaf

fect

edby

the

acti

onso

fth

ose

who

dore

ceiv

ean

dac

tupo

nth

ose

data

.(In

sider

tradi

ngis

acl

assic

exam

ple

ofth

isph

enom

enon

.)T

heno

tion

of"u

se"

ofac

coun

ting

data

can

bebr

oade

ned,

ther

efor

e,to

enco

mpa

ssth

ose

affe

cted

indi

rect

lyby

the

appe

aran

cean

dem

ploy

men

toft

heda

taby

thos

ehav

ing

dire

ctac

cess

.In

the

lim

it,t

hen,

the

enti

rem

embe

rsh

ipof

aso

ciet

ym

aybe

rega

rded

asco

nst

itutin

gth

eun

iver

sal

seto

f"us

ers"

;fu

ture

gene

rati

onsm

ight

wel

lbe

incl

uded

.1W

ith

hom

ogen

eous

user

s,or

hete

ro-

gene

ousu

sers

who

happ

ento

agre

eon

the

finan

cial

data

they

wan

t,th

ere

solu

tion

ofac

coun

ting

issu

esis

fair

lyst

raig

htfo

rw

ard.

The

pref

eren

ces

ofa

"rep

rese

nta

tive

"in

divi

dual

wil

lgov

ern,

beca

usea

llin

divi

dual

sar

e—by

assu

mpt

ion—

unan

imou

sin

thei

rpre

fere

nces

for

finan

cial

repo

rting

.B

utw

ith

hete

roge

neity

,th

epr

oble

mbe

com

esqu

ite

com

plex

.E

ven

whe

nth

epo

pula

tion

ofus

ers

issp

ecifi

ed,a

num

ber

ofqu

esti

onsa

bout

the

beha

vior

ofus

ers

rem

ain.

Fore

xam

ple,

tow

hat

exte

ntdo

belie

fsan

dpr

efer

ence

sva

ryac

ross

user

san

dho

w,

ifa

tall

,do

dif

fere

nces

inus

ers'

belie

fsan

dpr

efer

ence

saf

fect

the

dem

and

for

and

use

ofac

coun

tin

gre

port

s?W

hen

face

dw

ith

anac

coun

tin

gre

port,

how

dous

ersr

eact

?Ar

eth

eyin

fluen

ced

byth

e"f

orm

"as

wel

las

"con

tent

"of

the

data

pres

ente

d?A

reth

ey"r

atio

nal"

proc

esso

rsof

the

data

?Do

they

adop

t"n

aive

"in

terp

retiv

epr

oces

ses

or"s

ophi

stic

ated

"pr

oces

ses?

Whe

nfa

ced

with

mul

tiple

sour

ces

ofda

ta,h

owdo

they

com

bine

thev

ario

usda

tain

toa

com

posi

teas

sess

men

t?F

urth

erm

ore,

doan

swer

sto

thes

equ

estio

nsch

ange

whe

nw

elo

okat

user

sin

the

aggr

egat

era

ther

than

indi

vid

ua

lly? To

illu

stra

teth

eim

port

ance

ofth

ese

issu

esto

theo

ryco

nstr

ucti

on,

cons

ider

the

follo

win

g.W

hile

ques

tion

sab

out

whe

rece

rtai

nda

tash

ould

bedi

sclo

sed

wit

hin

are

port

ma

ybe

cruc

ial

ifon

evi

ewof

user

beha

vior

isas

sum

ed,

they

may

beir

rele

vant

ifan

othe

rvi

eww

erea

dopt

ed.F

orex

ampl

e,a

rati

onal

proc

esso

r,un

infl

uenc

edby

form

,wou

ldbe

tota

llyin

diffe

rent

toth

edi

sclo

sure

ofda

tain

foot

note

sra

ther

than

inth

est

atem

ent

prop

er.

Sim

ilarly

,qu

estio

nsof

whe

ther

finan

cial

repo

rts

shou

ldju

stpo

rtra

yev

ents

,or

port

ray

and

inte

rpr

etev

ents

,bec

omei

nter

esti

ngor

unin

ter

esti

ngde

pend

ing

onon

e's

view

ofus

erbe

ha

vio

r.

Ina

"gen

eral

"th

eory

,the

spec

ifica

tion

ofus

erss

houl

dbe

able

toco

pew

ith

ava

-

'How

ever

,thi

sobs

erva

tion

does

noti

mpl

yth

atfo

cusin

gup

ona

subs

et,e

ven

aqu

itena

rrow

subs

et,o

fth

isbr

oadg

roup

isun

desir

able

pers

e.On

thec

ontra

ry,

give

nth

epu

rpos

eof

thea

naly

sisan

dth

eaud

ienc

eto

which

itis

addr

esse

d,su

cha

restr

ictio

nm

aybe

eith

erth

eore

tical

lyso

undo

rpra

gmat

ical

lydi

ctat

ed.

Wec

ansa

ytha

tthe

focu

s(o

na

subj

ect)

is"t

heor

eti

cally

soun

d"if

itca

nbe

show

ntob

ebas

edup

on:f

irsta

deco

mpo

sitio

nof

the

seto

fall

user

sin

tosu

bset

s;an

dse

cond

,an

anal

ytic

alse

para

tion

ofth

esub

sett

obe

con

sider

edfro

mth

eoth

ersu

bset

s.Fo

rexa

mpl

e,ca

nwef

ocu

sup

on"n

aive

"us

ersw

ithou

tals

ogi

ving

dire

ctat

tentio

nto

thei

mpa

ct(s

)tha

tthe

yha

veup

on"s

ophi

stica

ted"

user

s,an

d/or

thos

eim

pact

swhi

chth

eact

ions

ofso

phist

icat

edus

ersh

ave

upon

them

?Th

epo

ssib

ility

ofin

tera

ctiv

eef

fect

can

serio

usly

wea

ken,

ifno

tdes

troy

the

effi

cacy

offo

cusi

ngon

asu

bset

.

Eve

nw

here

theo

retic

also

undn

ess

wou

ldbe

wea

ken

edby

focu

sing

ona

subs

et,th

efoc

usm

aybe

"pra

gm

atic

ally

just

ified

"w

here

the

loss

esin

"the

oret

ical

soun

dnes

s"ar

emor

etha

nof

fset

byth

egai

nsin

(rat

abi

lity

inth

eana

lysi

s(i

nth

eju

dgm

ento

ftho

sefo

rmu

latin

gan

dtho

sein

tend

ing

toem

ploy

thet

heor

y).

Intr

od

ucti

on

riet

yof

indi

vidu

alan

dm

ulti

-per

son

conf

igu

ratio

ns.

Such

ast

rong

requ

irem

ent

impl

ies

that

ifa

"gen

eral

"th

eory

isto

exis

tan

dsu

rviv

eit

shou

lda

cco

mm

od

ate

wha

tis

know

nor

beli

eved

abou

tth

eva

riet

yof

user

s.In

deed

,suc

ha

theo

rysh

ould

beab

leto

cope

wit

han

dsa

yso

met

hing

abou

tthe

desi

rabi

lity

ofa

vari

ety

ofin

stit

utio

nal

arra

ngem

ents

asw

ell.

Seve

ral

crit

ical

ques

tion

sar

eem

piri

cal

and

larg

ely

unex

plor

ed.A

sa

cons

eque

nce,

seve

rala

ccou

ntin

gth

eori

es—

each

inco

rpor

atin

ghi

ghly

spec

ific

view

sof

user

beha

vior

—co

exis

t.A

nd

the

view

sof

use

rbe

havi

ora

reus

uall

yin

con

sist

ent

acr

oss

the

theo

ries

tha

tem

ploy

this

appr

oach

.

En

vir

on

men

tA

sim

ila

rse

tof

ob

serv

ati

on

sm

ay

bem

ad

ew

hen

we

exa

min

eth

een

viro

nm

enti

nw

hich

user

san

dpr

epar

ers

ofac

coun

ting

data

are

thou

ghtt

obe

have

.O

nepr

omin

ent

envi

ronm

enta

lis

sue

con

cern

sth

eex

iste

nce

ofco

mp

etin

gso

urc

eso

fin

form

ati

on

ab

ou

tth

een

tity

inqu

esti

on.

Som

eth

eori

esdo

not

expl

icit

lyco

nsi

der

alt

ern

ati

veso

urc

esof

fina

ncia

lin

form

atio

n;ot

hers

acti

vely

exam

ine

acco

unti

ngin

form

atio

nas

one

ofse

vera

lso

urc

esof

fina

ncia

lin

form

ati

on

.S

imil

ar

ly,

inex

amin

ing

acco

unti

ngin

form

atio

n,so

me

theo

ries

focu

son

the

prep

arer

sof

such

info

rma

tio

nw

hil

eo

ther

str

eat

them

pass

ivel

y.In

am

ulti

ple-

sour

cese

ttin

g,co

nsid

era

tions

ofef

ficie

ncy

wou

ldim

ply

that

acco

unti

ngsy

stem

ssh

ould

beor

ient

edto

repo

rtin

gth

ose

attr

ibut

esof

the

enti

tyfo

rw

hich

they

have

aco

mpa

rati

vead

vant

age

ove

rot

her

sour

ces,

eith

eras

ap

rim

ary

sou

rce

of

"new

s"or

asan

inst

rum

ent

for

conf

irm

ing

(or

disc

onfi

rmin

g)"n

ews"

repo

rted

bym

ore

tim

ely

sour

ces.

An

oth

er

en

vir

on

men

tal

issu

eco

nce

rns

the

natu

reof

the

mul

ti-p

erso

nse

ttin

gfo

rac

tion

sby

user

s.B

oth

"ma

rket

"in

tera

cti

ons

(exc

hang

es)

and

"non

mar

ket"

inte

rac

tion

s(e

xter

nali

ties

)ap

pear

tobe

pre

sent

inre

lati

onsh

ips

amon

gpr

epar

ers

and

user

s.W

ith

resp

ectt

om

ark

etin

tera

ctio

ns,

such

ques

tion

sas

the

follo

win

gar

eas

ked:

To

wh

at

exte

nt

are

secu

riti

esm

ark

ets

"ef

fici

ent"

inre

flec

ting

avai

labl

ein

form

ati

on

?W

ha

tis

the

na

ture

ofth

em

ark

etfo

r

info

rmat

ion?

Inpa

rtic

ular

,do

we

obse

rve

ma

rket

fail

ure

tosu

cha

degr

eeth

at

inte

rve

ntio

nis

soci

ally

desi

rabl

e?2

Wit

hre

spec

tto

"ext

erna

liti

es,"

tow

hat

exte

nt

are

no

nm

ark

etin

tera

ctio

ns

am

on

gus

ers

pres

ent?

To

wha

tex

tent

are

they

pert

inen

t?E

xter

nali

ties

may

also

exis

tam

ong

prep

arer

s.In

form

atio

nco

nvey

edby

one

sour

cem

ayaf

fect

that

from

anot

her.

Sim

ilar

ly,

repo

rtin

gfo

rin

com

eta

xpu

rpos

es,f

orex

ampl

e,ca

nin

tera

ctw

ith

fina

ncia

lrep

orti

ng.A

lso,

ther

eis

anin

tera

cti

onbe

twee

nm

anag

ers'

deci

sion

sand

exte

rnal

fina

ncia

lre

port

s.A

thir

dcl

ass

of

issu

esd

eals

wit

ha

ttri

bute

sof

acco

unti

ngpr

oces

ses,

and

ofth

eco

ntro

lofs

uch

proc

esse

s.W

esh

ould

reco

gni

ze,

for

exam

ple,

that

audi

ting

isan

acce

pted

com

pone

ntof

the

exte

rnal

repo

rtin

gen

viro

nmen

tan

d,by

impl

icat

ion,

tha

tre

port

ing

conc

erns

cann

otbe

sepa

rate

dfr

omau

diti

ngco

ncer

ns.

We

mus

tal

sose

ekto

unde

rsta

ndth

em

an

yen

viro

nmen

t-us

erin

flue

nces

that

have

anim

pact

upon

audi

ting

ifw

ear

esu

bseq

uent

lyto

unde

rsta

ndth

ena

ture

ofau

diti

ng's

infl

uenc

eon

repo

rtin

g.

Su

mm

ary

Itis

easy

toob

tain

agre

emen

ton

the

broa

dro

leof

fina

ncia

lac

coun

ting

,a

sw

ell

ason

the

fact

tha

tac

coun

ting

syst

em

sa

re"c

ostl

y,"

and

henc

eth

atth

eir

desi

gnpr

ese

nts

acl

ass

ofre

sour

ceal

loca

tion

prob

lem

s.Su

chag

reem

ent,

how

ever

,ove

rloo

ksth

eex

iste

nce

of

ba

sic

dif

fere

nce

sin

the

way

vari

ous

theo

ries

view

user

san

dth

epr

epar

er-u

sere

nvir

onm

ents

.One

'ssp

ecif

ica

tio

ns

ofu

sers

an

den

viro

nm

ent

are

inte

gral

toth

em

odel

ing

ofac

coun

ting

choi

ces

and

wil

lsig

nifi

cant

lysh

ape

the

resu

ltin

gth

eori

es.

Wha

tw

ese

ekis

ath

eory

that

isge

nera

len

ough

toco

pew

ith

this

vari

ety

and

spec

ifi

cen

ough

toof

fer

assi

stan

ceto

acco

unti

ngpo

licy

mak

ers.

But

none

ofth

eav

aila

ble

theo

ries

isac

cept

able

toal

lac

coun

tant

s.

2A

ma

rket

fail

ure

issa

idto

exis

tw

hen

the

ma

rket

-de

term

ined

prod

uctio

nan

dco

nsum

ptio

nle

vels

gene

ra

tean

inef

fici

ent

allo

cati

onof

reso

urce

s.S

eeB

eave

ran

dD

emsk

i(1

974)

and

Gon

edes

,Dop

uch.

and

Pen

man

(197

5)fo

rfu

rthe

rdi

scus

sion

ofm

ark

etfa

ilur

eco

nsid

era

tio

ns.

4 Statement on Accounting Theory and Theory AcceptanceInstead, we have a collection of theories, pects of this specification problem buteach perhaps capable of coping with as- only by sacrificing attention toothers.'

Chapter 2

Alternative Theory Approaches

In this chapter we classify accountingtheories and summarize the dominant approaches that have evolved within each ofthe classifications. We have discernedthree basic theoretical approaches: (1)classical ("true income" and inductive)models; (2) decision usefulness; and (3)information economics.1 This survey is intended to provide the foundation necessaryfor our subsequent examination of thereasons that the profession has beenunable to achieve consensus on a"general" theory of external reporting.

Classical Approaches toTheory Development

"The history [of accounting thought)."writes Chambers, "is not a history of development, but a series of disconnectedepisodes" (1965, p. 33). The authors ofearly accounting theories failed to supply adevelopmental thread from one argumentto the next. Few referred to the works ofothers, and none built explicitly on thewritings of predecessors or contemporaries in the accounting literature.

At the outset, it will be useful to classifya number of major writers discussed inthis section according lo their roles withinthe literature. We begin by chronologicallylisting their principal writings:

William A. Paton, Accounting Theory —With Special Reference to the Corporate Enterprise (1922).

Henry Rand Hatfield, Accounting —ItsPrinciples and Problems (1927).

John B. Canning, The Economics of Accountancy (1929).

I We should observe (hat some authors who havehelped develop these approaches may disavow any intention of having contributed to an accounting theory.Nevertheless, each of these approaches has variouslybeen perceived as a potential source of a theory ofaccounting or as a basis for determining the conlent ofexternal financial reporting. The widespread acceptanceof these perceptions leads the committee, tentatively atleast, lo recognize each as an approach to theorizing inaccounting.

Henry W. Sweenev. Stabilired Accounting (193G).

Stephen Oilman, Accounting Conceptsof Profit (1939).

Kenneth MacNeal. Truth in Accounting(1939).

W. A. Paton and A. C. Littleton, An Introduction to Corporate AccountingStandards (1940).

Sidney S. Alexander. "Income Measurement in A Dynamic Economv"(1950).

A. C. Littleton, Structure of AccountingTheory (1953).

Edgar 0. Edwards and Philip W. Bell.The Theory and Measurement ofBusiness Income (1961).

Maurice Moonitz, The Basic PostulatesofAccounting (1961).

Robert T. Sprouse and Maurice Moonitz.A Tentative Set of Broad AccountingPrinciples for Business Enterprises(1962).

Yuji Ijiri, Theory of Accounting Measurement (1975).

Writers such as Paton, Sweeney, MacNeal, Edwards and Bell, Moonitz, andSprouse are advocates: they have arguedfor the primacy of new theories or approaches. They have been, in every sense,reformers. Canning and Alexander, bycontrast, are more analysts and expli-cators than advocates. They analyze andassess what accountants do and seek to do,they undertake to explain an economicmodel lo accountants, and they concludeby endeavoring to adapt the economicmodel to the pragmatic world of the accountant. All the writers in this first groupare normative theorists. They are primarily deductivists. Canning and Alexander differ from the rest of the group more in degree than in kind; they are motivated lessby missionary zeal than by a desire toanalyze, criticize, and suggest.

Hatfield, Gilmaii, Littleton, the Paton-Littleton partnership, and Ijiri are chieflypositivist, inductive writers, some of whom

Stat

emen

ton

Acc

ount

ing

The

orya

ndT

heor

yAcc

epta

nce

inte

rpos

eoc

casi

onal

norm

ativ

ede

duct

ive

reas

onin

gder

ived

from

unst

ated

prem

ises

.O

fthe

five

,Gil

man

and

Hat

fiel

din

duce

ata

mor

eco

ncre

tele

vel

ofge

nera

liza

tion

than

the

othe

rth

ree;

they

sum

mar

ize,

com

pare

,an

dex

pose

elem

ents

ofill

ogic

and

inco

nsis

tenc

y;th

eyse

ekto

rati

onal

ize

dive

rse

theo

ries

and

prac

tices

.L

ittle

ton,

Pato

n-L

ittle

ton,

and

Ijir

i,co

ntra

riw

ise,

unde

rtak

eto

deve

lop

cohe

rent

theo

ries

ofex

tan

tp

ract

ice.

Wha

tpr

oble

ms

doth

ese

auth

ors

expl

ore?

Tot

hein

duct

ivis

ts,w

hen

they

dono

tin

terp

ose

norm

ativ

ede

duct

ive

reas

onin

g,th

epr

oble

ms

are

perf

orce

thos

eth

atar

eim

plic

itly

bein

gad

dres

sed

inac

tual

acco

untin

gpr

actic

e.In

the

proc

ess

ofin

duc

tion,

how

ever

,th

eau

thor

sm

aydi

sagr

eeov

erge

nera

lized

prin

cipl

esan

dgo

als.

To

the

norm

ativ

ede

duct

ivis

ts,

the

answ

eris

less

evid

ent.

Inth

ebr

oade

stte

rms,

both

dedu

ctiv

ists

and

indu

ctiv

ists

are

con

cern

edw

ithde

sign

ing

fina

ncia

lre

port

sth

atco

mm

unic

ate

pert

inen

tin

form

atio

nto

ase

to

fre

ad

ers

.W

hydi

dm

any

ofth

epr

omin

ent

theo

rist

sad

voca

tecu

rren

tco

sts

or

valu

es?

To

asi

gnif

ican

tdeg

ree,

they

may

have

been

infl

uenc

edby

the

neoc

lass

ical

econ

omic

theo

ryof

the

firm

,in

whi

chhi

stor

ical

cost

sar

eig

nore

den

tirel

y.T

hey

may

have

obse

rved

the

beha

vior

ofin

vest

ors

and

othe

rec

onom

icde

cisi

onm

aker

san

dco

nclu

ded

with

ava

lidat

edhy

poth

esis

that

such

deci

sion

mak

ers

seek

curr

ent

valu

e,no

thi

storic

alco

st,in

form

atio

n.No

findi

ngs

from

any

form

alhy

poth

esis

test

ing

are

repo

rted

;pr

ior

toth

e19

60s,

itw

asno

tcu

stom

ary

for

acco

untin

gw

rite

rsto

con

duct

such

rese

arch

.U

nlik

eth

elat

ergr

oup

ofno

rmat

ive

wri

ters

who

embr

aced

ade

cisio

n-us

eful

ness

appr

oach

,th

eea

rlyno

rmat

ive

wri

ters

did

noti

nqui

rede

eply

into

the

deci

sion

alfr

amew

ork

ofus

ers.

They

posit

edth

eex

iste

nce

ofsp

ecifi

edus

ers,

and

proc

eede

dto

argu

eth

atcu

rren

tva

lues

are

asu

peri

orty

peof

info

rmat

ion

tohi

stor

ical

cost

s.A

sw

ew

illno

tebe

low

,pr

agm

atic

cons

ider

atio

nsob

liged

seve

ral

ofth

ew

rite

rsto

soft

enth

eir

reco

mm

en

dati

on

s.

No

rma

tiv

eD

edu

ctiv

eS

cho

ol

Inth

efi

rstd

ecad

esof

the

cent

ury,

mos

t

acco

untin

gw

rite

rsde

alt

with

the

par

ticu

lars

ofac

coun

ting

prac

tice

rath

erth

anw

ithco

ncep

tual

unde

rpin

ning

s.A

ccou

ntin

gth

eory

,es

peci

ally

inth

em

ajor

ityof

text

book

s,co

nsis

ted

ofsp

ecia

lple

adin

gs,

with

out

anev

iden

ttr

ail

oflo

gic,

for

orag

ains

tspe

cifi

cac

coun

ting

prac

tice

s.D

octr

ine

was

defe

nded

agai

nsth

eret

ics.

The

deve

lopm

ent

ofa

cohe

rent

acco

untin

gth

eory

may

have

been

thou

ghtb

yp

ract

itio

ner

sto

bea

task

mo

resu

ited

toth

eac

adem

icm

ind.

Infa

ct,

acad

emic

acco

un

tan

tsp

rio

rto

the

1950

swer

ea

dich

oto-

mou

slo

t.M

ost

acco

untin

gpr

ofes

sors

inA

mer

ican

univ

ersi

ties

wer

efa

rm

ore

orie

nted

tow

ards

prac

tice

than

tow

ards

acad

emic

rese

arch

.M

any

wer

efo

rmer

prac

titio

ners

orw

ere

cont

empo

rane

ousl

yen

gage

din

acco

unti

ngpr

acti

ce.

Aco

mp

arat

ivel

ysm

all

nu

mb

erof

acad

emic

sp

oss

esse

dd

oct

ora

tes.

Onl

ya

rela

tive

lyfe

wun

iver

siti

esaw

ard

edd

oct

ora

tes

inb

usi

nes

sad

min

istr

atio

no

rco

mm

erce

bef

ore

the

1950

s—

and

mos

toft

hose

appe

arto

have

requ

ired

ast

rong

dose

ofec

onom

icth

eory

.T

hus,

man

yof

the

earl

ydo

ctor

ates

awar

ded

toac

cou

nti

ng

acad

emic

s,to

Hat

fiel

dan

dP

aton

,fo

rex

ampl

e,w

ere

inec

onom

ics.

Itsh

ould

com

eas

nosu

rpri

se,

ther

efor

e,th

atec

onom

icth

eory

wou

ldin

flue

nce

toa

con

side

rabl

edeg

reet

hede

duct

ive

writ

ings

ofth

ose

com

para

tive

lyfe

wac

adem

ics

who

chal

leng

eddo

ctri

new

ithco

here

ntac

coun

tin

gth

eori

es.

Inde

ed,

seve

ral

ofth

em

ost

prom

inen

tw

rite

rs,

nota

bly

Can

ning

and

Ale

xan

der

,wer

eec

on

om

ists

.T

hede

duct

ive

theo

rist

s,w

ithth

eex

cept

ion

ofA

lexa

nder

(sec

App

endi

x),c

oncl

uded

that

inco

me

mea

sure

dus

ing

asi

ngle

valu

atio

nba

sew

ould

idea

llym

eet

the

need

sof

allu

sers

.In

that

rega

rd,t

hey

have

som

etim

esbe

ende

scri

bed

asad

voca

tes

ofa

"tru

ein

com

e"th

eory

.(A

sum

mar

yan

alys

isof

thei

rre

spec

tive

posi

tions

isgi

ven

inE

xhib

itI.

)T

hey

drew

onth

ete

achi

ngs

ofne

ocla

ssic

alec

onom

icth

eory

and

onth

eir

obse

rvat

ions

ofec

onom

icbe

havi

orto

prop

ose

that

acco

untin

g,w

hich

had

been

preo

ccup

ied

with

hist

oric

alre

cord

-kee

ping

and

con

serv

ativ

ere

ckon

ings

,sh

ould

bere

cons

truc

ted

tore

flec

tcu

rren

tco

sts

orva

lues

.T

hey

borr

owed

from

econ

omic

sth

ete

rms

"in

com

e"an

d

Alt

ern

ativ

eT

heo

ryA

pp

roac

hes

EX

HIB

ITI

nm

ttry

An

oty

taof

the

Ap

p'o

ach

eto

flh

tD

edu

ctiv

eT

fceo

ruU

Sa

me

fn/r

f»rd

t/ie

rtif

Pat

onT

opr

omot

eef

fici

ent

man

agem

ent,

H9

Ci

whi

chfu

rthe

rsth

ein

trre

sts

of.il

lfr

uit

yho

lder

s,al

soat

are

port

onen

terp

rise

pro

grr

sato

equ

ity

ho

lders

(ban

nin

g

Sw

rtT

K-y

•t£

tt'

MarM

rjI

Ale

xa

nd

er

US

TJO

'

Kd

ward

vB

ell

*The

prop

riet

orjr

nlth

u-a-

larn

rfir

ul

lyin

lcre

-vtr

dm

pro

prt

rto

fiii

pw

ish

chie

fly

Inkn

oww

hat

net

r*tj

i>t+

tir

tr*

M"*

»!•

•«-

in*»

fwn«

rtfi

tlu

f#fi

na

lii

*i(

"i#

h.n

«"

miu

fii-

il*

,fh

.".1

)«

jl

fc»

f«J«

—>

•(Ih

rrn

lorp

'llr

U>

I'vll

iri

*pp

l(,'

M?

H'

All

use

rs,

but

pri

mari

lyl«

u»in

»-*s

man

ag

em

en

t

"T

oin

form

the

ow

ner*

ofa

hu

sm

rs*

ofal

lth

ep

rofi

tsan

dto

w-i

inw

hic

hth

ryh

atr

ann

pji

ly>

pT

Ji*.

oth

erp

art

ir.

te*p

man

ager

*an

dcre

dit

ors

1at

inte

rest

als

oh

av

ea

rig

ht

loth

esa

me

info

rm.*

turn

•pp

ICO

lfS»

Assert

sd

iffe

rm

lm

cu

mr*

Ua

dif

fere

iil

purp

ose*

wh

ere

n-un

otny

isch

ara

cte

rized

byclu

nk

ing

pn

tvn

and

chan

gin

gei

pec

tatK

imof

fulu

reearH

int

po

wer

To

faci

lita

tem

anag

emt-

nipl

anni

ngan

dto

assi

stse

curi

tyan

aly

sts,

ow

ners

ofb

usi

ness

firm

*an

d

pti

trn

lial

entr

epre

neu

r*in

mak

ing

rati

onal

com

pari

son.

*am

ong

com

pan

ies

and

ind

ust

ries

tnf<

rt*

dA

lod

Wu

nd

erld

«o

lC

ecu

mil

un

ra>

i

"In

com

ein

the

bro

ades

tse

nse

may

be

co

nceiv

ed

as

incl

ud

ing

(he

en

tire

net

incr

ease

inth

e(t

rue

econ

omic

posi

tion

ofa

busi

ness

)aft

er

du

eall

ow

an

ce

has

bre

nm

ad

efo

rrw

*win

vest

men

t*an

dw

ith

dra

wab

'ip

p4*

0.U

A>

Mea

sure

the

ann

ual

chan

ge

inca

pit

alv

alu

el»

>re

fere

nce

loth

ed

:rert

*a

1ua

1wn

ofIh

ea

i^ets

Measu

rrctu

ng

e*

intr

iere

al

valu

atio

nof

cap

ital

byrH

erm

ce

toch

ang

esin

itsfu

ture

pro

du

ctiv

ity

toth

em

arg

inal

use

r

Mra

ture

ch

an

ges

in"eco

no

mic

valu

e,'

defi

ned

a*Ih

em

ark

et

pru

r*ir

fthe

firm

'sa*

sels

ina

free

,m

mp

rtil

ive.

hrn

.id

.an

dac

tiv

ern

arh

rt

Mea

sure

thr

cap

ital

ized

val

ue

ofth

een

terp

rim

tand

ctu

ng

esth

erei

n

Mea

sure

Ihe

Mth

jrrt

ne

valw

and

subj

ecti

vepr

ofit

oflh

een

lerp

ri*

r

Mn

um

U1

Sp

rou

seM

oo

nit

z

•IS

O

To

faci

lita

tem

anag

emen

tp

lan

nin

gan

dco

ntr

ol,

an

dto

aid

ow

ners

cre

dit

ors

,an

dg

ov

ern

men

tin

esah

uil

inr,

man

agem

ent

per

form

Mea

sure

the

chan

ges

inen

terp

rise

wra

ith

evid

entl

yb

ein

gth

rp

rese

nt

valu

eof

futu

recash

flo

ws

Krcu

mm

rtd

rd

Mft

isu

rrn

trn

t\t

Vt*

i«i*

Incl

ude

appr

ecia

tion

ofm

ark

etab

lesecu

riti

es

an

dsta

nd

ard

raw

mate

rials

inn

on

-op

erat

ing

inco

me,

tore

co

gn

ize

ap

pre

cia

tio

no

no

ther

inv

en

tori

es

wo

uld

he

mo

redu

biou

s,ap

pre

ciat

ion

onfi

xed

asse

tsan

dth

eco

nse

qu

ent

dep

reci

atto

non

appr

ecia

tion

mig

htbe

dis

pla

yed

ina

sup

ple

men

tary

sta

tem

en

t

Mea

sure

asse

tsan

dli

abil

itie

sby

disc

ount

ing

futu

reca

bh(l

ows,

iffe

asib

le,

ifn

ot.

reso

rtto

ind

irec

tv

alu

ati

on

sis

urh

as

rtrJ

'In

co

me

isth

ech

an

ge

inn

etassrt

s

Acc

ou

nt

for

ch

an

ges

inre

pla

cem

en

tco

at

'wh

ich

are

den

om

inate

das

un

reali

zed

un

til

the

assets

are

ci

chan

ged

1.al

sou

>eO

PI.

chan

ge*

L*e

mar

ket

pri

refo

r"m

ark

etab

leas

sets

."ap

pra

isal

sor

rep

lace

men

!co

stfo

r"r

epro

du

cib

le,

non

mark

eta

ble

ass

ets

."an

do

rig

inal

cost

less

am

ort

izati

on

ord

eple

tio

nfo

r"r

ton

rep

rod

uci

ble

nun

mar

kefa

blc

assets

"W

ou

ldin

clu

de-

un

real

ized

hold

ing

gai

ns

and

loss

eson

mer

chan

dis

ein

ven

tory

inne

tin

co

me,

oth

er

un

reali

zed

item

s,

wh

ile

dis

clo

sed

inIh

ein

co

me

sta

lem

enl.

are

tran

sfer

red

toC

apit

alS

urpl

us

Pro

po

ses

var

iou

sin

ean

ure

*d

epen

din

gon

use

ran

du

seIs

skep

tica

lof

Ihe

usefu

lness

of<

;PI,

acco

un

tin

g

Acc

ount

for

chan

ges

inre

pla

cem

ent

co

st.

dis

tin

gu

ish

ing

lielw

een

•!<

the

ex

cels

olreali

ze

rev

en

ue

ov

er

the

curr

ent

rep

lace

men

tco

stw

(no

nm

on

etar

yas

sets

con

sum

ed,

an

d<2

>th

eu

nre

aliz

edch

an

ce*

inth

ere

pla

cem

ent

co-»

tof

no

rm

onet

ary

asse

tsT

hr

gra

nd

tota

lis

call

ed"b

usi

nes

sp

ntf

it"

Als

ou

seG

PL

ch

an

ges

Use

dis

cou

nte

dp

rese

nt

val

ue

•at

his

tori

cal

inte

rest

rate

s-

for

receiv

a

bles

And

pay

able

sto

hese

ttle

din

cash

,n

etre

ali

zab

lev

alu

esfo

rre

adil

ysa

lab

lein

vent

orie

s,an

dre

pla

cem

ent

cost

lor

oth

erin

ven

tori

esan

dfo

rta

ng

ible

li»

edass

ets

Kej

ect

real

izat

ion

asta

ckin

g"a

naly

tica

lpr

ecis

ion

Als

ofa

vor

CP

Lch

an

ges

"wea

lth

"an

dso

ught

tom

ake

them

oper

atio

nal

inan

acco

unti

ngco

ntex

t.In

the

act

ofre

nder

ing

them

oper

atio

nal,

they

disa

gree

don

par

ticu

lars

.T

hei

rdi

ffer

ence

sov

erho

wto

redu

ceab

stra

ctco

ncep

tsto

pra

gm

atic

reco

mm

enda

tion

spe

rhap

sre

flecte

dth

eir

co

nclu

sio

ns

ab

ou

tu

sers

'eco

nom

icbe

havi

or.D

oubt

less

they

had

diff

eren

tpe

rcep

tion

sof

the

acce

ssib

ilit

yof

ac

cept

able

dat

a.P

ato

n,

Mac

Nea

l,an

dS

prou

sean

dM

ooni

tzap

peal

edto

the

mar

ket

asth

ep

erti

nen

tso

urc

eof

info

rmat

ion

.

8 Statement on Accounting Theory and Theory Acceptance

Canning, Sweeney, Alexander, Edwardsand Bell, and Moonitz were more influenced by the vision of capital as the presentvalue of future incomes. In theory, however, asset prices correspond with capitalized future incomes only under very restrictive conditions. Thus, the differencesin the specific policy recommendations ofPaton, Canning, Sweeney, MacNeal, Alexander, Edwards and Bell, Moonitz, andSprouse and Moonitz can be traced to therelaxations in conceptual rigor which theybelieve to be warranted by the presence ofuncertainty.

Several of the writers cite particularusers (ordinarily shareholders, creditors,and managers) and occasionally suggestthe information that users would finduseful; however, it is not possible to employ these casual references to explain thesurrogate choices that the writers made.Typically, the pragmatic modifications ina model are matter-of-factly asserted bythe writer. Except in the case of Alexander, who proposes different models for different users, each writer offers his policyrecommendation as a universally validproposal, as if the entire hierarchy of userswould be sufficiently well served by theresulting information. Nothing was saidabout the possible repercussions on resource allocation decisions made elsewhere in society as a consequence of decisions taken on the basis of information generated by a model.

What factors did the writers regard asinfluential in shaping their policy recommendations? Paton — and perhaps Sweeney — opted for replacement cost in preference to net realizable value in order not torecognize gross margin prior to realization. In that respect, Paton and Sweeneyare more conservative than is generallybelieved. Both attached considerable importance to the distinction between realized and unrealized outcomes. Sweeneywas under pressure from his dissertationadviser to be "practical," and his book stabilized Accounting was a product of the dissertation (see Appendix).

Canning had a dual mission: to lend advice to the profession as an economist, andto interpret accountants' handiwork toother economists. He acknowledged that it

would seldom be possible for accountantsto compute the capitalized values of nonmonetary assets. He therefore proposed amethod of indirect valuation by which anasset would be assessed at the current costto reacquire by the least costly means theremaining services inherent in the asset.This approach, Canning believed, constituted a sound application of the economist's opportunity cost analysis, and he devoted the latter part of his book to the practical problems of implementation.

Edwards and Bell imposed upon themselves the constraint that the accountingprofession not be rcvulsed by their policyrecommendation; consequently, like Patonand Sweeney, they proposed a replacementcost system in which gross margin wouldnot be recognized until sale. But unlikePaton and Sweeney, they preferred to include in income the currently accruedholding gain on nonmonetary assets.

MacNeal had been a practicing accountant; he alone among the deductive writersdid not possess a doctorate. To MacNeal,"truth" was ascertainable in current markets ; he shared the practitioner's aversionto relying on management's opinion of thevalues of assets. He recommended replacement costs — and, as a last resort, historical costs — only when acceptable marketselling prices were not accessible (see Appendix).

An interesting comparison may bedrawn among the deductive writers according to whether their orientation wastoward income-statement analysis or comparative balance-sheet analysis. Paton andSweeney, oriented toward income statements, were mindful of the accountant'sreluctance to record or report unrealizedgross margins and unrealized holdinggains. MacNeal, Edwards and Bell,Moonitz, and Sprouse and Moonitz were inclined lo view accounting as a successionof balance sheets. It is therefore not surprising that they would include unrealizedholding gains in net income and, except forEdwards and Bell's preoccupation with thereaction of the accounting profession totheir proposed system, they would havebeen unanimous in their disavowal of theconventional realization concept.

Of the deductive theorists, only Alexan-


der proposed that different models mightbe needed for different users. Yet he onlyintimates how and why different usersmight require different information; hestops well short of constructing decisionmodels.

It may be observed that the deductivewriters operated independently of oneanother, rarely comparing their work withthat of predecessors or contemporaries.The logic of their analyses is difficult tomonitor, as it reflects implicit criteria andjudgments. Of their writings, it may besaid that they neither proved their pointsnor were disproven by others. A commonthread may be discerned in their diverserecommendations: the implicit agreementthat users seek (or should seek) current-price information in making economicdecisions. In this important respect, notwithstanding the diversity of their recommendations, their cause was united.

Inductive SchoolHatfield and Gilman are inductivists in

a special sense. They were annotators ofthe literature. They compared and contrasted different practices and policyrecommendations, drew attention to similar and discordant notes among diversesources of authority, and commented uponillogic and inconsistencies in practice andin the literature. While they did not attempt to formulate coherent theories of extant practice, they did endeavor to identifygaps and (Hatfield especially) to distinguish doctrine from principle.

Littleton, upon observing the evolutionof accounting practice over a considerableperiod of time, concluded that theaccountant endeavors to help the readersof financial reports understand thebusiness enterprise by confining hismeasures to objectively verifiable transactions to which the firm is a party. By alargely implicit argument of a normativevariety, Littleton maintained thataccounting reports can best serve theneeds of those who contemplate future actions by reporting factually on the immediate consequences of actionspreviously taken by the firm. Currentprices and index numbers are eitherirrelevant to the transactional experience

of the enterprise or are not susceptible toobjective measurement; their inclusion infinancial statements would disturb thehomogeneity of the contents and mightwell reduce the integrity of the objectivelydetermined historical results.

Littleton induced accounting principlesfrom observing accounting practice, andargued in a normative deductive vein thatthe goals implicit in this generalized practice should be retained. He seemed to beconcerned that an accounting that dealswith actual transactions could become debauched by integrating in the financialreports subjective judgments about the impact of economic forces on the enterprise.While Littleton is remembered primarilyas a historian and an inductive theorist, hispersistent defense of the generalized principles induced from actual practice partake of normative deductive logic developed from unstated premises. In that respect, he has been characterized as proposing an "accounting Darwinism." According to this view, accounting is continuallyin evolution, and the elements that havewithstood the succession of challengershave, at least thus far, earned a place incurrent practice.

When a predominantly deductive writerof decidedly strong views (Paton) and apredominantly inductive writer possessingequally strong views (Littleton) join in thewriting of a monograph, a compromisemethodology should not be unexpected.The resulting monograph (1940), probablythe most influential work in American accounting literature, was a rationalizationof then extant accounting practice, explicated at a level of theoretical abstraction that had known few precedents.The elegance of the writers' generalizations even deceived a leading practicingaccountant who, in a review of the monograph for a journal, asserted that it bore noperceptible relation to current practice. Indeed, the work never could have attainedthe degree of popularity it has since enjoyed had it not been a largely faithful induction of actual accounting practice, evendespite an occasional oughtness spliced into the exposition. In addition, vivid metaphors employed in the inductive exercise(e.g., "matching," "attaching," "costs are


assembled... as if they had a power of cohesion") conferred a greater respectability on practices that could be so colorfullyand attractively characterized. (A moredetailed analysis of this monograph isgiven in the Appendix to this chapter.)

In comparison with Littleton, Ijiri(1975) shows a clearer separation of deductive and inductive reasoning. Ijiri undertakes to generalize the goals implicit incurrent accounting practice, and thendefends historical cost against thecriticisms of current-cost and currentvalue advocates by a reverse-inductiveanalysis. The defense is embedded in theparticular manner of his induction. Ijiriconcludes that accounting practice maybest be interpreted in terms of accountability, which he defines as economicperformance measurement that is notsusceptible to manipulation by interestedparties. Unequivocal and unambiguousmeasures are viewed as the sine qua non ofaccountability. The continuous recordingof current values is rejected because theyare predicated on hypothetical actions ofthe entity and, as such, are not verifiable.Ijiri explains forthrightly his strategy forpreferring inductive to deductive reasoning:

This type of inductive reasoning toderive goals implicit in the behaviorof an existing system is not intendedtobepro-establishment orto promotethe maintenance of the status quo.The purpose of such an exercise is tohighlight where changes are mostneeded and where they arc feasible.Changes suggested as a result of sucha study have a much better chance ofbeing actually implemented. Goalassumptions in normative models orgoals advocated in policy discussionsare often stated purely on the basis ofone's conviction and preference,rather than on the basis of inductivestudy of the existing system. Thismay perhaps be the most crucialreason why so many normativemodels or policy proposals are notimplemented in the real world, (p. 28)

Other Possible ApproachesIt is sometimes suggested that "postu

lates and principles" should be viewed as afundamentally distinct approach to theorydevelopment. Yet it may represent nothingmore than a different packaging of the deductive methodology and could accommodate either the "true income" or "decisionmodel" approach. "Postulates and principles" could also be adopted as the framework for developing a hierarchy of generalizations in an inductive study.

SummaryTwo approaches to theory development

may be discerned apart from the more recent emergence of decision-usefulness formulations, which are discussed in the following sections of this chapter. These arethe "true income" (an example of the normative deductive school >and the inductivemethodologies. The former attempts to formulate implicit accounting models ofglobal application, while the latter attempts to rationalize, and sometimes evento justify (by the interposition of normativedeductive reasoning), major elements ofextant accounting practice.

The Decision-Usefulness

ApproachAnother approach by which accounting

theory has been generated is based on explicit recognition of the usefulness objective. The decision-usefulness approach is abroad one from which two major branchesarise. In the first, decision models arestressed. Information relevant to a decision model or criterion is isolated and various accounting alternatives are comparedto the data presumably necessary for implementing these decision models. In thesecond branch of the decision-usefulnessapproach, decision makers are the focus ofattention. Their reactions to alternative accounting data are studied as a means forinductively deriving preferred reportingalternatives.

Decision ModelsMost of the earliest research on deci

sion-usefulness implicitly adopted the decision model emphasis although the assumeddecision model was often not specified indetail. This approach first began to appearin the literature in the 1950s. By 1973 it had


achieved both professional recognition andbroad exposure through publication of thereport of the American Institute of Certified Public Accountants (AICPA) StudyGroup on the Objectives of Financial Statements (1973). In that document, alsoknown as the Trueblood Report, it wasstated that "the basic objective of financialstatements is to provide information usefulfor making economic decisions" (p. 13).

Prior to the 1950s,a number of carefullyprepared works on accounting theory didrefer to users of accounting outputs, butthe theoretical structures in those workswere not demonstrably based on the alleged information "needs" of users. Several instances of passing references to usersand the uses of financial information willserve to illustrate that practice. The 1936"Tentative Statement" of the AmericanAccounting Association (AAA) included,but did not build upon, this paragraph:

The most important applicationsof accounting principles lie in thefield of corporate accounting, particularly in the preparation ofpublished reports of profits and financial position. On the interpretation ofsuch reports depend so many vitaldecisions of business and governmentthat they have come to be of greateconomic and social significance, (p.187)

The 1941 AAA statement reiterated thispoint, again without subsequent development (p. 133).

The Sanders, Hatfield, and Mooremonograph (1938) was more explicit inrecognizing the groups who have information "needs" in relation to an entity.The authors' list of functions of accountingincluded:

... preparing from time to timestatements showing all the more important aspects of the capital and income of the business and of the legalequities in them, satisfying therebythe need for information of all theparties in interest, especially of:

(a) the management of the business,

<b> outside groups, such as investors and creditors.

11

(c) government, in such matters astaxation and regulation, (p. 4)

Paton and Littleton (1940) gave user needseven more prominent attention, includingthem in their statement of the purpose ofaccounting:

The purpose of accounting is tofurnish financial data concerning abusiness enterprise, compiled andpresented to meet the needs of management, investors, and the public,(p.l)May (1943, pp. 19-21) went still further

when he listed ten "uses of financial accounts," such as "to determine the legalityof dividends" and "as a basis for price orrate regulation." May followed thatenumeration with a chapter on "The Usesof Accounts and Their Influence on Accounting," although he did not modeldecision processes or use decision modelsand their information requirements as abasis for accounting theory.

Vatter (1947, pp. 7-9) also gave a gooddeal of attention to the uses of accounting— in management, in the various taxingand social control activities of governments, and in the general area of credit extension and investment — but he rejectedthe idea of basing a general theory on anyparticular point of view. Instead, he developed the operations-oriented "fund theory," which ignores all user groups in an effort to be impersonal and unbiased. Thistheory later provided part of the basis forthe "data bank" and "events theory" approaches to accounting.

During the 1950s there was a stronguser-oriented movement in the manageria"accounting literature. That movemenimay have served as the stimulus for theinitial acceptance of the decision-usefulness objective in external reporting at thatime. Early works include a doctoral disscrtation (Staubus, 1954) and Chambersarticle, "Blueprint for a Theory of Accounting," published in 1955, whichstressed that "the basic function of accounting .. . [is] the provision of information to be used in making rational decisions" (p. 25). The July 1955 issue of TheAccounting Review included the Committee on Concepts and Standards' Supplementary Statement No. 8:


Since the ultimate test of thequality of any communication is itseffectiveness in conveying pertinentinformation, the initial step in thedevelopment of standards of disclosure for published financial statements is the establishment of the purposes to be served. The potentialusers of corporate reports includegovernmental agencies, short- andlong-term creditors, labor organizations, stockholders, and potential investors. Since in all likelihood theneeds of these groups cannot beserved equally well by a single set ofstatements, the interest of some oneaudience should be identified as primary. Traditionally, this has been thestockholder group....

In considering disclosure standards,therefore, the Committee has beenconcerned primarily with the use offinancial statements (l) in making investment decisions and (2) in theexercise of investor control overmanagement, (p. 401)

The Committee's acceptance of the userorientation was not merely superficial; itshaped the entire Statement. However, thenext major statement of the Committee onConcepts and Standards, issued in 1957,ignored the user approach except in thelast section, "Standards of Disclosure,"which was based on Supplementary Statement No. 8. As a result, AAA literature didnot contribute heavily to the acceptance ofthe decision-usefulness approach to accounting theory until A Statement of BasicAccounting Theory was published in 1966.

Chambers' "Blueprint" article mightwell have served as the starting point for adecision-usefulness theory of financial accounting:

... a formal information-providingsystem should conform with twogeneral propositions.

The first is a condition of all logical discourse. The system should belogically consistent; no rule or pro

cess can be permitted which is contrary to any other rule or process

The second proposition arisesfrom the use of accounting statements as a basis for making decisionsof practical consequence. The information yielded by any such system should be relevant to the kinds ofdecision the making of which it is expected to facilitate, (pp. 21-22)

When considering "the introduction of additional rules,... in every case the test isrequired to be: 'What is necessary forrational decision-making' " (p. 24). Whilethat view provided an excellent foundationfor the development of a detailed decisionmodel approach, in his subsequent worksChambers apparently rejected the idea ofbasing an accounting theory on the decision models of specific user groups. Instead, he emphasized the general usefulness of "current cash equivalents" muchin the manner in which MacNeal acceptedcurrent market values and Canning accepted the present value measure of capital. Chambers explained that "the effect ofthe arguments of this book is to shift the focus of attention from the parties of interest(creditors, investors, managers) to the entity under consideration..." (1966, p. 375)."To provide the corroborable and corroborated financial statements which willserve as foundations for everyman's evaluations and actions is the business of thekind of accounting here envisaged" (1966,p.376).

Work on the decision-usefulnessapproach eventually led to more specificreliance on decision models. Staubus(1961) emphasized that "accountantsshould explicitly and continuously recognize an objective or objectives of accounting," and "that a major objective of accounting is to provide quantitative economic information that will be useful inmaking investment decisions" (p. viii). Hethen constructed an interrelated set ofbasic concepts and described a cash-flow-oriented measurement system keyed to thedominant theme of finance theory of thetime — the discounting view of securities


value. This led to the position that the attribute of assets and liabilities that is mostrelevant to security investment decisionsis discounted future cash flows.

The current status of the decision-usefulness, decision model approach to accounting theory may be summarized asfollows:

1. The primary objective of accountingis to provide financial informationabout the economic affairs of an entity to interested parties for use inmaking decisions. This objectivestatement is a premise which mostpeople seem to find acceptable, subject to slight variations. But it maynot be suitable unless one interprets"decisions" broadly. For "decisions"to encompass what might be calledthe control objective, the term mustinclude making choices regarding theinvestigation of variances, choices byemployees of the honest or dishonestcourse, and choices as to treatment ofoffending employees. If such an interpretation is not acceptable, a secondobjective — the control objective —must be given equal billing.

2. To be useful in making decisions,financial information must possessseveral normative qualities. The primary one is the relevance to the particular decision at hand of the attributeselected for measurement. The secondary one is the reliability of themeasurement of the (relevant) attribute. Objectivity, verifiability, freedom from bias, and accuracy areterms for overlapping parts of thereliability quality. Other qualities,such as comparability, undcrstanda-bility, timeliness, and economy, arealso emphasized. A set of such desirable qualities is used as criteria forevaluating alternative accountingmethods.

3. The relevance criterion is used toselect the attribute(s) of an object orevent to be emphasized in financialreporting. Information about an attribute of an object or event is relevant

13

to a decision if knowledge of that attribute can help the decision makerdetermine alternative courses of action or to evaluate an outcome of analternative course of action. Selectionof the attribute most relevant to a decision requires familiarity with thedecision processes of the user of financial data. "Modeling" decisionprocesses is often helpful to accounting theorists.

4. The decision-usefulness approachprovides for the development of theory on the basis of knowledge of decision processes of investors, taxingauthorities, union negotiators, regulatory agencies, and other externalusers of accounting data, as well asmanagers. To date, however, only thedecisions of investors (in the broadsense) have served as the basis forfairly complete theories of externalreporting.

5. The investment decision models utilized by decision-usefulness theoristshave been either simple present valuemodels or two-parameter expectedreturn and risk models. Risk lias notbeen formally incorporated in accounting theories to any great extent;expected future cash flows to investors have. Building on these commondecision-usefulness aspects, variouswriters have followed alternativepaths.

6. Investors' desires to predict cashflows from the firm have led manydecision-usefulness theorists to acash flow orientation (Staubus, 1961;AAA, 1969; Revsine, 1973; StudyGroup on the Objectives of FinancialStatements, 1973).

Cash returns to investors dependupon the firm's capacity lo pay,which, in turn, depends upon itspresent cash balance and its cashflow potentials. Present cash andpositive cash flow potentials areassets; negative cash flow potentialsare liabilities. When reliable evidenceof future cash flows is available it


should be used in the measurement ofan asset or a liability. Otherwise, asurrogate measurement methodmust be used. Accordingly, severalstudies have explored the theoreticalconditions under which various financial measurement alternatives wouldtend to provide surrogates for futurecash flows (Revsine, 1970a, 1973;Staubus, 1967,1971).

When the measurements of all assets and liabilities reflect or havesome relation to cash flow potential,the difference—owners' equity—hassome relation to the cash flow potential of the owners' interests. Revenues, expenses, and income, aschanges in net assets, are estimatesof the change in the net cash flow potential of the firm from operating activities.

7. One variant of the cash flow approach(Revsine, 1973) concentrates on "distributable operating flows" as amajor basis for dividends. Distributable operating flows make up "thatportion of net operating flows whichcan be distributed to owners withoutreducing future physical operatinglevels" (p. 47).

Maintenance of physical operatinglevels requires replacement of services consumed, thereby requiringcash equal to the replacement cost ofsuch services. When expenses aremeasured on the basis of replacementcost, the resulting income is ameasure of distributable operatingflows. This emphasis on maintenanceof productive capacity was also incorporated into the Securities andExchange Commission's replacement cost disclosure requirement(SEC, 1976).

8. According to an alternative view(Chambers, 1966; Sterling, 1972b),the impossibility of measuring afuture event rules out the cash floworientation; furthermore, the irrelevance of replacement cost of an assetnow owned to future events involvingthat asset rules out replacement cost.

Only present exit values are bothrelevant and objective. Even in casesof assets not likely to be sold in theirpresent condition, exit values arerelevant to (a) an appraisal of thefirm's liquidity, (b) a judgment of itscapacity to engage in indirect exchanges (Chambers, p. 101), and (c)an assessment of the risk a shareholder bears when he foregoes alarger sum of money (the price of ashare) in order to hold an interest in acollection of assets (Sterling, p. 206).

Proponents of the decision-usefulness,decision model approach to accountingtheory acknowledge certain implementation issues. Of particular importance arethe issues associated with identification ofthe decision processes of users. First, thereis the question of whether to rely on normative models or descriptive models.Sterling (1967) poses the choice as a fundamental dilemma:

If we are convinced that the receivers(decision makers] are using thewrong decision model, we have adilemma. (1) We can transmit the information specified by their [decisionmakers'] wrong model which willyield right decisions only by chance.(2) We can transmit the informationspecified by the right model whichwill be irrelevant to their [decisionmakers'] model, and hence right decisions will result only by chance, (p.106)

Differences of opinion about how to avoidthe dilemma have arisen. Some haveargued that we ought to focus on users andignore the difficult, perhaps impossible,problem of determining when a decisionmodel is right or wrong. This is what wehave classified as the decision-usefulness,decision-maker approach (see Abdel-khalik, 1971) and is discussed below.Others have argued that the nature of theprocess requires that accountants becomeinvolved in the evaluation, refinement, andconstruction of decision models and thenthat they educate the receivers in the use ofthose models (seeSterling, 1970a, pp. 59ff).This is the decision-usefulness, decisionmodel approach that is discussed in this


section. Debate over decision-model versusdecision-maker approaches to theory construction continues.

Another problem is how to deal with different user groups with different information wants (Revsine, 1973, pp. 5-8). Forexample, is there a conflict between theconcern of short-term creditors with short-term ability to pay and the shareholders'concern with long-run payoffs? Does thispossible difference imply the need for different measures of the same asset or liability, or only for recognition of the difference in reporting on short-term and long-term items in the balance sheet and relatedflow statements? Not only might differentuser groups want different data, but different individuals within one user group maydiffer in their decision processes in waysthat call for different financial data. Issuesof this type are still open.

Criteria Useful in Making AccountingChoices. Normative standards, or criterianecessary for information to be useful,serve a major supporting role in the decision-usefulness, decision model approachto accounting theory. Long before the decision-usefulness objective was explicitlyadopted by accounting theorists, severalqualities of financial information wererecognized. For example, Sanders, Hatfield, and Moore (1938) emphasized conservatism and consistency of accountingmethods. The AAA statement (1941) gavepassing attention to objectivity, inter-period and intercompany comparability,and consistency. Moonitz (1961) includedobjectivity, consistency, and disclosure inhis group-C postulates (imperatives). Butnone of those authors accepted the ideathat criteria of useful information shouldbe a cornerstone of their theories, becausethey did not emphasize the usefulness of information.

The first authors to advocate decision-usefulness also gave more attention to thequalities that were believed to make information useful, but they did not make theconnection a very direct one. Thus, although Chambers (1955) was the firstwriter to emphasize the relevance of information, he did not develop the concept tothe point of delineating the relevance of

IS

specific information to selected decisions.The AAA Supplementary Statement No. 8(1955) underscored interperiod and intercompany comparability, recognized theimportance of timeliness of reporting, andcame close to the relevance criterion whenit suggested that "the information to beconveyed should be selected on the basis ofits significance to the investor ...." (p.401). Staubus (1961) included a section on"Criteria for the Selection of MeasurementTechniques" in which he discussed relevance, accessibility of evidence, cost, bias,and reliability, and ranked six measurement methods according to the criterion ofrelevance (p. 51).

The role of criteria of useful information in accounting theory achieved prominence in the mid-1960s with the work of theAAA Committee to Prepare a Statement ofBasic Accounting Theory (1966). A Statement of Basic Accounting Theory(ASOBAT) began with a discussion of accounting objectives that specified information for making decisions, for directing andcontrolling an organization's resources, forthe custodianship of resources, and for facilitating social functions and controls. Itthen turned to standards or criteria to beused in evaluating potential accounting information, the all-embracing criterion being usefulness. Other criteria were formulated by asking, "What characteristicsshould accounting information have in order to be useful?" Relevance, verifiability,freedom from bias, and quantifiabilitywere suggested as standards. Consistencyand uniformity were included among thesecond-tier "guidelines for communicatingaccounting information." Sterling (1967)provided further explication of these standards in a review of ASOBAT. Snavely(1967) modified the ASOBAT structure byadding understandability, "worth morethan cost," timeliness, and significance(materiality), and by relating verifiabilityand freedom from bias under the headingof reliability. During the same period, Ijiriand Jaedicke (1966) contributed a discussion of the measurement of reliability,objectivity, and bias, while McDonald(1967) elaborated on those and other"feasibility criteria" in a treatment of dispersion, displacement, timeliness, cost.


and correlation (of the accounting numbers with an attribute to be predicted). Healso mentioned the all-inclusive criterion ofnet social benefit. Subsequent works making heavy use of the multiple-criteria approachwere Accounting Principles BoardStatement No. 4 (1970, Chapter 5), Staubus(1970), Sterling (1971), Kenley and Staubus(1972), Revsine (1973), and the Report ofthe Study Group on the Objectives of Financial Statements (1973).

It is crucial to observe the relationshipbetween the decision-usefulness objectiveand the evolution of independent normative standards (suchas comparability andobjectivity) into a coherent set of criteriaof useful information. First, the acceptance of the usefulness objective raised andemphasized the question, "What attributesof financial data make them useful to decision makers?" Without acceptance of thedecision-usefulness objective, it seemsunlikely that the multiple-criteria approach would have been developed to theextent that it has. Second, the normativeconcept that is generally recognizedas primary — relevance to the decisions of usersof financial statements — grew out of, andis dependent upon, the decision-usefulnessobjective.

Some criteria which have been advocated in contemporary literature for use inevaluating proposed accounting methods(information systems) and comparingcompeting proposals are presented in thefollowing paragraphs.

Relevance. "Relevance" has beendefined in a variety of ways, some of themrather vague and not helpful. Forexample,a general usage definition of "relevant,"such as "bearing upon the matter athand," may not be operational. Otherpossible definitions would make the termalmost synonymous with usefulness, thusruling it out as a criterion of usefulness.The current trend is to give relevance aspecific meaning related to the potentialroleof information about an attribute (e.g.,historical cost or net realizable value) ofan object or event in the decision process.For example, "An attribute of an object orevent is relevant to a decision if knowledgeof it would help the decision-maker evaluate an outcome of one or more of the alter

native courses of action under consideration" (Staubus, 1976). Or, "If a property isspecified by a decision model, then ameasure of that property is relevant (tothat decision model)" (Sterling, 1972b, p.199). Relevance is viewed as a necessarybut not a sufficient condition for usefulness.

Reliability. Users of financial information prefer that it have a high degree of reliability. Reliability is that quality whichpermits users of data to depend upon itwith confidence as representative of whatit purports to represent. But reliable information is not necessarily useful. It could,for example, be reliable but unrelated tothe use at hand. Several relatively generalterms are often used as synonyms for, or tocover parts of, the concept of reliability.Thus, verifiability, objectivity, lack ofbias, neutrality, and accuracy all are related to reliability. Like relevance, reliability (above some minimal level) is anecessary but not a sufficient condition forusefulness of data.

Other Criteria. Several writers haveemphasized the value of a criterion that focuses attention on the receiving phase ofthe communication process. Undcrstanda-bility and readability are sometimes suggested.

Timeliness, including both frequency(interval) and lag (delay), is widely accepted as a criterion of useful information.

Optima) quantity and cost are closelyrelated criteria. The literature of information economics stresses optimal partitioning and optimal aggregation. The possibility that certain information may not justifythe cost of producing and utilizing it is generally recognized. Issues regarding cognitive processing levels (Revsine, 1970b;Driver and Mock, 1975) and their impact ondecisions have also been recognized.

Many accountants believe that comparability is an important criterion forchoosing among accounting methods.Others point out that the several types ofcomparability that are desirable may beincluded in relevance, reliability, or othercriteria.

A number of accounting writers believethat the usefulness of accounting data inpredicting phenomena of interest to deci-


sion makers is a criterion that should beused in evaluating alternative accountingproposals. Thus, Beaver, Kennelly, andVoss (1968) recommended the use of "predictive ability" as a criterion. Althoughthese authors do not necessarily suggestthat the accountant do the forecasting,other theorists have attempted to distinguish measurements from forecasts andhave argued that the measurement function should be performed by the accountantand the forecasting function performed bythe decision maker (Sterling, 1970a, 1972b;Revsine, 1973, pp. 40ff, 1975).That is, whileit is clearly recognized that forecasts arerelevant (indeed indispensable) to decisions, there have been some who haveargued that the very nature of the marketmakes it necessary for the risk-bearing decision maker, as opposed to the accountant, to prepare the forecasts.

A review of the criteria discussed abovewill disclose that some criteria (e.g., relevance) are applicable primarily to the selection of attributes to be measured.Others (e.g., timeliness) play a significantrole only in the choice of methods of dataprocessing and communication. Stillothers may be important in almost anytype of accounting issue. Finally, thosewho advocate the decision-usefulness, decision model approach have proposed as anultimate criterion that the benefits yieldedby accounting activities should exceedtheir costs.

Decision MakersIn the previous section we focused on

decision models; we now shift to decisionmakers and review certain empirical research bearing upon various issues of financial reporting. Such research can beclassified according to the level at whichthe behavior of decision makers is observed: the individual level or the aggregate market level. First, we focus on empirical research on individual behavior;the second section discusses aggregatemarket behavior.

Individual User Behavior. Empirical research involving observation of individualbehavior as it relates to accounting information has ordinarily been associated with

17

the term "behavioral accounting research" (BAR). The objective of BAR is tounderstand, explain, and predict aspects ofhuman behavior relevant to accountingproblems; that is, "to establish generalizations about human behavior that are supported by empirical evidence .... Behavioral science thus represents the systematic observation of man's behavior for thepurpose of experimentally confirming specific hypotheses by reference to observablechanges in behavior" (AAA, 1971b, p. 248).

Behavioral accounting research is relatively new. Devine's critical remarks in1960 expose the failure of accountants toexamine user behavior empirically beforethat time:

Let us now turn to... the psychological reactions of those who consumeaccounting output or are caught in itsthreads of control. On balance, itseems fair to conclude that accountants seem to have wadedthrough their relationships to the intricate psychological network ofhuman activity with a heavy-handedcrudity that is beyond belief. Somedegree of crudity may be excused in anew discipline, but failure to recognize that much of what passes as accounting theory is hopelessly entwined with unsupported behavior assumptions is unforgivable, (p. 394)

Subsequent to Devine's remonstration, agrowing literature in behavioral accounting has emerged. For example, Hofstedt(1976) provides evidence of the increasingshare of the literature over the last tenyears that has dealt with accounting andhuman behavior.

In this discussion, we are interested inempirical research on individuals as it relates to the selection of alternative accounting techniques in an external environment.

BAR studies ordinarily lack anyagreed-upon basis by which their resultsmay be assessed. Instead, BAR has beenprimarily concerned with studying thetechniques of data collection and analysis;there has been little attempt to developgeneral theoretical formulations of problems or of hypotheses to be tested. Thestudies represent diverse attempts to un-


derstand, explain, and predict human behavior in an accounting setting without anagreed-upon universal theoretical perspective. As a consequence, BAR lacks a theoretical base to facilitate the selection of appropriate accounting procedures for individuals in economic decision making.

For these reasons it is difficult to classify BAR as it relates to external reporting.Nonetheless, some writers have attemptedto develop systems of classification (Birn-berg and Nath, 1967; Hofstedt, 1972; andRhode, 1972). The most recent andprobably the most exhaustive attempt wasby Dyckman, Gibbins, and Swieringa(DCS, 1975). We have chosen one part oftheir classification to illustrate the generalnature of studies conducted in BAR relating to the external reporting environment.2It should be emphasized that ourdiscussion uses only the first section of theDGS paper and provides only a brief overview; readers should consult the originalsource for a more detailed discussion of thestudies involved.

BAR studies may be divided into fourgeneral classes according to financialstatement disclosure and the usefulness offinancial statement data: (1) the adequacyof financial statement disclosure, (2) theusefulness of financial statement data, (3)attitudes about corporate reporting practices, and (4) materiality judgments.

In testing for the adequacy of financialstatement disclosure, researchers haveused many different strategies. For example, one strategy develops a description ofa user's approach to financial statementanalysis in order to evaluate the reasoningunderlying that approach; it then assessesthe implications of that approach and reasoning for various disclosure issues (e.g.,Horngren, 1955,1956,1957). Another strategy focuses on certain interest groups andsurveys their perceptions and attitudesabout disclosure (e.g., Bradish, 1965;Ecton, 1969). A third strategy has been todetermine the extent to which specificitems of important information are dis-

2 More specifically, the following five paragraphsare essentially a condensation of pp. 4-17 of the DGSpaper. This area of research Is important. However,our discussion is brief because of the recency of theDGS paper and its excellent synthesis of the area.

dosed in corporate annual reports, using anormative index of disclosure as a basis forassessment (e.g., Cerf, 1961; Singhvi andDesai, 1971; andBuzby, 1974).

A second set of studies has focused onthe usefulness offinancial statement infor-motion to investors in making resource allocation decisions. In one method, users offinancial statements were asked to indicate the relative importance as information items of various factors in investmentanalysis (e.g., Baker and Haslem, 1973;Chandra, 1974). Another approach hasbeen to try to determine whether financialstatement data are used in decision making and whether their use is affected byother variables. This is done by creating arepresentation of a decision maker underquasi-laboratory conditions,and by studying the behavior of the subjects who makethe decisions (e.g., Falk and Ophir, 1973a,1973b; Libby, 1975a, 1975b). A final approach has been to attempt to measure theeffectiveness of the communication ofthose data (e.g., Soper and Dolphin, 1964;Smith and Smith, 1971; Haried, 1972,1973).

A third set of studies has attempted tomeasure the attitudes and preferences ofvarious groups toward current and proposed corporate reporting practices. Oneapproach has focused on preferences foralternative methods of accounting for specific transactions or events, such as a preference for a given form of business combination such as poolings of interest (Nelsonand Strawser, 1970; Brenner and Shuey,1972). Other attempts to measure attitudeshave focused, more generally, on attitudesabout how much information should beavailable, how much information is available, and how important given items are(Copeland, Francia, and Strawser, 1973;Godwin, 1975).

Another set of studies has focused onmateriality Judgments that affect financialreporting. Research on materiality judgments has generally attempted to determine what factors influence the collection,classification, and final summarization ofaccounting information (e.g., Woolsey,1973; Dyer, 1973; Boatsman and Robertson, 1974; Pattillo, 1975; Pattillo andSiebel, 1973, 1974; Rose et al., 1970;Dickhaut and Eggleton, 1975). The major


thrust of the research attempts to determine how much of a difference in accounting data is required beforeusers perceive a difference.

It should be noted that these citationsare not intended to be all-inclusive, nor isthere any attempt to explain thesestudiesand subsequent recommendations in anydepth. However, this brief overview shouldprovide a perspective on the ways inwhichsomeBARspecialists haveconducted theirempirical research inthebehavioral area.

Aggregate Marfeet-Level Research. Theinterest in decision usefulness extends notonly to individual user responses to accounting variables, but to aggregate userresponse as well. To be sure, aggregatemarket behavior is a manifestation of individual action. However, according to proponents ofmarketlevel research, therearefactors that are difficult to simulate in individual level research (such as competinginformation sources, incentives, and userinteractions) that are important in thestudyofgroups; those factors thusprohibita simplistic extension from the individualto the aggregate (e.g., see Gonedes, 1972;and Gonedes and Dopuch, 1974). Indeed,they may be so significant that theoriesabout individual behavior and theoriesabout market behavior become, in fact,theories about distinctly different things.^Therefore, some researchers believe thataggregating individual users' responsesmay not provide an apt description ofmarket-wide user behavior.

Much early research regarding relations between accounting variables andmarket behavior has been based on thetheory of capital market efficiency. Thetheory predicts that, on average, the abnormal return (return in excess of the equilibrium-expected return) to be earnedfrom employing asetofextantinformationis zero (see Fama, 1970). This implies thatan alteration in the information set will result in a prompt transition to a new equilibrium.

3 A discussion of selection of paradigms or research programs is provided inChapter 4. For arecentreview and analysis of aggregate market-level research, see Kaplan (1975).

19

The theoryis notspecificwith respecttothe information set, and technical problems arise when it is admitted that theprice actually reflects theunderlying information (Grossman and Stiglitz, 1976).However, studies of accounting data haveshown that upon release of those data,there is (at best) only a short-lived opportunity for abnormal returns. The promptadjustment to a new equilibrium in conjunction withthedissemination ofaccounting data is consistent with the notion thatthose data are useful or possess pragmaticinformation content.4 Following that logic,researchers have assessed the pragmaticinformation content of various accountingdata by studying the timing of the incidence of abnormal returns. To do so, onemust identifytherelevantdates(i.e.,whenthe data were externally available,andthelength of the adjustment period to be observed) and a model by which to specifyabnormal returns. One such model is thediagonal model attributed to Sharpe(1963). The diagonalmodelimplies that theexpected return ona particular security isa stable, linear function of the return on aportfolio of all securities in the market.*Operationally, the intercept and slope parameters are estimated by regressing anindividual security return against the return on a market portfolio. Those parameters are then utilized to generate a prediction of the return on an individual securityduring a particular period. Any disparitybetween the actual return and the predicted return is considered to be abnormal.

4"Pragmatic information" refersto the reactionof individualsor groupsto a signor message. It is contrasted to "semantic information."whichrefers to therelation betweena sign or messageand the objectorevent that it signifies. For example, the message.•There is a fire in the building."will be said to havepragmatic information content if thereceivers ofthatmessage react to it, if, say, someone calls the fire department —thereaction of receivers determines themeasure of pragmatic information. Bycontrast, thesemantic informationof that same message is tested byobserving the building toseewhether ornotthere is infacta fire.SeeSterling (1970b. pp.44S-46. 453) for further discussion.

5 Evidence on the applicability of this importantmodel is provided by King (1966). Fama, Fisher, Jensen, andRoll (1969), andGonedes<1973).

20 Statement onAccounting Theoryand TheoryAcceptance

Since the mechanics of least squares regression produce, on average, a zero abnormal return, observation of large numbers of firms with abnormal returns coincidental with dissemination of accountingdata is taken as an indicant of pragmaticinformation content.

A number of studies have been conducted along those lines. Ball and Brown(1968), Beaver (1968), and Gonedes (1974)consistently observed abnormal returns inconjunction with the announcement of theannual earnings number. May (1971) observed similar reactions to the quarterlyannouncement of firm earnings. In addition, anticipatory market reactions havebeen observed, suggesting the presence ofcompeting information sources. Finally,the Gonedes study noted that the joint information implicit in several annual accounting numbers differed little from theinformation implicit in the earnings number by itself.

In other words, these studies are consistent with the notion that financial reports are useful. Stronger statements,though, must deal with questions about internal validity. The mere presence of anabnormal return coincidental withthe publication of accounting earnings provides asomewhat tenuous basis from which to infer that the observed price movement wascaused by the earnings signal. The principal basis upon which information contentin attributed to accounting earningsis therather impressive consistency with whichabnormal returns concurrent with the publication of accounting earnings have beenobserved.

The early studies of the pragmatic information content of accounting earningssignals were extended to changes in accounting policy. Researchers investigatedwhether "artificial" accounting changesthemselves affected security price behavior. The notion that security pricescould be alteredby merebookkeeping phenomena devoid of economic import (orsemantic information) dominated theliterature prior to the early 1970s. However, studies of security price reactions toaccounting policy changes by Kaplan andRoll (1972), Baskin (1972), Archibald(1972), and Ball (1972) generally indicated

that alterations in reportedearnings whichwere more or less devoid of present orfuture cash flow consequences could notproduce significant long-lasting effects onstock prices. In addition, Sunder (1973)hasprovided evidence that changes in inventory valuation methods which reduceearnings but enhance future cash flowstend to be associated with increases instock prices.6

The widespread belief that the meanand variance of an investment portfoliocomprise the interesting parameters of investor choice led to use of the diagonalmodel in examining the extent to which accounting risk measures are associatedwith a security's contribution to portfoliovariance.7 A measure of that contributionis captured in the slope parameter (beta)of the diagonal model. Studies by Beaver,Kettler, and Scholes (1970), Gonedes(1973), Lev and Kunitzky (1974), and BU-dersee (1975) indicated that accounting-based risk measures were contemporaneously correlated with beta. Thus, accounting risk measures appear to reflect whatever economic phenomena produce thevariance in a portfolio return.

Further, Beaver, Kettler, and Scholesexamined the usefulness of several accounting risk measures in predictingfuture values of beta; they concluded thataccounting-based forecasts of beta weresuperior to naive forecasts based solely onpresent values of beta. Thus, accountingrisk signals appear to have private valuetoan investor who seeks to predict the holding period risk of the portfolio. Perhapsmore importantly, Beaver, Kettler, andScholes suggested that security price research could be applied to the evaluation ofcertain accounting measurement issues.Specifically they posited that a simplifiedordering of competing methods of accounting for common economic phenomenamight be obtainable by observing the asso-

6 Curiously, however, Sander's data did not indicate significant decreases in stock pricesin conjunction with inventory changes which increase earningsbut reduce cash flow.

7 Either quadraticutility functions over returns ornormal return distributions are sufficient to demonstrate an exclusive interest in mean and variance. SeeHaleyand Schall (1973, pp. 101-05).


ciations between earnings variability andbeta, where earnings were defined usingcompeting measurement methods.

Beaver and Dukes (1972) subsequentlyexamined the associations between security prices and three competing definitionsof accounting earnings: cash flow earnings, earnings with deferral of.incometaxes, and flow-through earnings. Their association metric was a variant of the Abnormal Performance Index (API) used byBall and Brown (1968). The API measuresthe abnormal return that could have beenearned from holding a portfolio for a specific number of periods culminating in theperiod of an earnings announcement. Securities comprising the portfolio are selected with hindsight provided by someearnings expectation model. More specifically, securities that subsequently havehigher earnings than were originally forecasted are placed in one portfolio and securities with lower-than-forecasted earnings are placed in another. A positive APIassociated with the unexpectedly higher-than-forecasted earnings is alleged to beindicative of positive association betweenunexpected earnings and unexpected orabnormal price movements. Likewise, anegative API associated with a portfolio ofsecurities with lower-than-forecasted accounting earnings is alleged to indicatepositive association between unexpectedearnings and prices.8Notice, however, thatthe entire line of research is conditionalupon the particular model used to developan expectation of accounting earnings.Beaver and Dukes utilized a variety ofearnings expectations models to assess theassociations between security prices (viathe API metric) and their three competingdefinitions of earnings.9 They concludedthat deferral earnings generally possessedthe highest association.

8 It is important to note that the API is only ameasure of information in a foreknowledge sense. Fora discussion of the nature of foreknowledge, see Hirsn-leifer(1971l.

9The importance of expectations models to thistype of research is obvious. Several studies of the timeseries properties of accounting earnings have been undertaken. See Brown and Ball (1967), Beaver (1970).Ball and Watts (1972), Ball, Lev, and Watts (1976).Gonedes (1973).and Magee (1974).

21

Since their findings were inconsistentwith their expectations, Beaver and Dukesundertook a second study (1973) in whichthey treated various amounts of differencebetween tax expense and tax payable notas deferred charges or credits, but as aform of depreciation. They concluded thatearnings numbers computed upon deduction of depreciation amounts exceeding reported depreciation tended to display thehighest association with security prices.

To date, no other studies utilizing theAPI as a metric for ordering accounting alternatives have been forthcoming. As wediscuss in the next chapter, questions havebeen raised whether the API indeed can beused to assess the association betweenprice movements and accounting earnings.Further, Beaver and Dukes only suggestedthat their analysis provided informationregarding the associations between expectations sequences and a price sequence;nevertheless, other questions have arisenregarding the wisdom of interpreting suchanalyses as assessments of the relative desirability of accounting alternatives. Thosequestions are also discussed in Chapter 3.

Information Economics

The applications of economics to accounting problems discussed in the previous sections utilized economic theory forthe purpose of specifying what kinds of information were needed to make economicdecisions. In this sense, information wasnot incorporated in the economic theorybeing applied. It was often treated as a freegood, despite numerous references to acost-benefit criterion. By contrast, thearea of inquiry known as "information economics" treats information as a conventional economic commodity, the acquisition of which constitutes a problem of economic choice. That is, the product, information, is internalized in this approach tothe formulation of the problem. Previousapplications of economic theory had concerned themselves with the costs andprices of commodities other than information; this approach concerns itself withcosts and prices of information as well.

In the following two sections weexamine some insights provided by information economics. Initially, we review a


model for information choice that focuseson a specific individual's demand for information. Then we extend the analysis into adomain of multiple individuals.

The Single Individual CaseIn the single individual case, we view

the demand for information in terms of itsability to improve the quality of the ultimate choice in some choice problem thatthe individual faces. Put another way, thedemand for information is a derived demand; information is valued because it improves the quality of decisions.

To analyze a demand for information,we posit a basic choice problem that an individual must resolve. He must select oneaction from a set of at least two availablealternative actions. The precise outcomethat will follow selection of any specific action is not generally known. Rather, a set ofpossible outcomes is envisioned; only onewill occur, but precisely which one is uncertain.

If the individual's preferences are represented by the expected utility hypothesis, however, we have a precise representation of how his choice will be made. Eachpossible action gives rise to a lottery of outcomes, in which each possible outcome isassigned an action-conditional probability.(These probability assignments may beexogenously specified, as in the von Neu-mann-Morgenstern [1947] axiomatization,or endogenously specified, as in the Savage[1954] axiomatization.) Each possible outcome is also assigned a utility measure.The expected utility of each action is thenmerely the expected value of the outcome-utility measure. One action is preferred toanother if and only if its expected utilitymeasure is greater than that of the other.

Information can now be readily introduced into the analysis. The motive for acquiring information is to learn about whatoutcomes will be associated with what actions. Thus, with outcome possibilitiesassessed with a probability measure, performing an experiment or producing an accounting report is viewed in terms of providing a random outcome whose observance leads to (Bayesian) revision of theoriginal outcome probabilities.

A decision of what information to ac

quire, then, is viewed in terms of a two-stagelotteryorgamble. First, the information source or accounting system will produce some signal or message. Preciselywhat signal will be produced is uncertainand probabilities are assigned to eachpossibility. Second, following signal observance, outcome probabilities are revised and a conditional best action is selected, thereby resulting in an outcome lottery. In turn, theexpected utility ofthespecific source is the joint expectation of theutility of the outcomes associated with theconditional best actions. One reporting system is preferred to another if and only if itsexpected utility measure is higher thanthat of the other. Moreover, at zero cost,less information would never be preferredto more. "> (Further note that the cost of theinformation is accounted for by appropriately adjusting the outcomes that may materialize.)11

We have, in other words, a specificdescription of how information is used(Bayesian revision) and a specific criterion for designing an information or accountingsystem (subjective expected utility maximization). One limitation of thisapproach results from the traditional economic assumption of consistent, rationalchoice behavior.

The Multi-Individual CaseNow consider a situation in which a

number of heterogeneous individuals havea demand for information, such as publicfinancial reports. Analyzing this case in aneconomic manner pushes us into the realmof welfare economics. Two questionsemerge. First, given that we treat information as an economic good, is there any reason to worry about regulating its production, or can we treat the multiple-usersituation in a conventional market setting?Second, if we cannot rely on a market solution and thus must consider regulatory in

to This is Blackwell's theorem. See Marschak andRadner (1972) or Demski and Feltham (1976).

II The basic work in information economics is attributed to Jacob Marschak. See Marschak and Radner(1972). Application to accountingissues is discussedinDemski (1972). Feltham (1972). and Demski and Feltham (1976).


tervention, what is the nature and extent ofthe regulatory activity that should be employed?

The Regulation Question. To shed light onthe first question, we initially review theeconomic arguments for intervention in amarket economy. We then apply thesestandard concerns to an examination of theproperties of information.

If we adopt a normative social choicecriterion of Pareto optimally, to be consistent with this criterion, an allocation ofresources should be interpersonally efficient in the sense that no reallocation canstrictly improve one group's well-beingwithout harming some other group's well-being. Quite naturally, then, Tor intervention or regulation to be an issue, a laissezfaire determination of the production andconsumption of goods and services mustprovide an allocation of resources that isnot efficient.12 Under the usually assumedindividual, firm, and market conditions,however, the laissez faire solution is, infact, efficient.13 That is, a further increasein any group's well-being can occur only atthe expense of some other group. Underthese conditions, there is no incentive forregulation at this point.

That pleasant situation is altered, however, if we recognize the possibility in in-terdependencies that are not properly mitigated by the prevailing market structure.That is, if the market system fails to incorporate all of the interaction effects ofresource allocations, efficiency is notassured. Several examples of these inter-dependencies (or externalities) exist: consumption or production by one may polluteanother; increasing returns to scale maybe present; or the extreme case of a purepublic good (where acquisition by one

12A second argument for intervention is based on"ethical" or "humanitarian" grounds; in such a casethe distribution of goods and services is at issue and.stricUy speaking, we address the questionof movementfrom one efficient point to another. Of course, publicgood and externality aspects are also involved in thedistribution question.

13 See Arrow and Hahn (1971). Debreu( 1959). Henderson and Quandt (1971). Malinvaud (19721,or Quirkand Saposnik (19681 for further discussion.

23

makes the identical amount costlessly andautomatically available to all) may be theissue at hand.

The difficulty in such situations is thatindividual decision agents do not internalize the effect of their choices on others.A polluter, for example, does not naturallyaccount for the effect his pollution has onothers — a fact stressed by recent concernfor social accounting programs. A more extreme case of externalities is the phenomenon of a public good, that is, one availableto a single consumer only if the producerprovides the good to all members of the society. Such goods cannot be appropriatedby some individuals lo the exclusion ofothers. Thus, a price cannot be charged forthe good that will effectively reflect differing individual preferences for the good.For example, consider national defense.The protective benefits of national defenseare enjoyed by all members of society; thebenefit that an individual receives from national defense is not reflected in a price heis willing to pay, since he receives the benefit of national defense production whetherhe pays or not. Thus, private producers ofpublic goods are not able to incorporate intheir decision making the positive effectsof production on nonpurchasers (therebycreating the suspicion of under-production). Without such internalization,market-based solutions may not be efficient and the question of regulationarises. •«

Only the question arises, however.Existence of an externality effect (or itspublic good extreme) is not a sufficientcondition for intervening in the laissezfaire solution. Establishing a correctivemechanism will be costly, and those costsare one of the determinants of whether intervention is desirable. Similarly, the affected parties may be naturally motivatedto resolve their differences privately (e.g.,merger). In other words, intervention is avastlydeeper issuethan mereobservationof a potential inefficiency.

Next we consider whether externalities

14 Analyses are available in the literature. See Dueand Friedlaendcr (1973). Henderson and Quandt(1971). and Malinvaud (1972).

24Statement on Accounting Theory and Theory Acceptance

arelikelytobeassociated withinformationin an otherwise classical economy.

Analysis indicates that financial accounting information sharesmuchin common with the more traditional examples ofexternalities. Its use by one does not necessarily preclude subsequent enjoyment byothers. Once produced, anannual reportornewspaper mayberead bynumerous individuals. And in the extreme, its value toany specific individual may be completelyindependent of who else possesses it.«Coupling the possibility of nondestructioninconsumption with the facts that productionof information is,by nature,a risky activity(uncertainty isanecessary conditionfor information to have value), and thatprevailing property rights may make thetotalbenefit from production notcompletelyappropriable bythe producer, weareledto conclude that, without intervention, toolittle information will be produced (seeArrow, 1962, or Due and Friedlaender,1973). This is one of the standard arguments for such activities as patent lawsanddisclosure policies. Ofcourse,we musttemper this argument by recognizing thecostliness of intervention (see Demsetz,1969)

The possible difficulties are, however,further complicated if oneadmits that thedistributionof informationacross individuals is alsoan important determinantof theconsumption schedules ultimately enjoyedby each. That is, if everyone had the sameinformation, blackmail, sign-stealing inbaseball, and code-breaking in war wouldbe unnecessary. Other illustrations areprovided by moral hazard andadverse selection phenomena. In the phenomenon ofmoral hazard, two parties make an insuranceagreement. Buttheobserved event onwhich the insurance is based is an outcomethat is jointly determined by nature andthe insured. That, in turn, reduces the insured's decision-making incentives (as incost-plus contracting), and we therebyar

ts Gonedes (1975> formulates a problem where theprivate value of information is independent of howmany use it; with costly production this leads to amonopolistic producer. Under conditions thatallow forexclusion ofnonpurchasers heis thenabletocharacterizeefficientproduction anduseconditions.

rive at insurancecontracts that are less efficient in risk sharing than would be thecase with better information.^ This is, infact, a special case of the problem of incomplete markets(seeRadner, 1974).

Adverse selection, on the other hand, occurs when private information provides abasis for self-selection. Individuals with"bad" used automobiles have an incentivetomisrepresent theirautos intheusedautomarket as "good" because a buyer's informationabout the quality of the auto is inferior to that of the seller. Indeed, a potential seller of a "good" used automobilemaybeunable toattracta price thatisconsistent with the auto's quality because ofthe "bad" autos that are misrepresented inthe market. In a similar fashion, sickness-prone people haveastrong incentive toapply for health insurance. In turn,countervailing institutions such as advertising orindependent appraising arise; interventionmay be sought, as in the caseof licensingor mandatory health insurance (see Aker-lof, 1970; Hirshleifer, 1973; and Spence,1973).

These distributive effects associatedwith information differences across individuals have begun to be examined in aninvestment setting. The work of Hirshleifer(1971) and Fama and Laffer (1971), (andcontinuing with that of Gonedes, Dopuch,and Penman, 1975; Jaffe, 1974; Kihlstromand Mirman, 1975; Marshall, 1974; Ng,1975; Radner, 1974; and Wilson, 1974), provide examples. A major implication oftheserelatedexplorations is that it is possible to have overproduction as well as underproduction of information. For example, in a pure exchange economy withhomogeneous beliefs, information production merely redistributes the wealthamong the individuals; no social gainwhatever is produced. Indeed, it is alsopossible to construct examples where —due to the distributive effects — all individuals are unanimous in opposing the production of information (see Marshall, 1974;and Ng, 1975).

16See Kihlstrom and Pauly (1971) and SpenceandZeckhauser(1971) foranalysisof the effect of basinginsurance contractson incomplete knowledge of the insured's behavior.


In summary, externally reported financial information is a public good and thisintroduces the possibility — as is true withother public goods — of underproduction(in this case, underproduction of information). On the other hand, distributive effects of information differences introducethe possibility of overproduction of information. Thus, putting these two phenomena together, it is not at all clear whether inefficiency results in a market setting. Intheory, the answer is clearly affirmative.But in a real-world situation, we simply donot know. No research, to our knowledge,has systematically addressed that question.

The Normative Criterion Question. With inefficiency an open question, discussion ofintervention and regulation is, of course,somewhat premature. Furthermore,issues of "fairness" are apparently also involved. Blackmail is generally frownedupon in the sense that an "unfair" distribution of information produces the transfer ofwealth. Similar thoughts are expressedabout "unfair" advertising, incompletecandor in real estate and security sales,and insider trading. (Of course, allowinginsider trading may be a least cost methodof producing the information.) In anyevent, the precise motivation for intervention is ambiguous at this point.

Suppose, however, that we face a situation in which the market allocation is inefficient, or is the product of an undesirableinitial allocation of resources (includingaccess to information), or both. Some formof intervention will be pursued in order toachieve a more desirable, efficient allocation. The question then involves what criterion should govern this policy decision.

Quite simply, the problem is one of finding a criterion to guide nonmarket allocations. Cost-benefit analysis is a popularterm for a large subset of such activity.The concept is deceptively appealing: the"costs" and "benefits" are tallied and analternative with maximal gain is selected.Unfortunately, the concept of employingsome well-defined objective function islargely illusory in that it runs afoul of theArrow Paradox. This concept is furtherdiscussed in Chapter 3.

25

In summary, the information economics approach offers an explicit individual-demand-based analysis of accounting policy questions. Rationality is themajor assumption employed. This approach provides a means for examiningwhether regulatory intervention (i.e., rulesspecifying form and content) is desirablein an external reporting context. Thepower of the approach is in isolatinggeneral relationships and effects of alternative scenarios. At present, however, theapproach is still too general to provide definitive answers to existing policy issues.

Summary

In this chapter we have identified andclassified three prominent approaches tothe construction of accounting theory: (1)classical ("true income" and inductive)models; (2) decision-usefulness; and (3)information economics. Our synopsisclearly indicates many differences in problems addressed, assumptions made, andanalytic methods employed across thevarious approaches. While the differencesare fundamental and starkly visible, andwhile the issues and conclusions are ofteninconsistent among alternative approaches, theorists have had little successin reconciling these differences or in persuading critics that their unique structureis superior to others. Issues are continuallyrecycled and closure appears to be nonearer.

This situation naturally requires explanation, and in the next two chapters thisstatement endeavors to provide it. In Chapter 3 we explore conceptual issues thatarise in the specific accounting environment. As we will see, these issues are currently unresolved and consequently serveto preclude theory closure at this time.Worse still, many of these issues are notamenable to a solution using the mode oflogic and rigorous analysis upon whichtheorists have long relied. In Chapter 4 wegeneralize our Chapter 3 discussion and offer a philosophy of science interpretationof the continuing disagreement (recognizing, of course, that competing modelsexist). That chapter provides insight into

Statement onAccounting Theory andTheoryAcceptance

narvasive basic problem - and other disciplines - such an arduousthe nature of_ thei P«'ress in acCounting task,that makes theory P* e


Appendix

This Appendix contains analyses of thewritings of Sweeney, MacNeal, Paton andLittleton, and Alexander, which are intended to supplement the discussion in Chapter2. Full reference to the works cited in theAppendix will be found in the bibliography.

Henry W.SweeneyStabilized Accounting (1936)

Sweeney's other writings include"Maintenance of Capital" (1930), "Stabilized Appreciation" (1932), "Capital"(1933), and "Income" (1933).

Sweeney was much influenced byFisher. His decision on the dichotomy between capital and income is predicated onthe intrinsic (i.e., economic) attributes ofeach notion. Sweeney aligns himself withFetter in the famous Fisher-Fetter argument over whether the inclusion of capitalgains in income is double counting.

The closest that Sweeney comes to appealing to the needs of users is his statement that "... the true function of accounting is, or should be, to summarize financial data in such a way as to enable amaximum of helpful information regarding an enterprise to be obtained at a minimum of cost" (1936, p. 199). Only in thePreface to Stabilized Accounting does hespecify the users: primarily business management, but also the owners, bankers,federal and state tax authorities, and thegeneral financial public. Truthfulness isthe controlling tenet, but Sweeney does notshow how truthfulness leads to better decisions.

The entire thrust of Sweeney's argument is of the "true income" school, notwithstanding isolated appeals to the needsof various users.

Kenneth MacNealTruth in Accounting (1939)

MacNeal's work is predominantly of the"true income" school, although elementsof a decision model approach recur. Thedecision model is implicit.

Here is the epitome of the "true income" argument:

27

There is one correct definition ofprofits in an accounting sense. Aprofit is an increase in net wealth. Aloss is a decrease in net wealth. Thisis an economist's definition. It isterse, obvious, and mathematicallydemonstrable .... The function ofaccounting is to record, collate, andpresent economic truths.... (p. 295)

MacNeal, however, is not a partisan of discounted present value. The truth, he says,should be "unmodified by future improbabilities, probabilities, [and] even certainties" (p. 46). His preference is for "economic value," which he translates asmarket prices in a free, competitive,broad, and active market (Chapter 7). Inthe case of inventories, market price wouldseem to be selling price, yet MacNeal ordinarily is satisfied by the use of replacement cost as a surrogate. In general, herecommends market prices for "marketable assets," imputed economic values(e.g., appraisals or replacement cost) for"reproducible, nonmarketable assets,"and original cost less amortization or depletion for "nonreproducible, nonmarketable assets" (Chapter 9). These choicesare evidently a concession to the practicalproblems of measurement.

Notwithstanding his dominant true income orientation, MacNeal frequently refers to the information which stockholdersrequire in order to know the worth of theirshares (Chapters 9, 10, and 14, and p. 2).Moreover, "Creditors have the same rightto know the true status of a company as dostockholders. If the facts are misrepresented, creditors are deprived of a sound basisfor deciding whether to extend or withholdcredit. The same reasoning applies witheven more force to depositors and policyholders" (p. 221).

The use of discounted present values,contends MacNeal, "deceives stockholdersand creditors, and deprives them of atruthful basis of present fact upon which tobase their own estimates of the future" (p.146). The requirements of managers arementioned (pp. 180-81). Thus, MacNeal'sconstruct, while an appeal to truth as economic values, nonetheless reflects thevarious uses to which truth is put by diverse readers. MacNeal evinces a greater

TStatement on Accounting Theory and Theory Acceptance

explicit concern for the lot of specific usersthan does Sweeney. Neither author linkstruth to users' specific decision models.

W. A. Paton andA. C. Littleton

An Introduction to CorporateAccounting Standards (1940)

To many readers, this monographmight defy classification as either deductiveor inductive. Insofar as it represents adeductive argument, one would opt for thedecision model approach. Yet notwithstanding the ostensibly normative vein inwhich the argument is sometimes conveyed, the frequent use of the indicativemood recalls the Liltletonian alloy of deduction and induction —a massaged generalization. An example: "The accountant. . . deals primarily with the administration of the affairs of the continuing business institution and accordingly emphasizes the flow of costs and the interpretation of assets as balances of unamortizedcosts" (pp. 10-11). Do the authors implythat this is what the accountant should do,or only what he doesdo? Are they contentwith generalizing from practice and interpolating an integrated rationalization, or isit coincidence that their explanation ofproper accounting largely coincides, atleast in terms of result, with what accounting is? On at least some points (e.g., purchase discounts and "cost or market"), theauthors recommend procedures that appear to depart from settled practice. It isworth noting that one of the reviewers ofthe monograph, a leading practitioner,said that it "attempts to build a coherentstructure on a philosophical foundation,rather than to discover any coherence inthe field of accounting as it is" (Wilcox,1941, p. 75). In that day. practice was diverse, and practitioners were not accustomed to coherent structures. It is doubtfulthat a coherent structure any more faithfullo then extant practice could have beenproduced.

The decision model, if any, is implicit.The authors go no further than. "The purpose of accounting is to furnish financialdata concerning a business enterprise,compiled and presented to meet the needs

ofmanagement, investors, and the public"(headnote, p. 1). Although they underscorethe importance of the absentee owner,their discussion of the possible uses towhich current value information might beput is related to the decisions made bymanagement (e.g., pricing and replacement).

Transaction-based price aggregatesdominate the analysis. Few exceptions inthe accounts and in the body of the financial statements arc tolerated. "The emphasis on the cost basis . . . rejects the proposition that periodic revaluation has a settled place within the regular accountingframework" (pp. 125-20); that is a disturbing tautology. Current values and generalprice-level data, althoughimpermissibleinthe body of the financial statements, maynonetheless appear in supplementaryschedules or reports. These constitute "interpretive" information in the Littletoniansense.

An argument frequently given (pp. 123.135, 141) in support of historical cost is theconstraint imposed by legal requirements.At one place, the term "contractual earnings" is used, evidently connoting the notion of earnings contemplated by contracting parties, or perhaps by legislation andjudicial opinion. The authors do not arguefor an idealized notion of income, and theyare conscious of institutional boundaries.

The authors' persistently positivistform of expression confounds the reader.Another example: "Prom the point of viewof the managers and owners of the enterprise the cost incurred is regularly the significant measure of operating effort . . ."(p. 23). "Ought to be" or "is"? While thestated purpose of accounting, quotedabove, is couched in the indicative mood,the authors elsewhere state that accounting "must represent a practical tool ofbusiness and finance, competent to meetthe legitimate needs of managers, investors, government, and the public atlarge" (pp. 4-5). Fulsomcly normative.

Perhaps the clearest expression of theauthors' view of their mission appears onpage4:

A statement of accounting standardsshould represent an integrated conception of the function of accounting


as a means of expressing the financial facts of business in a significantmanner. It must inevitably embodysome conflict with existing accounting practices, since existing practiceis in conflict with itself at a hundredpoints. It should avoid any appearance of encouraging violations ofexisting law, but it need not accept asgood accounting all definitions, policies, and practices which legislatorsand courts have added to the accounting structure.

In vivid contrast to MacNeal's work, thePaton and Littleton monograph was evidently written so as not to annoy the practitioner audience. At that time, the American Accounting Association was attempting to replace rivalry by collaboration inits relations with the American Institute ofAccountants, as the AICPA was thencalled. Copies of the monograph were distributed without charge to members ofboth bodies, and the authors clearly soughtto achieve a maximum favorable impacton practitioners without sacrificing theirprinciples. This motive could explain theindicative mood.

Much of the writing is Paton's but allthe ideas are not. Price aggregates, andthe dominant "interpretation" theme ofthe final chapter are pure Littleton, Yetmost of Paton's ideas find expression in themonograph.

Notwithstanding the authors' frequentreference to "standards." even in the titleof their monograph, their approach cannotfairly be characterized as falling in thesame class as ASOBAT. Paton and Littleton enumerate measurement rules whichare assertedly predicated on the several

29

concepts, conventions, and assumptionsset forth in Chapter 2. If there is a pervasive standard (i.e., normative standard), itis dependability, which is frequently invoked in defense of traditional historicalcost. In fact, the authors did not undertaketo enumerate standards, but instead to laythe groundwork for a future exposition ofstandards.

Sidney S. Alexander"Income Measurement in a

Dynamic Economy" (1950)(revised by D. Solomons in

Studies in Accounting Theory, 1962)

Alexander's attitude toward the existence of "true income" is not easily categorized. Throughout the essay he repeatedlystresses that the utility of a given incomesystem must be assessed in relation to theintended use for the resultant measurements (see Baxter and Davidson. 1902. pp.130, 146,and 164). This strongly hints of thedecision model, rather than the true income, approach to income theory. However, in several places in the essay,Alexander makes statements which suggest that in theory at least an ideal incomeconcept does indeed exist. For example:

... we must find out whether economic income is an ideal from whichaccounting income differs only to thedegree that the ideal is practicallyunattainable, or whether economicincome is inappropriate even if itcould conveniently be measured, (p.159)

In summary, Alexander probably fallssomewhere between the "true income"and decision alternatives.

Chapter 3

Criticisms of Present Theory Approaches

This committee has looked for reasonswhy none of the alternative theory approaches discussed in Chapter 2 has risento a position of singular prominence in contemporary accounting thought. None ofthem has been left unchallenged; indeed, asteady stream of counterarguments andcriticisms appears to have prevented anyof the approaches from gaining a clear majority of accounting theory students as supporters.

Many prominent criticisms of alternative accounting theory approaches wereclearly discernible in 1966, when the Association's most recent major theoreticaldocument, A Statement of Basic Accounting Theory, was written. Since then, additional points of dissatisfaction with one ormore of these approaches have appearedin the literature. Further, some of the traditional criticisms have received renewedsupport.

In this chapter we survey some of thecontemporary points of conflict that arisein the accounting literature. Clearly, eachpoint is not shared by all critics. In the aggregate, however, they apparently serve topreclude a meaningful consensus regarding financial accounting theory. Our survey is not intended to be exhaustive; however, it should convey the fundamentallevels at which objections to proposed theoretical approaches arc raised. Six different issues are summarized in this chapter:(1) the general problem of relating theories to practice; (2) the allocation problem; (3) the problem with normative standards; (4) the problem with basing accounting theories on securities price research ; (5) the problem of cost-benefit considerations in accounting theories; and (6)the danger in assuming that more information is preferable to less. Collectively thesepoints of conflict illustrate why no theoretical approach has yet achieved dominantacceptance within the accounting community.

The Problem of Relating Theories toPractice

Apparently, accountantsexpect or hopethat the process of accounting theorizingwill lead to a sufficient and compellingbasis forspecifying the content of externalfinancial reports. However,even if a singletheoretical approach should rise to prominence,many of the day-to-day issues facedby practicing accountants and auditorsprobably would persist. For example, inmany proposed theories, issues of materiality wouldremain issues. Furthermore,to the extent accountants fail to perceiveand accept limits to the practical coverageprovided by a theory, all theoretical approaches are apt to be controversial because of the gaps they leave in guidingpractical applications.

One possible source of practical dissatisfaction with a theory is its lack of coverage or breadth. For example, some expositions of inductive theories have leanedheavily on statements to the effect that assets are valued at historical cost, and haveneglected those assets, such as cash andreceivables, which are not customarilyvalued at cost, along with all liabilities. Another example is the case of the theoristwho limits his concern with flows to incomestatement changes to the exclusion ofchanges in working capital, even thoughstatements of changes in financial positionare acknowledged to be a vital part of practical accounting. Theories with such gapsare bound to be perceived as inadequatefor yielding the answers that practitionersseek.

Another practical problem with theories is that they do not always recognizealternative possible features of the environment and specify how to deal withthem. This "weakness" of theories may beassociated with unrealistic assumptionsmade by the theory, such as competitiveequilibrium, zero transaction costs, immaterial fringe acquisition costs of commodi-

31


ties and services, complete markets, zerocostsof producing and utilizing accountingdata, stability ofthemonetary measuringunit, homogeneity of decision modelsamong members of a user group, and mutuality of interests of owners and managers. Observed inconsistencies betweensuch assumptions (and the resulting simplifications of the theories) and the environment in which accountants work maycause practitioners to lose faith in ailtheories.'

It is not clear that theorists can overcome the problem of incomplete specification by adding realism to their assumptions; the real world is very complex andcomplexity is alsoa weakness in a theory.A complex theory may be more difficult toapply, and therefore its popularity amongpractitioners will be lessened. For example, the evaluationof information systemsin accordance with the approach of information economics may escape the criticisms of narrow coverage or inadequateprovision for variations in the environment. But the complexity of the information evaluation task, as outlined in the information economics section of the previous chapter, may well cause many accountants to avoid it. Similarly, if a theorist specifies a large number of normativecriteria that an accounting method shouldmeet, the practitioner may not be able tomake the complex set of trade-offs that arerequired to apply the theory. Perhaps thenotion of an optimal quantity of data thataccountants should provide to users in order to maximize the net value of information can be adapted to apply to the amountofdetailed specifications in an accountingtheory. Butdifferencesamongaccountantsin their tolerance (or desire) for detailedguidance are bound tosubjectany theoryto criticism by some of its users! An admirably simple accounting theory is notlikely toyield immediate and unequivocalanswers to all accounting decision situa-

1The question ot whether a theoryshouldhe evaluated on the basis of the realismof its assumptionsinaddiiion to its predictive power remains unsettled. Cf.Friedman(195.11. Nagel(195.1), Samuelson (196,1), andCarsbcrg. Arnold,and Hope(1977).

lions. To the extent that accounting theories are evaluated on these criteria, notheory is likely to be wholly satisfactory.That is one possible explanation for thelack of general acceptance of any theoretical approach.

The Allocation Problem

Most of the alternative approaches tospecifying the content of external financialreports rely heavily upon allocation processes. Even if allocation procedures areavoided in formulating the general framework of a proposal, they are frequently introduced at a more specific level. Thomas(1969, 1974) has argued that the allocationproblem is an inherently insoluble one thatpervades most external reporting proposals.

The most serious objection to allocationtechniques is that many are inherently arbitrary. A familiar example of arbitraryallocation in U.S. accounting principles isthe oft-quoted description of depreciationaccounting published by the AICPA in1953: ' 'Depreciation accounting is a systemof accounting which aims to distribute thecost or other basic value of tangible capitalassets, less salvage (ifany), over theestimated useful life of the unit (which may bea group of assets) in a systematic and rational manner. It is a process of allocation,not of valuation . . . ." The absence of anyguidance as to the meaning of "rational" inthis context has enabled accountants to defend widely varying time-patterns of depreciation. Many accountants agree withThomas that this is a weakness in thosetheoretical statements reflecting U.S. accounting principles, such as ARS No. 7(Grady, 1965,pp. 148-51).

The extent of dissatisfaction with accounting procedures involving allocationsappears to vary widely. One probable reason that dissatisfaction varies is that themeanings attributed to the term "allocation" vary. Some accountants limit theirapplication of the term to cases in whichone monetary amount is split among several accounts or periods without direct reference to either physical or monetary evidence as a basis for the split. Others extendthe term to such cases as splitting rentamong departments on the basis of floor

Criticisms ofPresentTheory Approaches

space occupied, or even to charging thesalaryofan employee tothedepartment inwhich he worked. In the view of many accountants, there is a vast difference in thearbitrariness of these various procedures.Nevertheless, it seems safe to say thatthere is widespread dissatisfaction withthose procedures that appear to fall at themore arbitrary end of the spectrum.

The fact that an allocation technique isarbitrary does not mean that it is indefensible. On the contrary, in Thomas' ownwords:

The difficulty is more subtle: for eachsituation in which allocation is contemplated, there isa variety of possible allocation methods, each of which could bedefended. The allocation problem arisesbecause there is no conclusive way tochoose one method in preference to allothers, except arbitrarily. (1974, p. 2)

Thomas illustrates at length the arbitrariness of various classes of allocation techniques and the inability to defend thesetechniques rigorously to the exclusion ofavailable alternatives. Since conclusive selection is impossible, endless argumentation ensues. While some may disputeThomas' conclusion that (given our current state of knowledge) allocation techniques are totally arbitrary, at a minimumit is apparent that such allocations giverise to vexing controversies.• The irresolv-abilitv of these allocation-induced controversies explains, at least in part, why the-

2 An interesting manifestation of the allocationdilemma arises whenweattempt to define the conceptofincome inan uncertainty setting Following theHick-sian tradition, we think in terms of temporal wealthmeasurement. Yet future cash flows, in an uncertainlysetting, arccharacterized by probability distributions.Ina complete embracing ofstate Indexed commodityproduction and exchange (eg.. Debrcu 1959), however,nodifficulty is encountered. K.ach possible distributionof future cash flows has a known markc! priceandwealth calculations arc straightforward. But difficulties emerge if we donot have access to these marketprices-lhat is. ifthe markets are incomplete. Here thetraditional allocation dilemma again emerges,bothinterms of "assigning" periodby period components ofthecashflow as well as (recognizing slate dependence)in interpreting the resultant income measure. However, market incompleteness andtheresultant allocation dilemma, notuncertainty perir, IsIhecause ofthedifficulty.

33

oretical approaches in which allocationplays an integral role have not gainedwidespread acceptance.

The Difficulty with NormativeStandards

As discussed in Chapter 2. some theories of external reporting incorporate nor-matively posited standards that provideabasis for choosing information to be includedinexternal reports. Such standardsinclude objectivity, verifiability, and timeliness, amongothers. Oneimportant characteristic ofsome ofthese standards is thattheyrefer toproperties thatare intrinsic totheaccounting numbers themselves; theyare not inferred from the personal preferences of the individual users of accountingdata. As a consequence, the relationshipbetween theproposed normativestandardsand the individual preferences of decisionmakers is indirect. Since normative standards that arc divorced from individualpreferences cannot completely surrogatethose preferences, the imposition of suchstandards presents serious difficulties.Specifically, if weassume that an accounting system should be designed to maximize,say, the expected utility of the usersof accounting data, the data chosen to conform to imposed standards will not consistently lead to that result.

Consider an individual user of accounting data whose preferences can be represented by the expected utility hypothesis;that is, he guides his behavior in a mannerthat results in maximum expected utility.Further assume that our individual usesaccounting information to make more thanone type of economic decision. Since eachtype of decision incorporates potentially diverse dimensions, each distinct decisioncategorymay result inemphasison different properties of the financial phenomenaat hand. For example, if accounting policymakers adopted objectivity as a criterionfor selectinginformation to be reported tothis individual, the resultant informationsystem might actually maximize the expected utility forone typeofdecision madeby the individual. That is, the preferencesof the individual in the context of a specifictype ofdecision mayrequire that objectivity be the dominant criterion in construct-

34 Statement onAccounting Theory andTheory Acceptance

ing the system from which he obtains information input to his decision process. However, if the same individual uses the sameinformation in the context of a differenttype of decision, his preferences may require that objectivity not be the dominantcriterion in the second context.

Furthermore, no two decision problemsare identical. New environmental dimensions can suddenly materialize even in recurring, somewhat standardized decisionareas. In other words, the processof defininga "type of decision" for which a specified normative standard will be appropriate cannot be expected to be stable as environmental changes occur. As a consequence, there is no guarantee that a normative standard consistent with utilitymaximization on one occasion will necessarily lead to utility maximization in an altered decision environment. (Indeed, thesetting becomes more perverse if we allowthe individual's preferences to shift overtime, and we have no basis for assumingthat this is a rare occurrence.)3

3Ofcourse, one might argue that,intheideal case,the standardswillbe rich enough to "work" in all settings. However, this cannot be the case because thenecessary and sufficient conditions for such a state ofaffairs, as specifiedby Blackwell's theorem (see Marschak and Radner, 1972), preclude any such universalset of standards.

This argument is presentedby Demski (1973a). Itsessence is as follows. Without loss of generality, assumethe policy maker is faced withchoosing betweentwo accountingalternatives, if andrf. Further assumethatrj" andrj" can be compared in termsof fineness,wherein T is said to beas fine asr)" if1' provides asmuchinformation as rf and possibly more. Forexample,deferral of incometaxeswould appearlobeas fineas nondeferral, since deferral contains as much information as nondeferral, and perhaps more. If f andri*are costless, il is clear that rras fine asry is sufficienttoinduce the policy maker todesirerf at leastasmuchasrf. However, rj" as fine asn- isalsoa necessary con-dition for rf tobeatleast asdesirable asrf. <This istheessence of Blackwell's theorem.) Therefore, anystandards which are used to surrogate preferencesmust also surrogate fineness.

Thedifficulty is thatfineness isnotacomplete relation—some Information systemsarenotcomparable intermsof fineness. For example, UFO andFIFO probably cannot be compared on a fineness dimension.Since fineness is incomplete, there are coses in whichstandards cannot possibly surrogate fineness and, inturn, preferences. Therefore, standards surrogatingpreferences are. ingeneral,animpossibility.

In a more realistic portrayal of the reporting environment, the problems of using normative standards to specify the information system are complicated evenfurther by the obvious existenceof manyindividual users. Observation of human behavior suggests that the preferences of individuals are diverse. To cite a few examples, some individuals may valueprosperity highly, whileothersstrive merely fora"comfortable" living standard; some pursue increments to wealth under highlyrisky conditions,whileothers will not; andsome have a more highlydeveloped socialconscience than do others. Since individuals in a multi-person setting have diversepreferences, no single set of standards forexternal reporting is likely to be consistentwith thosediversepreferences.4 Moreover,the question of how diverse individualswould or should specify public reportingpolicies is largelyopen. And we simply donot know the relationship betweenany setof standards and such group-level choicebehavior.

Thus whether viewed in a single- ormulti-person setting, normative standardsdo not appear to offer a noncontroversialbasis for theory building or policychoice.

Difficultiesin InterpretingSecurityPrice-Behavior Research

As we noted in Chapter 2, considerableempirical research based on the efficientmarket hypothesis has attempted to discern how investors, viewed in the aggregate, use accounting measures. Briefly,that research tries to interpret the contemporaneous association of information flowwith unexplained variations in rates of return on securities as an indication of "informationcontent." Twocaveats are, however, important when we undertake to interpret these types of studies.5

4Some of the objections raisedagainstthe decision-usefulness, decisionmodelapproach are based on thispoint. Specifically, that approach seeks to provide information specified by the decision modelof a certaincategory of users. But this Ignores the potential redis-tributive effects on other users whose information desires are excluded fromthe analysis.

5Some, such asThomas 11874, pp. KWl),reject allsuch evidence on grounds that the individual behaviorsthat are aggregated in market statistics are them-


First, whether we interpret the studiesas attempts to assess consequences or desirability of measurement methodchanges, we nevertheless confront the traditional experimental concerns of internaland external validity (as we do in all empirical studies). Most clearly, thetestsarebasedon controlling for all "obvious" return factors (such as market-wide phenomena, arrival of other information, andrisk changes); the tests are thereforelimited by our ability to identify and allowfor the effect of such factors. Whether acorrelated variable is overlooked remainsproblematic (see Gonedes and Dopuch,1974).

Second, when one interprets the association measure as an indicator of either theconsequences or the desirability of measurement method changes, he must alsorecognize the inherent simplificationsbeing relied upon. In terms of consequences, price changes per se are not likely to indicate fully the impact of reallocation in a regime of incomplete markets,simply because of less than complete marketability. For example, questions of amanager's nonmarketable human capitalbecome an issue at this point, and with lessthan complete marketability, strict focuson market aspects may miss important dimensions of the consequences.

In terms of desirability, however, theargument is, perhaps, more subtle. Consider a strict presentation issue, as inBeaver and Dukes (1972,1973), where theunderlying depreciation and tax allocationdata were available to the reader and thenarrow question addressed was how to present these data. Subject to internal and external validity concerns, one might conclude that the "statement-proper" presentation yielding the highest contemporaneous association provides the data that individuals prefer to employ, and shouldtherefore be the accountant's choice(thereby saving on users' processingcosts).

But even if this line of attack is successful in pinpointing desirable (that is, efficient) modes of presentation, it falls short

selves the results of "bad" judgments or prior beliefsand hence cannot be relied upon in the search for a"good" measurement system.

35

of answering more basic questions.6 Forexample, Beaver (1972), and May andSun-dem (1973), point out that this approachcannot be employed to assess the desirability of unreported alternatives such as current values (assuming that the underlyingevents are not reported by competingsources of information). More fundamentally (as noted by Beaver and Demski,1974; and Gonedes and Dopuch, 1974), theapproach breaks down when it is pushedbeyond narrow format questions to the extreme of using market association studiesto rank accounting alternatives in general.Other criticisms are summarized by Sterling and Harrison (1974,pp. 146-48).

A similar question is whether the traditional guides of market prices and the market value maximization rule can be used torank accounting alternatives. The answer,unfortunately, is negative if the marketstructure is incomplete (Radner, 1974).This result is also articulated by Gonedesand Dopuch (1974). In their analysis, markets are initially assumed to be perfect andthe goal of firms is to maximize the currentmarket value of common shares; that is,firm decisions follow the market valuerule. Information is regarded as merelyone of the outputs that firms may producefor sale. Gonedes and Dopuch further assume that firms are capable of limitingconsumption of their information output tothose who actually purchase it and that nointer-user externalities arise. Under suchconditions, a market for information wouldexist; firms could observe prices in thismarket and assess marginal revenue andcost relationships to determine information output. Further, share prices would reflect changes in the value of the firm associated with its information production activities. Hence market value maximizationwould lead to efficient resource allocation.

However, difficulties emerge if we positthat firms do not have the ability to limituse of produced information to those whoactually purchase it. The inability to exclude nonpurchasers causes evaporation of

* Marshall (1975). for example, stresses the internal validity concerns in specifying the "abnormal performance index" as a measure of the association between unexpected earnings and unexpected pricemovements.

36 Statement onAccounting Theory andTheory Acceptance

the market for information, since information canbecostlessly obtained by all.? Thisinduced market incompleteness precludesdecisions about information productionthat are based on considerations of marginal revenue and cost.To illustrate, considera firmproducing information that, inthe presence of an information market,would add valueto the firm. However, theinformation produced has negative cashflow implications for the firm's noninfor-mation activities. In the absence of an information market, share price will fall,since no increase in the firm's value due toinformation production will occur to offsetthe loss in value related to noninformationactivities. Hence, as long as informationmarketsareincomplete, nogeneral correspondence between the desirability of information production decisions and sharepricemovementswillexist. Consequently,observation of market reactions may notbe helpful in resolving certain controversiesregardingaccountingalternatives.

The Problem of Cost-BenefitConsiderations inAccounting

TheoriesIn assessing the desirability of various

financial reporting alternatives, accounting theorists have generally not determined the benefits and costs related totheirrecommendations. The opinion that aspecified reportingalternative, such as accounting in units of general purchasingpower or extensions of interim reporting,should not be adopted because it is toocostly for the benefits received is expressed frequently, but supporting dataarerarelypresented. Thisis notsurprisingsince thebenefits andcosts ofanyproposalare exceedingly difficult to identify andmeasure. However, until such a demonstration occurs, it is not possibleto defendconclusively a particular accounting policyagainst competing alternatives. More generally, except for the informationeconomics approach that incorporates cost-benefitconsiderations at anabstract level, allapproaches to accounting theory fail to provide explicitly for the measurement and

7 Similarobservations emergewhenwereflectonthenotion ofa price that"reflects" allcurrently availableinformation (seeGrossmanandStiglitz,1976).

comparison of benefits and costs of accounting options. Many accountants feelthat this is a crippling weakness which willalways prevent them from achieving consensus regarding solutions to accountingissues.

Thus, formally incorporating the cost-benefit orientation of information economics offers an appealingavenue of theoretical pursuit. However, difficulties exist.Most obvious is the practical issue of howone operationalizes the expected utilitymeasures. With the proper axiomatic foundation, the mathematical existence of suchmeasures surely is guaranteed. But thistheoretical existence provides little guidance forconstructing such a measure andincorporating actual numbers—as must bedone to implement the approach in a real-world setting.

An even deeper problem also exists,however. That is, even the mathematicalexistenceof the preference, orcost-benefit,measure may be denied in a multi-personsetting. Recall that in a multiperson setting we seek nonmarket allocation criteriawhen considering how to intervene in orsupplement a laissez faire allocation process. An example arises when the FASBconsiders some disclosure policy. The issueis, what criterion should govern theFASB's deliberations. And the difficulty,known as Arrow's Paradox, is that a completeandtransitiveranking of the alternative criteria does not, given a set of seemingly innocuous assumptions, exist. Thatis, existence is denied.

Briefly, the problem as posited byArrow (1963) is oneof moving from individualtosocial preference. Tomake the problem interesting, at least two individualsand three alternatives are recognized(which is hardly a problem in an accounting setting). Arrow imposed four conditions that this relationshipshouldsatisfy:

1) Universal Domain. The individualsare unrestricted except that theymust display complete and transitivepreferences. Similarly, the socialpreference that emerges must becomplete and transitive.

2) Pareto Optimality. If all individualsstrictly prefer one choice to another,


then the social ordering should strictly prefer the one to the other. Thisconfines the choice to those alternatives that are efficient and hardlyseems debatable. Abandoning itwould amount to proposinga methodof choice based on systematicallydenying those affected what theywant.

3) Independence of Irrelevant Alternatives. If two alternative sets of individual preferences agree on a subsetof the alternatives then the corresponding social preferences mustagree on that subset of alternatives.This condition rules out interpersonalutility comparisons. The choice between any pair of alternatives mustbe based solely on the individual preferences for the two alternatives andnot on any other preferences theyhave (such as might be used to "calibrate" interpersonal comparisons).

4) Nondicta torship. There does not existan individual whose preferences arealways identical to society's.

Difficulty arises because these conditions are mutually inconsistent. No suchmethod of moving from individual to socialpreferences exists. As a consequence, thepolicy problem cannot be formulated interms of maximizing some conceptuallywell-defined criterion. That is, even with awell-defined criterion for accounting system design at the individual level, we donot emerge with a specific criterion forregulation.

How, then, does one formulate the problem of regulation or policychoice? Clearly,these choices are presently made and willcontinue to be made. Since whatever method might be employed violates at least oneof Arrow's conditions, one might structurethe problem in terms of selecting whichconditions to violate. Fundamentally, however, the criterion question is an open one.

Quitesimply, then, cost-benefitanalysisdoes not provide the basis for constructinga theory of accounting. Indeed, one cannoteven formulate the question of guiding non-market allocations in terms of costs andbenefits because, accepting Arrow's for

37

mulation of the problem, such measurements do notexists

To summarize, for some accountants,the failure to incorporate cost-benefit considerations in most theoretical approachesserves as a sufficient reason to reject thoseapproaches. On the other hand, propercost-benefit analysis does not appear possible with the present state of economicknowledge. This is not to say, however,that unambiguous social choice criteriacannot emerge in a multi-person setting.To the extent that the individuals involvedcan work out mutually acceptable sidepayments, any externality can beremovedand Pareto-optimal solutions achieved.Nevertheless, in general cases, cost-benefit analysis cannot provide a basis for constructing an accounting theory at the present time.

Limitations of Data ExpansionIn recent years a theoretical approach

to external reporting issues called "dataexpansion" has emerged. Simply stated,proponents of this approach argue thatmore information is assuredly preferableto less. Accordingly, they suggest thatmany types of accounting controversiescan be resolved through expanded disclosures.

Impetus for the data expansion approach arises from two distinct sources. Inone view, accountants' limited knowledgeof users' preferences and behavior makesit difficult to defend a single reporting option against available competing alternatives. As a consequence, it is argued thatthese difficult choices can be avoided bypresenting more detailed disclosures or byreporting simultaneously several alternatives that satisfy some (usually unspecified) logical criteria. Sorter's "events approach" (1969) illustrates this position.

A second perspective, market efficiency, also lends support to the data expansion approach. In this view, the "efficient" market in the aggregate presumably is not misled by the form or content ofaccounting disclosures. As long as there

8 See Beaver and Demski (19741 for an analysis ofthe problem in terms of the political institutions we devise to answer such questions.

V38 Statement on Accounting Theory and Theory Acceptance

are sufficient disclosures, the discloseddata will be evaluated and efficiently impounded in security prices. From the perspective of market efficiency, it makes littlesenseto argue over, say,which reporting format is preferable; disclosure isviewedas a means for resolving such controversies. Debates regarding alternativeforms of lease accounting, for example,would constitute a trivial issue. Completedisclosure is viewed as the critical issueand where—or in what manner—thedisclosure is effected is unimportant. In a similar vein, increased disclosure would beviewed by some as a means of resolvingtraditional measurement controversies.Reporting historical and current-cost numbers to market participants, for example,eliminates the controversy over inflationaccounting.

The net result of both views is the same.There has emerged a theoretical positioncentering around increased disclosure inlieu of resolution of issues by reliance ontraditionaltheoreticalanalysis. But the appeal of data expansion has been resisted onthree grounds: (1) physiological considerations, (2) cognitive processing limitations, and (3) economic efficiency.

The most obvious criticism of data expansion is physiological. Beyond somepoint, additional data could overwhelm adecision maker and make it physically impossible for him to consider and impoundincremental messages—the familiar "information overload" problem. Of course,this criticism does not apply when the incremental data are of modest proportionsand the user is not inundated with data.

However, even when the proposeddataexpansion is small, an important psychologicaleffect suggests that even modest increases in disclosure are not "costless."Increased disclosure levels increase theperceived complexity of the environment.Abundant psychological testing evidenceindicates that such changes in perceivedenvironmental complexity induce changesin decision makers' cognitive processingcapabilities (e.g., Schroder, Driver, andStreufert, 1967). These cognitive processing changes, in turn, can decrease the effectiveness of decision making by causingdecision makers to revert to a more con

crete conceptual level in an attempt to copewith the new, more complex environment(Driver and Streufert, 1969).

The relevance of the cognitive processing effect to accounting reporting issueswas initially raised at a theoretical level(Revsine, 1970b, 1970c, 1973). Subsequentempirical tests in an accounting setting explored the potentially adverse effects of increasing environmental complexity onvarious kinds of decision styles (Driverand Mock, 1975). Dermer (1973) also examined the effect of personality variables onthis process. From a psychological perspective, results of preliminary researchindicate that data expansion is not a panacea. Beyond a certain point, even relatively modest increases in reported dataraise the possibility of adverse decisionmaking effects.!)

The desirability of a data expansionapproach to external reporting is also unclearwhen the issue is examined from the traditional perspective of the economic consequences of information. Here two deep concerns arise, even if we assume that the reporting alternatives are costless. First,placement—as long as the disclosure ismade—may not be a totally neutral aspectof the problem. With alternative placementoptions, choice of one as opposed to anothermay "signal" some additional economicaspect. For example, footnote versus statement-proper disclosure of a potential liability from pendinglitigationmay "signal"different management beliefs as to thelikelihood that the litigation will be adverse. Similarly, the manner in which forecasts are presented may convey informa-

' This brief summary necessarilyomits certain issues. For example, art information specialists subjectto similar adverse effects of increasing complexity?What little evidence is available suggests that the adverse decision effect arises among both abstract andconcrete decision makers (Streufert and Schroder,1965; Driver and Streufert, 1969; and Streufert, 19701.However, the issues of personality variables (Dermer.1973)and decision style (Driver and Mock, I97S) navealso been explored to determine the generalizabib'tyofthese findings. Furthermore, over time the level ofcomplexity that leads lo adverse effects may be increased as learning takes place (Miller and Gordon,1975). However, the general point remains: the possibility of adverse decision effects cannot be ignored indata expansion proposals.


tion about the quality of the forecasts. Thistype of signalling effect is discussed bySpence(1973) andStiglitz (1975).">

Second, disclosing or producing "more"information even if it is costless may not bebeneficial. In fact, it may be strictly harmful. In a multi-personsetting, such as thatof numerous users of accounting reports, itdoes not necessarily follow that public information is a superior good. Some may beharmed by productionof such information,even if it is costless (see Ho and Chu, 1974;Baiman, 1975; and Demski and Feltham,1976). Perhaps more significantly, in a recently published paper, Hart (1975) demonstrates that in a regime of incomplete markets, openingnewmarkets, say, by providing additional accounting measures onwhich to make event-contingent trades,may not be an economicallydefensibleact.More information, in other words, even iffree, need not be beneficial in a multi-person setting.

In summary, the basic tenet of the dataexpansion approach—more information isassuredly preferable to less—does not apply in either a general economicsetting or

10 Further note that a firm may have elected tohave its statements audited before specific requirements were imposed simply because such behaviorwasa signal of the quality of the statements and the integrity of the management.

39

from a cognitive information processingperspective. Accordingly, theories thatrely on data expansion have not achievedwidespread acceptance.

SummaryIn this chapter, we have reviewed a va

riety of reasons why individual accountants may be dissatisfied with the variousapproaches to developing an accountingtheory. We have not attempted to providean exhaustive enumeration; instead, ourdiscussion illustrates major issues. Although the reader may well have additional points of contention or criticism, thislimited discussion should suggest thebreadth of the problem.

We urge the reader to avoid reading anypartisan intent into our use of the words"criticisms" and "points of contention."We are not assessing specific formulationsof accounting theories. Nor are we defending or criticizing any particular theoreticalapproach. Rather, we have attempted topropound reasons why none of the available theoretical approaches has yielded asufficient and compelling basis for specifying the content of external financial reports. That is, we have examined severaldetailed points of conflict which collectively explain why no theoretical approach hasachieved consensus acceptance.

IF"

Chapter 4

Difficulties in Achieving Consensus: A General View

Accounting researchers, academiccommittees, and professionally sponsoredpolicy groups have attempted, over theyears, to formulate theories of accounting.In Chapter 3, we acknowledged a prevailing expectation among accountants thatthose efforts would lead to some sort of unified theory that would provide a sufficientand compelling basis for specifying thecontent of external financial reports. Wethen summarized a number of problemsthat may have prevented any of the available theoretical approaches from rising toa position of prominence relative to theothers.

Yet, even if each separate approach isperceived to contain serious flaws, members of the accounting community mightstill expect that a process of argument andcounterargument would eventually promote theoretical closure. That is, manymight consider the process of theoreticaldevelopment in accounting to be a type ofevolution. If that be the case, we need onlyto continue the collective struggle to adaptand modify our theoretical structures asenvironmental changes occur.

From an evolutionary perspective, however, specific formulations of accountingtheory will continually face the prospect ofdiminishing relevance unless piecemealmodifications are undertaken to adapt thetheory formulation to environmentalchanges. Furthermore, if environmentalchanges are frequent, the existing theoretical structure is bound to be under constantpressure; that is, the implications of thetheory may not be judged as compellingreasons to constrain or specify the contentof external reports.

The evolutionary view of accountingtheory formation has considerableappeal.It obviously allows for the existence of important, unresolved issues. Furthermore,it also holds the optimistic promise ofmovement toward resolution of these issues.

In spite of these features, the committeeis led to consider an alternative view. Our

reason is that the accounting literature ofthe past decade or two appears inconsistent with the evolutionary view of accounting theory development. Consider these examples: (1) the apparent consensus on the"matching and attaching" approach totheory formation is disintegrating; (2) issues that are either irrelevant or unresolv-able under that approach continue to be recycled; and (3) there is an expanded arrayof alternative approaches to theory formation, none of which looms as an apparentsuccessor to the cost-based approach ofmatching and attaching. All these factorssuggest that changes in the process of theorizing in accounting may be more revolutionary than evolutionary. Such a perspective was developed as a general view of science by Thomas S. Kuhn (1970) .>

Achieving Paradigm AcceptanceIn his treatise on the pattern of changes

in scientific thought, Kuhn introduces theconcept of "paradigms," which he contex-tually defines as conceptual and instrumental frameworks that "provide modelsfrom which spring particular coherent tra-

) We are sensitive to the potential validity of criticisms that analyses (such as Kuhn's) of scientific practices and methodologies probably were not intended toapply and, in fact, may not be applicable to such diverse areas of intellectual activity as physical sciences, social sciences, and accounting. For example,we note Mark Blaug's conclusion that "both Kuhn andLakatos jeer at modern psychology and sociology aspre-paradigmatic, prolo-scienccs. and although economics seems to be exempted from the charge, Lakatosseems to think that even economists have never seriously committed themselves to the principle of falsifi-ability" (1976). No doubt the vast bulk of accountants'work falls victim to the charge levied upon psychologists, sociologists, and economists. Thus, we must recognize the possibility that Kuhn's views on the development of science may apply less forcefully to accountingthan would be appropriate for other areas of intellectual endeavor which have richer scientific traditions.Nevertheless, we note striking similarities in the general objectives of science and accounting, and our searchfor rationales that apply to the existing stage of accounting thought suggests significant insights to begained from the comparisons drawn in this chapter.

41


ditions of scientific research" (p. 10). Theparadigm of an individual scientist specifies the kinds of problems he perceives tobe interesting. It circumscribes the empirical domain over which the individual's theories and research are applied. It indicatesthe kinds of tests or standards that areused to adjudicate contradictory theoretical propositions. And, after a time of research and theorizing, if a paradigm isshared by the scientific community, itleads the research and theorizing processes to a single, prevailing theory of thearea. In other words, a paradigm constitutes a kind of world view and focus for research. Kuhn observes that "men whoseresearch is based on shared paradigms arecommitted to the same rules and standardsforscientific practice" (p. 11).

The current accounting scene includes awide diversity of issues that are deemed tobe important by individual researchers; italso includes a wide diversity of researchmethods employed to address these issues.Such diversity is pervasive. Some authors,for example, consider the issues raised byothers to be trivial and undeserving of attention. Similarly, debate on the relativemerits and weaknesses of alternative research methods provokes a constantstream of argumentative articles, symposia, and speeches. In terms of Kuhn's description, this state of affairs would suggest that accounting theorists do not have ashared paradigm. Rather, we tend to havedifferent perspectives of the world, perceive different issues to be of significance,appeal to different sets of empirical phenomena in searching for answers to theseissues, and accept different tests or standards forresolving theseissues. Further, ifone subscribes to the Kuhnian view, it ishighly unlikely that consensus on thesematters will occur by a cumulative processof scientific inquiry.2

2Wewould emphasize theintention of thischapterlobringadditional perspective to the arrayof theoretical conflict in accounting; the chapter presents onepossibleview of the factors that collectively representa lack of clear direction and focus within the community of accounting theorists and researchers. Readersshould not infer a committee decision to reject alternative perspectives, such as might be provided by ImreLakatos. Indeed.Blaug's (1976) thoughtfulcomparison

Contrary to the idea that scientificknowledge progresses in a cumulative,evolutionary pattern, Kuhn offers a revolutionary description of scientific progress,which proceeds in the following order:

1. acceptance of a paradigm;2. working within that paradigm by do

ing "normal" science;3. becoming dissatisfied with that para

digm;4. searching for a new paradigm;5. accepting a new paradigm.

The most difficult thing to explain is thereason scientists collectively accept a newparadigm. It seems clear that the reason

of Kuhn'sand Lakatos' alternativeviewsonthe historyof scientific thought is recommended for assistance inassessing the significance of Kuhn's arguments, as wellas forsuggesting some of the potential shortcomings inherent in the present application of Kuhn's work. Seealso Lakatos and Musgrave (1970).

In addiiion, this chapter is not intended to infer thatthe existing state of theoretical conflict in accountingmust necessarily be understood in terms of a philosophical perspective of the history of science. Oilier plausible views arc possible.Oneexample of an alternativeview arises from an economicinterpretation, wherebywe (1) assume rational behavior, (2) recognize sufficient externalities to yield some kind of central policymaking as an efficient means of establishingaccounting practices,and then (3) perceivethe alternativeapproaches to developing accounting theories not as"paradigms" or as "scientific theories" but rather asindividual or group proposals for specifying how accounting ought to be done, or specifying how accountants ought to go about deciding how accounting will bedone. In this situation, we observethat the problem ofselectingthe best accountingpoliciescannotbe set upin a cost-benefit settingbecause the requisite conceptof social preference is missing (a la Arrow's theorem).Also, whatever policies will finally be chosen will depend upon the institution orstructureby which thepoliciesarechosen.Even if the decisionis to selectthe policies by voting,the outcomewill dependuponIhevotingsystem we impose.

From this economic perspective, disagreement ensues because we have heterogeneous opinions andtastes, and there is no neat, defensible method of putting our conflicting tastes and beliefs into a grand social function.Thus, the problem is by nature oneofconfrontation, and there is no generally accepted way ofcoming to agreement. Different "theories" or proposals lead to different resourceallocations, andwe generally have conflicting views over which resource allocations are better, and there is no theory to guide us incounterbalancing conflicting opinions. So it is understandable that the current state of "theoretical" conflict in accounting exists and does not appear to be resolvable at the present time.

Difficulties in Achieving Consensus: A General View 43

they reject an existing paradigm is because anomalies arise, which cause scientists to become dissatisfied with that paradigm. This dissatisfaction motivates a subset of the scientists to search for a new paradigm while other scientists continue to tryto work within the older paradigm and toeither rationalize or ignore the anomalies.

What is happening in accounting todaycan be described in similar terms. Thereare a number of theorists who have become dissatisfied with the old matching-at-taching approach to specifying the contentof financial reports. In its place, they haveturned to several alternative approaches toanswering accounting questions. Depending upon the level of generalization atwhich one might choose to apply this view,the decision-usefulness approach and theeconomic approach to analyzing accounting information issues might each betreated as an alternative paradigm. At another level, one might treat each of the alternative valuation schemes as reflectingalternative approaches to resolving accounting questions.

Not all of these alternative paradigmsare necessarily advanced as a sufficientbasis for making accounting policy decisions; often proponents of a particular approach readily acknowledge numerous uncertainties and unsettled implementationissues. Indeed, some proponents of an alternative may caution against the suggestion that the alternative constitutes an approach to theory formation in accountingat all. In these cases it would appear thatan alternative is intended only to bring interesting evidence to bear upon the matterof making policy decisions. Nevertheless,each of the accounting approaches currently advocatedinvolvesa unique way of looking at the accounting problem at hand,whatever it may happen to be. Indeed,eachapproachtends toself-select the problems with which it will deal as well as themeans it will employ to attack these problems. In short, each approach begins totake on the attributes of a distinctive paradigm. At the same time that many theorists and researchers are working out theaccounting implications of these new paradigms, others are unwilling to accept anyof the new paradigms. Rather they contin

ue to try to solve problems within the context of the old paradigms.

In this state of dissatisfaction with existing paradigms we can note that eachtheorist attempts to provide his own foundation for the field. In regard to Newton'stheory of optics, Kuhn writes :

Being able to take no common body ofbelief for granted, each writer on physical optics felt forced to build his fieldanew from its foundations. In doing so,his choice of supporting observation andexperiment was relatively free, forthere was no standard set of methods orof phenomena that every optical writerfelt forced to employ and explain. Underthese circumstances, the dialogue of theresulting books was often directed asmuch to the members of other schools asit was to nature. That pattern is not unfamiliar in a number of creative fieldstoday, nor is it incompatible with significant discovery and invention, (p.13)

This seems to be an apt description of whatis happening in accounting at the presenttime. Many theorists seem to feel the needto start from some basic foundations tobuild the field of accounting anew. For example, a recent managerial accountingstudy started with very fundamental kindsof observations and then attempted to reconstruct the discipline on the basis ofthose foundations (Demski and Feltham,1976). Theorizing from efficient marketsresearch has proceeded in a similar vein(Fama, 1970). Valuation theorists havealso started with very elemental or fundamental assumptions. For example, Chambers drew heavily on a large number ofother disciplines in order to derive his particular theory of valuation (1986).

One of the few attitudes that is sharedby most contemporary accounting theorists is dissatisfaction with the prevailingmatching-attaching paradigm. They sharelittle else in regard to problems that oughtto be solved or methods that ought to beemployed. Indeed, different individualsmay not even agree on what constitutes a"fact" in the absence of a shared paradigm(Sterling, 1970b). Therefore, since different theorists may each employ differentimplicit paradigms, they may simultaneously each perceive different facts. As


Kuhn makesclear, thisis not uniqueto accounting:

Excluding those fields, like mathematics and astronomy, in which the firstfirm paradigms date from prehistoryand also those, like biochemistry, thatarose by division and recombination ofspecialties already matured, the situations outlined aboveare historically typical Inparts ofbiology—the studyof heredity, for example—the first universally received paradigms are stillmore recent; and it remains an openquestion what parts of social scienceshave yet acquired such paradigms atall. History suggests that the road to afirm research consensus is extraordinarily arduous, (p. 15)

Obviously, the road to a research consensus in accounting has beenextraordinarilyarduous as is evidenced by the fact that wehave not achieved that consensus. Kuhncontinues:

History also suggests, however, somereasons for the difficulties encounteredon that road. In the absence of a paradigm or some candidate for paradigm,all of the facts that could possibly pertain to the development of a given science are likely to seem equally relevant.As a result, early fact-gathering is a farmore nearly random activity than theone that subsequent scientific development makes, (p. 15)

Thus, in the absence of an agreed-uponparadigm, thereis noconsensus regardingwhich environmental characteristics areespecially important as a basis for theorydevelopment. All facts appear to be equally relevant. Some theorists will then suggest a particular paradigm which, amongother things, specifies what are facts andwhat are nonfacts. The problem is that oneparadigm will lihely specify a different setoffacts from thatspecifiedby a competingparadigm.

This is the process that results in disjoint paradigms and, in turn, is the reasonthat oneparadigm cannot be used to judgeorassess another. Kuhnstates it directly:

We have already seen several reasonswhy the proponents of competing paradigms must fail to make complete contact with each other's viewpoints. Col

lectively these reasons have been described as the incommensurability ofthe pre- and postrevolutionary normal-scientific traditions, and we need onlyrecapitulate them briefly here. In thefirst place, the proponents of competingparadigms will often disagree about thelists of problems that any candidate forparadigm must resolve, (p. 148)

He goes on to point out that:Communication across the revolutionary divide [of different paradigms] isinevitably partial, (p. 149)

The proponents of competing paradigmsalmost inevitably find their argumentativediscourse to be fraught with communication failures.That is, those who employ different paradigms find it difficult to communicate with one another. Numerous examples can be observed in accounting debates.Consider a discussion of accountingfor special purpose manufacturing equipment that has no resale value, where theparadigm of one discussant has led him toadopt a current exit price theory of accountingvaluationand the paradigmof theother discussant has led him to adopt a replacement cost theory of accounting valuation. In proposing the balance sheetamount to be reported, the two discussantswould likely disagree. Further, in attempting to resolve the dispute, the two discussants would likely appeal to different facts.The exit price accountant would look to thezero resale value of the equipment, whilethe current entry price accountant wouldgenerally look to the replacement cost ofthe equipment. In the usual case, the exitprice would be an irrelevant fact to theentry price accountant, and to the exitprice accountant the entry price wouldusually be irrelevant to the investigation.Thus, the divergent paradigms employedby each partylead todifferent perceptionsof what "facts" should be observed to resolve the problem. After a while, eachparty to the discussion may conclude thatthe other simply doesn't understand theproblem, or doesn't understand the role ofaccounting. What one sees as relevant"facts," the other may reject as totally irrelevant to the situation.

Kuhn continues with:... the third and most fundamental as-


pect of the incommensurability of competing paradigms. In a sense that I amunable to explicate further, the proponents of competing paradigms practicetheir trades in different worlds ....Both are looking at the world, and whatthey look at has not changed but in someareas they see different things, and theysee them in different relations one to theother. That is why a law that cannoteven be demonstrated to one group ofscientists may occasionally seem intuitively obvious to another, (p 150)

This type of disagreement is evident in accounting also. Some things that appear tobe too obvious to require explicit statementby one group are seen by another group asunfounded assertions that are unacceptable unless supported by evidence and/ordeductive reasoning. Kuhn argues, however, that these different views cannot bereconciled by either logic or empirics. Instead, they are revolutionary shifts inworld views:

Just because it is a transition betweenincommensurables, the transition between competing paradigms cannot bemade a step at a time, forced by logicand neutral experience. Like the gestaltswitch, it must occur all at once (thoughnot necessarily in an instant) or not atall. (p. 150)

Thus, paradigm acceptance is like a gestalt switch, a change in world views. Thatis, it is not possible to prove that one paradigm is superior to another; instead a scientist either sees the new paradigm or hedoes not. Butterfield (1957) summarizedthe point in describing a switch from oneparadigm to another in the physical sciences. He said that the change was not inthe facts observed, but instead it was "atransposition in the minds of the scientists" (p. 17).

Just as Kuhn could not explicate whatprecise factors finally precipitate generalacceptance of a specific paradigm, thiscommittee is unable to identify the factorsthat would lead to such a consensus in accounting. It may be another instance of the"aha" phenomenon. That is, some scientists may look at a paradigm, make a gestalt switch, and, in effect, say, "Aha, I understand that better now." It is a psycho

logical question of great magnitude. Forexample, the paradigm of someaccountingtheorists results in an information economics approach to analyzing accounting issues, which is accepted perhaps becausethe theorists believe that it offers greatpromise for the resolution of accountingproblems. Other theorists have not madethis particular gestalt switch and resist theinformation economics paradigm. They accept other, equally defensible paradigms(such as the decision model approach) andbelieve that their paradigms offer morepromise for resolving accounting issues.The phenomenon is apparently psychological, as suggested by the fact that some ofour greatest scientists have continued toresist particular paradigms even afterthey had become well accepted by the majority of the scientific community. Even ifthe accounting community eventually develops a consensus on a single paradigm,we will not be able to criticize those thatwere discarded:

Still more men, convinced of the newview's fruitfulness, will adopt the newmode of practicing normal science, untilat last only a few elderly holdouts remain. And even they, we cannot say, arewrong. (Kuhn,p. 159)The revolutionary perspective of sci

ence is applied to accounting as a plausiblemeans of interpreting the current conditionof accounting theory. There are a numberof people offering different paradigms. Afew have been converted to each of the paradigms but there is no consensus at thepresent time. Furthermore, it would appear as though there is no quick way to achieve a consensus among competing paradigms. Exhaustive research efforts, fiatpronouncements, and exhortations will notnecessarily work. Perhaps one day soon allothers will say "aha" and accept one of thecompeting paradigms and begin to do thepuzzle-solving and mopping-up operationsof normal science. Note, however, that thisis a description of the process, not an explanation. Scientists must admit that atpresent the process of paradigm selectionis not well understood. The process can bedescribed, but describing it does not assistus in the actual selection of a particularparadigm.


Empirical and Logical Testing ofCompeting Paradigms

The traditional scientific response to aproblem of competing alternatives is to examine the logical consistency and empirical implications of each alternative. Competing paradigms should lead to competingtheories. Following the logical positivisttradition, the prerequisite criteria for acceptance of a theory are that it be logicallyconsistent and empirically supportable.3 Ifthese criteria can be applied to select between competing theories, perhaps theycan also serve, by inference, to select between the competing paradigms fromwhich the respective theories were derived.

There are, however, fundamental flawsin this approach, some of which are rathertechnical. One easily understandable flawis that all theories are incomplete, in thesense that additional research is always required. The research process in science isone of continuing to make logical connections or empirical tests. Thus, no theorycan pass those two tests completely, because the testing processes themselves arenever complete. Even within the context ola single paradigm, therefore, the testing ofcompeting propositions and theories maynot provide a permanent basis for selectionbetween the alternatives.

A second, and more damaging, flaw inthe approach of using logical and empiricaltests to select between competing theoriesbecomes apparent if the competing theories issue from competing paradigms.Kuhn explains the problem in the contextof discussing the reasons for chemists' acceptance of Dalton's paradigm:

... it is hard to make nature fit a paradigm. That is why the puzzles of normalscience are so challenging and also whymeasurements undertaken without aparadigm so seldom lead to any conclusions at all. Chemists could not, therefore, simply accept Dalton's theory onthe evidence, for much of that was stillnegative. Instead, even after acceptingthe theory, they had still to beat nature

3 Other criteria such as simplicity and elegancecould be introduced but these are clearly subordinate tological and empirical testability.

into line, a process which, in the event,took almost another generation. When itwas done, even the percentage composition of well-known compounds was different. The data themselves hadchanged, (p. 135)

Thus, at the time a theory is proposed, itmay not pass the typical empirical tests;the evidence may still be negative. Insteadof subjecting the theory to empirical tests,researchers may have to "beat nature intoline" by developing new tests within thecontext of the new paradigm. The acceptance of Dalton's paradigm involved a gestalt switch that changed the very perceptions of the scientists who were doing theresearch. It is this new world view whichallows the scientists to change the data.And these new, changed data are necessary if the new theory is to pass the empirical tests. A recent illustration of this process is the testing of the accounting implications of efficient securities markets. Inorder to assess the information content anddisclosure implications of that theory, analytic devices such as the API metric andother tests related to the capital asset pricing model had to be developed. Withoutthese devices, the theory's implications foraccounting were untestable. Thus, theremust be at least some threshold acceptance of a new paradigm before tests of thetheory arising from the paradigm arepossible.

The point is that, although theoriesmust pass tests of logical consistency andempirical predictability, one cannot expectthese tests to provide a conclusive basis forselecting from among competing paradigms. Ultimately, the design of the testsas well as the perception of the data towhich they apply are dependent upon, ordefined in terms of, the paradigm that is tobe tested. Logic and empirics do not, therefore, provide a sufficient basis for selectingbetween competing paradigms.

Judgmental Selection of a ParadigmAlthough scientists cannot explain the

precise reasons for the acceptance of different paradigms, it is possible to describecertain types of differences among paradigms. One set of such differences relatesto the paradigms' empirical domains.


Each of the currently competing accounting paradigms tends to specify a differentempirical domain over which an accounting theory ought to apply.

For example, one paradigm, whichcould be labeled the "anthropological approach," specifies the professional practices of accountants as the empirical domain of accounting. Following this paradigm, accounting theory is formulated as arationalization of, and by drawing inferences from, extant accounting practices.Another paradigm rests largely upon thebehavior of stock markets to provide theempirical domain over which accountingtheory is constructed and applied. Still another general view of accounting specifiesthe decision processes of individualsand/or extant decision theories as the empirical domain of accounting theory.4 Thistripartite categorization can be further expanded to incorporate both the ideal income approach and the information economics approach, each of which suggests asomewhat unique empirical domain of accounting. It was precisely these differences in empirical domain that providedthe organizational format of Chapters 1and 2.

As we have already suggested, the differences in empirical domain that pervadeeach paradigm reflect basic differences inworld views. Each paradigm has differentembedded premises and objectives; accordingly, each rather naturally focuses onprecisely those empirical phenomena thatare highlighted by those premises and arerelevant or controllable in achieving thedesired outcomes. When viewed from thisperspective, it is apparent that no view canbe per se "correct" or "incorrect." Instead, each must be evaluated in conjunction with contextual factors and personalpreferences.

Because the choice among paradigmsfrom this perspective must necessarily bejudgmental, proponents of different paradigms frequently make basic value judgments (either explicit or implicit) and thenargue that their paradigm, on the basis of

those value judgments, is superior to competing ones. Kuhn argues that this is theusual procedure in science:

Like the choice between competing political institutions, that between competing paradigms proves to be a choice between incompatible modes of community life. Because it has that character,the choice is not and cannot be determined merely by the evaluative procedures characteristic of normal science,for these depend in part upon a particular paradigm, and that paradigm is atissue. When paradigms enter, as theymust, into a debate about paradigmchoice, their role is necessarily circular.Each group uses its own paradigm toargue in that paradigm's defense.

The resulting circularity does not, ofcourse, make the arguments wrong oreven ineffectual. The man who premisesa paradigm when arguing in its defensecan nonetheless provide a clear exhibitof what scientific practice will be like forthose who adopt the new view of nature.That exhibit can be immensely persuasive, often compellingly so. Yet, whatever its force, the status of the circularargument is only that of persuasion. Itcannot be made logically or even probabilistically compelling for those whorefuse to step into the circle, (p. 94)The lengthy debate over alternative ac

counting theories obviously has not led toclosure. Perhaps the extent to which eachtheory's proponents employ circular arguments is most easily discerned by thosewho are aligned with the alternative theories. Kuhn concludes that such circulararguments are inevitable, but not necessarily ineffectual. Individuals can sometimes be persuaded by this kind of debate.Nevertheless, we should not expect that thecontinued debate over accounting theorieswill naturally lead to a dominant consensus.9 Note that we are not in any sense calling for a moratorium on this kind of debate. Rather, we are attempting to articu-

4 For an elaboration of this classification of paradigms by empirical domain, see Sterling (1970b. pp.449-M).

5 For a somewhat different viewon this point, seeWells (1976). That article, which parallels our discussion at several points, did not come to our attentionuntil this document was completed.

3'


late realistic expectations about the consequences of this activity.

SummaryThe purpose of this chapter has been to

develop a plausible explanation for thelack of progress in achieving accountingtheory consensus. An expanding array ofaccounting theories and/or theoretical approaches suggests the existence of severalcompeting paradigms. Each paradigm implicitly incorporates individual beliefs andpremises that cannot be proved or disproved in a logical sense. Further, the

"facts" towards which a theory is directedare themselves a function of the paradigmembraced by the observer. Thus, rigorousexperimental testing, while valuable in thesense of influencing long-run psychologicalsets, cannot be used to resolve the problemof selecting a single paradigm for accounting. As a consequence, defenders of a particular paradigm are forced to rely on persuasion rather than logic and empirics inattempting to defend a proposal. Whileconsensus may eventually develop, thetransformation that occurs is primarily apsychological matter rather than a dispassionately intellectual phenomenon.

Chapter 5

Implications

A committee such as this is formed periodically in order to assess the state of development of accounting theory. Ours wasnot a research assignment; committeescannot efficiently conduct research. Instead, our role was to survey the accounting theory literature. The benefits of thisprocess may be twofold. First, by its verynature, a survey requires generalizationand thus emphasizes important, dominantthemes. This focus eliminates irrelevan-cies and makes primary messages moredistinct. Highlighting important points thatare obscured by the sheer complexity ofthe unclassified environment often generates new perspectives and insights, aswell. Second, a survey may also clarify andunderscore issues that were not explicit inthe original sources. Different language ora more dispassionate, neutral perspectiveis frequently helpful in interpreting andbroadly disseminating somewhat esoteric—and hence potentially obscure—researchfindings. We hope that this report accomplishes these two objectives and therebyprovides a useful perspective regardingaccounting theory approaches.

Implications For ResearchIn addition to these objectives, commit

tee statements of this genre might be expected to have an impact on future researchers and research activities. In thisdomain three major "messages" emergefor researchers in accounting and for thosewho utilize the output of such research.

Perhaps the most fundamental implication of this document is that statementsand pronouncements by "authoritative"bodies—even those comprised of accounting theorists—cannot themselves be expected to lead to near-term closure in thetheory area. Our reason for emphasizingthis point is that it is in marked contrastwith the sentiment that apparently prevails in the accounting community. Manyhave looked with anticipation to past statements of accounting objectives and theoryreports as a means for resolving theoreti

cal debates. Once the awaited report appears, however, there is usually disappointment when it becomes apparent thatthe document does not receive immediateacceptance and thus does not lead to a cessation of debate. Some might then deemthe report to be a "failure." This committee statement will, it is hoped, put the issueof theory closure into proper perspectiveand thereby reduce unrealistic expectations. Our message is clear: theory closurecannot be dictated. Even if based upon rigorous analysis, theories are inevitably derived from the experiential "sets" of thoseproposing them. Such sets (or paradigms)utilized in viewing the world cannot be regarded as uniquely correct. Many divergent sets may simultaneously possess inherent "truth" content. The eventual acceptance of one such approach arises fromits explanatory power or ability to resolveanomalies. In other words, a theory achieves such dominant acceptance becauseof its demonstrated efficacy over time, notbecause of its espousal by some group (forexample, of nine academics) that subjectively deems it to be superior.

A second basic message of this document is that external reporting theory hasa wider scope than that which has beengenerally perceived. In the past, the theoryrealm has been thought to be limited tobasic measurement controversies, natureof users' information desires, cognitiveprocessing characteristics, and other similar issues. The institutional framework bywhich the policy-making process is accomplished has usually been taken as given.However, this committee report emphasizes that the institutional structure andprocedures by which this structure operates are themselves a part of external reporting theory. This perspective immediately introduces questions regarding thedesirability of regulatory intervention andthe form such intervention should take, ifneeded. At present, studies of the socio-political process as it relates to accountingpolicy information, voting strategies,

49


means for eliciting "true" preferences,and similar issues have not been widely applied to external reporting issues. However, as discussed in our report, these tooare a legitimate, albeit neglected, elementof external reporting theory deserving ofresearch attention.

Finally, this report has isolated areas ofdisagreement between theory proposals.Explicitly recognizing basic differencesbetween theories could lead to at least twopotential benefits in future research. First,it could serve to reduce the endless argumentation regarding imperfectly understood or even hidden assumptions and consequences that currently inhibits theoryprogress. Second, recognizing inherent differences among theories could place inbold relief the nature of the private and social welfare tradeoffs that may have to bemade in choosing among policy alternatives.

To illustrate this more concretely, consider the following example. At present,one of the most prevalent theoretical approaches employed in contemporary research is the decision-model approach. Inits standard form, this approach exploresmeans for providing relevant informationto a designated user group. Accordingly,the analytic focus is limited to that groupand explores only the expected consequences of the information for their decision effectiveness. But it was indicated inChapter 3 that this approach ignores thepotential economic impact of the theoryproposal on other users and even nonusersof financial information. In other words,the decision-model analysis is a partialequilibrium approach. This raises the possibility of objections to such proposals onseveral bases. Notwithstanding the benefits accruing to the user group that formedthe basis for the analysis, some may objectto a decision model proposal because of unfavorable repercussions to other societalsegments; that is, the issue may be one ofdistributive welfare. Similarly, the socialcosts incurred in generating "better" information to the selected group may not becommensurate with the benefits expectedto accrue to society. Frequently, the realissues underlying debates of this sort arenot recognized. Irresolvable arguments en

sue because the "benefits" perceived byproponents of a proposal may be irrelevantor inconsequential to its critics.

Prior to this committee report, partialities of this sort may not have been widelyrecognized as inherent in various theoryapproaches. By delineating these scopelimitations, needless argumentation maybe avoided. How might this take place? Wehope that this will follow from general acceptance of the view adopted herein regarding accounting theory. Thus, our thirdbasic message is that all theory approaches are flawed when viewed from theperspective of some alternative approach.This problem cannot be avoided. Hence, itis useless to seek the "universally appropriate" theory approach. At best, one approach can take cognizance of its own limitations and attempt to address omitted issues that are likely to generate disagreement when the results are viewed from theperspective of some alternative theory approach.

Summarizing, one of the primary objectives of this committee document is to promote a deeper understanding of the natureof accounting theories, how they differ, andthe issues that arise in selecting among thevarious alternatives. We concluded thattheory choices are vitally dependent uponthe experiential set and preferences of theobserver. Consequently, rigorous analytictechniques serve only a limited role in thisarea as an initial screening device, that is,as a means for identifying and eliminatingproposals that are logically inconsistent orerroneous. Proposals that survive this analytic screen must gain acceptance on someother basis—their explanatory or predictive power, ability to resolve existinganomalies, or conformity to prevailing social preferences. Clearly, research playsan important role in this post-screeningchoice process too. However, dispassionatelogical analysis is only a necessary, not asufficient, condition for eventual theory acceptance. Seemingly, the crucial role of thesufficient conditions has not been widelyunderstood in the past. But failure to comprehend this element of the theory choiceprocess leads to attempts to impose suchclosure by a process that might be termed"theoretical fiat." Such efforts have histor-

Implications

ically been largely unproductive. Our belief is that a better understanding of the nature of the total theory acceptance processwill reduce expectations for achieving theory closure by fiat and redirect researchers to more promising avenues. Thus, theapproach used in this committee documentwas selected with the expectation that itwill facilitate such research progress. It isto be hoped that this expectation will be fulfilled.

Implications For PolicyWhen considering issues of immediate

and crucial importance, policy makers donot have the luxury of sidestepping difficult, unresolved questions. Life goes on,decisions must be made, and policy choicesoccur. Recognizing these facts, our committee statement has certain implicationsfor policy makers seeking direction fromaccounting theories.

The central message to policy makers isthat until consensus paradigm acceptanceoccurs, the utility of accounting theories inaiding policy decisions is partial. Competing theories merely provide a basis forforming opinions on what must remain inherently subjective judgments. While it istrue that consensus will frequently developon certain points, usually this consensusonly narrows the range of disagreement; itoften does not resolve the basic issue thatgives rise to the underlying problem. Thatis, consensus about peripheral improvements may occasionally emanate from applying a theory, but the basic underlyingchoice issues continue to be disputatious.

A non-accounting example illustratesthis point. Consider the issue of whether ornot capital punishment is justified on a social welfare basis. At this level, numeroustheories abound. Most are inconsistentwith one another and it is impossible to objectively state that one such theory dominates all others. Hence the issue of capitalpunishment is unresolvable by theoreticaldebate alone. On this issue, social action isdependent upon the subjective beliefs andpreferences of policy makers. Various theories may be instrumental in shaping thebeliefs of individual policy makers, but nobasis for objectively demonstrating the superiority of any one theory exists.

51

Notwithstanding the continuing disagreement on the overriding basic issue,certain subordinated issues relating to capital punishment are virtually unanimouslyagreed upon. That is, we are often able toconclude that a particular change was forthe general good. Take, for example, thechange from processing executions bydrawing and quartering the criminal to execution by guillotine, hanging, electricchair, or any of the more modern forms ofexecution. The distributive and aggregative welfare effects of these changes defyrigorous analysis. Yet, we would expectgenera] consensus that this was an example of improvement in professional practice.

In a similar vein, virtually all externalreporting issues involve social welfaretrade-offs. Depending on the particulartopic, theories may occasionally narrowthe range of choice by indicating a dominant consensus regarding some facet of thebasic issue. Forms of execution is a non-accounting illustration; the policy changewhich long ago included income statements in published financial reports is anexample from an accounting setting. Frequently, accounting theories will also beuseful to policy makers by providing an improved basis for reaching what must nowbe considered to be an essentially subjective judgment. That is, theories may behelpful insofar as they apprise policymakers of the underlying issues and clarifythe trade-offs implicit in various approaches. But in the absence of paradigmacceptance, it is unrealistic to expect accounting theory to provide unequivocal policy guidance. Different theories will pointto different policies. These theories arisefrom different paradigms. Since there is norigorous analytic means for choosing between paradigms, there is similarly no rigorous means for choosing between theoriesor their derivative policy implications.

This limited theory role will not necessarily persist in the long run, nor is itunique to accounting.1 Changes will occur

1As an analogy, consider the equivocal guidancethat competing economic theories provide to economicpolicymakers.

!. I

52 Statement onAccountingTheory and Theory Acceptance

if, and when, consensus regarding an accounting paradigm is once again achieved.Until such time, however, expectationsabout the roleof theory in guiding policymust be realistic. In clarifying the role oftheory, one of our objectives is to indicatethe futility of attempting to dictate "ultimate theoretical truths." This theory consensus may or may not be achieved in thenear term. If it is achieved, it will arise be

cause of continued accounting theory research and debate about the product ofsuch research. However, the crucial pointis that theory consensus cannot be imposed, not by this group nor by any other.Whatever future influence theory has onpolicy making will be achieved by continued argumentation, new theory development, and debate, not by fiat.

Bibliography

Abdel-khalik, A. R. (1971). "User Preference Ordering Value: A Model." The AccountingReview 46 (July, 1971)457-71.

Akerlof, G. (1970). "The Market for 'Lemons': Quality Uncertainty and the MarketMechanism." The QuarterlyJournal of Economics 84 (August, 1970) 488-500.

Alexander, S. S. (1950). "Income Measurement in a Dynamic Economy." Five Monographs on Business Income. New York: The Study Group on Business Income, TheAmerican Institute of (Certified Public) Accountants, 1950. Also see: Alexander, S. S.,"Income Measurement in a Dynamic Economy" Revised by D. Solomons. Studies inAccounting Theory. W. T. Baxter and S. Davidson, eds. Homewood, 111.: Richard D. Irwin, Inc., 1962.

American Accounting Association (1936). Executive Committee. "Accounting PrinciplesUnderlying Corporate Financial Statements." The Accounting Review 11 (June, 1936)187-91.

American Accounting Association (1941). Executive Committee. "Accounting PrinciplesUnderlying Corporate Financial Statements." The Accounting Review 16(June, 1941)133-9.

American Accounting Association (1955). Committee on Concepts and Standards Underlying Corporate Financial Statements. "Standards of Disclosure for Published Financial Reports." Supplementary Statement No. 8. The Accounting Review 30(July, 1955)400-4.

American Accounting Association (1957). Committee on Concepts and Standards Underlying Corporate Financial Statements. "Accounting and Reporting Standards for Corporate Financial Statements, 1957 Revision." The Accounting Review 32 (October,1957) 536-546.

American Accounting Association (1966). Committee to Prepare a Statement of Basic Accounting Theory. A Statement of Basic Accounting Theory [ASOBAT]. Evanston, 111.:American Accounting Association, 1966.

American Accounting Association (1969). Committee on External Reporting. An Evaluation of External Reporting Practices — A Report of the 1966-68 Committee on ExternalReporting. The Accounting Review 44 (Supplement, 1969) 79-123.

American Accounting Association (1971a). Report of the Committee on AccountingTheory Construction and Verification. The Accounting Review 46 (Supplement, 1971)53-79.

American Accounting Association (1971b). Report of the Committee on BehavioralScience Content of the Accounting Curriculum. The Accounting Review 46 (Supplement, 1971) 247-85.

American Accounting Association (1972). Report of the Committee on Research Methodology in Accounting. The Accounting Review 47(Supplement, 1972) 399-520.

American Accounting Association (1974). Report of the Committee on Concepts and Standards — Internal Planning and Control. The Accounting Review 49(Supplement, 1974)79-96.

American Institute of (Certified Public) Accountants (1953). Committee on Terminology."Review and Resume." Accounting Terminology Bulletin No. I New York: AmericanInstitute of (Certified Public) Accountants, 1953.

American Institute of Certified Public Accountants (1970). Accounting Principles Board.Basic Concepts and Accounting Principles Underlying Financial Statements of Business Enterprises. Statement No. 4 of the Accounting Principles Board, 1970.

American Institute of Certified Public Accountants (1971). Accounting Principles Board."Interest on Receivables and Payables." APB Opinion No. 21. New York: AmericanInstitute of Certified Public Accountants, 1971.

53

'••:


American Institute of Certified Public Accountants (1973). Study Group on the Objectivesof Financial Statements. Objectives of Financial Statements. New York: American Institute of Certified Public Accountants, 1973.

Archibald, T. R. (1972). "Stock Market Reaction to the Depreciation Switch-Back." TheAccounting Review 47 (.January. 11)72) 22-30.

Arrow, K. J. (1962). "Economic Welfare and the Allocation of Resources for Invention."In The Rate and Direction of Economic Activity: Economic and Social Factors.Princeton, N.J.: National Bureau of Economic Research. Princeton University Press,1962,

Arrow, K. J. (1903). Social Choice and Individual Values. Revised Edition. New York:Yale University Press, 1963.

Arrow, K. J. (1964). "The Hole of Securities in the Optimal Allocation of Risk-Bearing."Review of EconomicStudies 31 (April, 1964) 91-6.

Arrow, K. J., and F. H. llahn (1971). General Competitive Analysis. San Francisco:Holden-Day. Inc., 1971.

Baiman. S. (1975). "The Evaluation and Choice of Internal Information Systems Within aMultiperson World." Journal of Accounting Research 13 (Spring. 1975)1-15,

Baker, H. K.. and J. A. Hasten (197.3). "Information Needs of Individual Investors." TheJournal of Accountancy 136 (November, 1973) 64-9.

Ball. K. J. (1972). "Changes in Accounting Techniques and Stock Prices." Journal of Accounting Research 10 (Supplement, 1972) 1-40.

Ball, K. J., and P. Brown (19611). "An Empirical Evaluation of Accounting Income Numbers." Journal of Accounting Research 6 (Autumn. 1968) 159-77.

Ball. R. .!., B. Lev, and R. Watts (1976). "Income Variation and Balance Sheet Compositions." Journal ofAccounting Research 14 (Spring, 197G) 1-9.

Ball, R,.)., and R. Watts (1!)72). "Some Time Series Properties of Accounting Income."The Journal ofFinance 27 (June, 1972) 663-81.

Baskin, E. F. (1972). "The Communicative Effectiveness of Consistency Exceptions."The Accounting Review47 (January. 1972) 38-51.

Beaver, W. H. (1968). "The Information Content of Annual Earnings Announcements."Journal of Accounting Reseurch 6 (Supplement. li)68) 67-92.

Beaver, W. H. (1970). "The Time Series Behavior of Earnings." Journal of AccountingResearch 8 (Supplement. 1970) 62-99.

Beaver. W. H. (1972). "The Behavior of Security Prices and Its Implications for Accounting Research (Methods)." Chapter II. Report of the Committee on ResearchMethodology in Accounting [AAA, 19721. The Accounting Review 47 (Supplement,1972)407-37.

Beaver, W. II.. and J.S. Demski (1974). "The Nature of Financial Accounting Objectives:A Summary and Synthesis." Journal of Accounting Research 12 (Supplement, 1974)170-187,

Beaver, W. II., and R. E. Dukes (1972). "Interperiod Tax Allocation, Earnings Expectations, and the Behavior of Security Prices." The Accounting Review -17 (April. 1972)320-32,

Beaver, W. II., and R. E. Dukes (1973). "Interperiod Tax Allocation and d-DepreciationMethods: Some Empirical Results." The Accounting Review 45(July. 1973) 549-59.

Beaver, W. H.. J. W. Kenneliy. and W. M. Voss (1968). "Predictive Ability as a Criterionfor the Evaluation of Accounting Data." The Accounting Review 43 (October, 1968)675-83.

Beaver, W. H., P. Kettler, and M. Scholes (1970). "The Association Between MarketDetermined and Accounting Determined Risk Measures." 7'he Accounting Review 45(October. 1970)654-82.

Benston, G. J. (1973). "Required Disclosure and the Stock Market: An Evaluation of theSecurities Exchange Act of 1934." The. American Economic Review 63 (March, 1973)132-55.

Bibliography 55

Bildersee. J.S.I 1975). "The Association Between a Market-Determined Measure of Riskand Alternative Measures of Risk." The Accounting Review 50 (January. 1975) 81-98.

Birnberg, J. G., and R. Nath (1967). "Implications of Behavioral Science for ManagerialAccounting." The Accounting Review 42 (July, 1967)468-79.

Blaug, M. (1976). "Kuhn versus Lakatos, or Paradigms versus Research Programmes inthe History of Economics." History of Political Economy 8 (January, 1976) 399-433.

Boatsman, J. R., and J. C. Robertson (1974). "Policy-Capturing on Selected MaterialityJudgments." The Accounting Review 49(April, 1974)342-52.

Bradish, R. D. (1965). "Corporate Reporting and the Financial Analyst." TheAccountingReview40 (October, 1965) 757-66.

Brenner, V.. and R. Shuey (1972). "An Empirical Study of Support for APB Opinion No.16." Journal of Accounting Research 10 (Spring, 1972) 200-8.

Brown, P., and R. Ball (1967). "Some Preliminary Findings on the Association Betweenthe Earnings of a Firm, Its Industry and the Economy." Journal of Accounting Research 5 (Supplement. 1967) 55-77.

Butterfield, II. (1957). The Origins of Modern Science 1300-1800. Revised Edition. NewYork: The Free Press, 1957.

Buzby, S. L. (1974). "Selected Items of Information and Their Disclosure in AnnualReports." The Accounting Review 49 (July, 1974)423-35.

Canning, J. B. (1929). The Economics of Accountancy. New York: The Ronald PressCompany. 1929.

Carsbcrg, B., J. Arnold and A. Hope (1977). "Predictive Value: A Criterion for Choice ofAccounting Method" in W. T. Baxter and Sidney Davidson, eds. Studies in AccountingTheory. 3rd Edition. Homewood, III.: Richard D. Irwin, Inc., 1977(forthcoming).

Cerf, A. R. (1961). Corporate Reporting and Investment Decisions. Berkeley: Universityof California Press, 1961.

Chambers, R. J. (1955). "Blueprint for a Theory of Accounting." Accounting Research 6(January, 1955) 17-25.

Chambers, K. J. (1965). "The Development of Accounting Theory" in R. J. Chambers, L.Goldberg, and R. L. Mathews, eds. The Accounting Frontier. Melbourne: Cheshire,1965, pp. 18-35.

Chambers, R. J. (1966). Accounting. Evaluation and Economic Behavior. EnglewoodCliffs: Prentice-Hall. Inc.. 1966.

Chandra, G. (1974). "A Study of the Consensus on Disclosure Among Public Accountantsand Security Analysts." The Accounting Review 49 (October, 1974)733-42.

Clark. J. M. (192.'!). Studies in the Economics of Overhead Costs. Chicago. 111.: Universityof Chicago Press, 1923.

Copeland, R. M., A. J. Francia. and R. H. Strawser (1973). "Students as Subjects inBehavioral Business Research." The Accounting Review 48 (April, 1973)365-71.

Debreu, G. (1959). Theory of Value. Cowles Foundation Monograph 17. New York: JohnWiley &Sons, Inc., 1959.

Demsetz, II. (1969). "Information and Efficiency: Another Viewpoint." The Journal ofLaw and Economics 12 (April, 1969) 1-22.

Demski, J. S. (1972). InformationAnalysis. Reading, Mass.: Addison-Wcsley PublishingCo., 1972.

Demski, J. S. (1973a). "The General Impossibility of Normative Accounting Standards."The Accounting Review48 (October. 1973)718-23.

Demski, J. S. (1973b). "Rational Choice of Accounting Method for a Class of Partnerships." Journal of Accounting Research 11 (Autumn, 1973) 176-90.

Demski. J. S., and G. A. Feltham (1976). Cost Determination: A Conceptual Approach.Ames. Iowa: The Iowa State University Press. 1976.

Dermer, J. D. (1973). "Cognitive Characteristics and the Perceived Importance of Information." The Accounting Review 48 (July, 1973) 511-19.

I

StatementonAccounting Theoryand TheoryAcceptance

Dcvine, C. T. (1960). "Research Methodology and Accounting Theory Formation " TheAccounting Review 35 (July, 1960)387-99.

Dickhaut J. W., and I.R. C. Eggleton (1975). "An Examination oftheProcesses Underlying Comparative Judgements of Numerical Stimuli." Journal of Accounting Research 13 (Spring, 1975)38-72.

Driver, M. J., and T. J. Mock (1975). "Human Information Processing, Decision StyleTheory, and Accounting Information Systems." The Accounting Review 50 (Julv.19(0) 490-308.

Driver M J., and S. Streufert (1969). "Integrative Complexity: An Approach to Individuals and Groups as Information-Processing Systems." Administrative ScienceQuarterly 14 (June, 1969) 272-85.

Due, J. F., and A. F. Friedlaender (1973). Government Finance: Economics of the PublicSector. Homewood, III.: Richard D. Irwin, Inc., 1973.

Dyckman, T. R„ M. Gibbins. and R. J.Swieringa (1975). "Experimental and Survev Research in Financial Accounting: AReview and Evaluation." The Impact of Accounting Research in Financial Accounting and Disclosure on Accounting PracticeDurham, N. C: Duke University, Unpublished Proceedings, 1975.

Dyer. J. (1973). "A Search for Objective Materiality Norms in Accounting and Auditing "Unpublished D.B.A. Dissertation. University ofKentucky, 1973.

Ecton, W. W. (1969). "Communication Through Accounting - Bankers' Views " TheJournal ofAccountancy 128 (August, 1969) 79-81

Edwards E. 0. and P.W. Bell (1961). The Theory and Measurement of Business Income.Berkeley,Calif.: UniversityofCaliforniaPress, 1961

Falk, II., and T.Ophir (1973a). "The Effect ofRisk on theUse ofFinancial Statements bv."~3 mcnl Decls">nMakers: ACase Study." The Accounting Review 48 (April, 1973')

Falk. H., and T. Ophir (1973b). "The Influence of Differences in Accounting Policies onInvestment Decisions." Journal ofAccounting Research 11 (Spring, 1973) 108-16,ma,wB,; (1970)- "Efficicnl Capital Markets: AReview of Theory and EmpiricalWork. Ihe Journalof Finance25 (Muy,1970)383-417.

Fama, E F L. Fisher, M. C. Jensen, and R. Roll (1969), "The Adjustment of StockJ-TicestoiNew Information." International Economic Review 10 (February 1969) 1-'>1

Fama, E. F.. and A. B. Laffer (1971). "Information and Capital Markets." The Journal ofBus mess 44 (July. 1971)289-98.

Feltham, G. A. (1972). Information Evaluation. Studies in Accounting Research #5.American AccountingAssociation,1972.

Fetter. F. A. (1927) "Interest Theory and Price Movements." The American EconomicReview 1/ (Supplement, 1927) 62-105.

Fisher. I. (1906). The Nature of Capital and Income. New York: The Macmillan Company. 1906.

Fisher, I. (1927). The Income Concept in the Light of Experience. New Haven' YaleUniversity Press, 1927.

Fisher. I. (1927). "Interest Theory and Price Movements - Discussion." The AmericanEconomic Review 17(Supplement, 1927) 106-8.

Friedman. M. (1953). "The Methodology of Positive Economics." Essavs in PositiveEconomics. Chicago: University ofChicagoPress, 1953.

Gilman. S. (1939). Accounting Concepts of Profit. New York: The Ronald Press Company, 1939.

Godwin, L. B. (1975). "CPA and User Opinions on Increased Corporate Disclosure " TheCPA Journal 45 (July, 1975) 31-5.

Gonedes, N. J. (1972). "Efficient Capital Markets and External Accounting." The Accounting Review47 (January, 1972) 11-21.

Gonedes, N, J. (1973). "Evidence on the Information Content ofAccounting Numbers:

Bibliography 57

Accounting-Based and Market-Based Estimates of Systematic Risk." Journal ofFinancial and Quantitative Analysis 8 (June, 1973)407-43.Gonedes, N. J. (1974). "Capital Market Equilibrium and Annual Accounting Numbers:

Empirical Evidence." Journalof AccountingResearch 12 (Spring, 1974) 26-62.Gonedes, N. J. (1975). "Information Production and Capital Market Equilibrium." The

Journal of Finance 30 (June, 1975)841-64.Gonedes, N. J., and N. Dopuch (1974). "Capital Market Equilibrium, Information

Production, and Selecting Accounting Techniques: Theoretical Framework andReview of Empirical Work." Journal of Accounting Research 12 (Supplement, 1974)48-129.

Gonedes, N. J., N. Dopuch. and S. Penman (1975). "Disclosure Rules, InformationProduction and Capital Market Equilibrium: The Case of Disclosure Rules." Unpublished Working Paper, Graduate School of Business, University of Chicago, June,1975.

Grady, P. (1965). An Inventory of Generally Accepted Accounting Principles for BusinessEnterprises. Accounting Research Study No. 7. New York: American Institute of Certified Public Accountants, 1965.

Grossman, S. J., and J. E. Stiglitz (1976). "Information and Competitive Price Systems."The American Economic Review 66 (May, 1976) 246-53.

Haley, C. W., and L. D. Schall (1973). The Theory of Financial Decisions. New York:McGraw-Hill Book Company. 1973.

Haried, A. A. (1972). "The Semantic Dimensions of Financial Statements." Journal ofAccounting Research 10 (Autumn, 1972) 376-91.

Haried, A. A. (1973). "Measurement of Meaning in Financial Reports." Journal of Accounting Research 11 (Spring, 1973) 117-45.

Hart, O.D. (1975). "On the Optimality of Equilibrium When Markets are Incomplete."Journal of Economic Theory II (December, 1975)418-43.

Hatfield, II. R. (1927). Accounting. New York: D. Appleton & Company, 1927.Hayami, Y., and W. Peterson (1972). "Social Returns to Public Information Services:

Statistical Reporting of U.S. Farm Commodities," The American Economic Review 62(March, 1972)119-30.

Henderson, J. M., and R. E. Quandt (1971). Microeconomic Theory: A Mathematical Approach, 2nd ed. New York: McGraw-Hill Book Company, 1971.

Hicks, J. R. (1939). Value and Capital. London: Oxford University Press, 1939.Hirshleifer. J. (1971). "The Private and Social Value of Information and the Reward to

Inventive Activity." The American Economic Review61 (September. 1971)561-74.Hirshleifer, J. (1973). "Where Are We in the Theory of Information?" The American Eco

nomic Review63 (May, 1973) 31-9.Ho, Y. C.and K. C. Chu (1974). "Information Structure in Dynamic Multi-Person Control

Problems." Automatica 10 (1974) 341-51.Hofstedt, T. R. (1972). "Some Behavioral Parameters of Financial Analysis." The Ac

counting Review 47 (October. 1972) 679-92.Hofstedt, T. R. (1976). "Behavioral Accounting Research: Pathologies. Paradigms and

Prescriptions." Accounting, Organizations andSociety1(Spring, 1976) 43-58.Horngren, C. T. (1955). "Security Analysts and the Price Level." The Accounting Review

30 (October, 1955)575-81.Horngren, C. T. (1956). "The Funds Statement and Its Use by Analysts." The Journal of

Accountancy 101 (January. 1956) 55-9.Horngren, C. t. (1957). "Depreciation, Flow of Funds and the Price Level." Financial

Analysts Journal 13 (August, 1957) 45-7.Ijiri, Y. (1975). Theory o/Accounting Measurement. Studies in Accounting Research «10.

American Accounting Association. 1975.Ijiri, Y., and R, K. Jaedicke (1906). "Reliability and Objectivity of Accounting Measure

ments." The Accounting Review 41 (July, 1966) 474-83.

1 58 Statement on Accounting Theory and Theory Acceptance

Jaffe,J. F. (1974). "Special Information and Insider Trading." The Journal of Business_47—July, 1974)410-28,

Kaplan. R S. (1975). "The Information Content of Financial Accounting Numbers: ASurvey of Empirical Evidence." The Impact of Accounting Research in Financial Accounting andDisclosure on Accounting Practice. Durham. N.C.: Duke University, Unpublished Proceedings, 1975.

Kaplan, R. S and R. Roll (1972). "Investor Evaluation of Accounting Information: SomeEmpirical Evidence." The Journal ofBusiness 45 (April, 1972) 225-57.

Kenley, W. J., and G. J. Staubus (1972). Objectives and Concepts of Financial Statements. Accounting Research Study No. 3. Accountancy Research Foundation(Australia), 1972,pp. 52-8.

Kihlstrom, R„ and L. J. Mirman (1975). "Information and Market Equilibrium." TheBell Journalof Economics 6 (Spring, 1975) 357-76.

Kihlstrom, R., and M. Pauly (1971), "The Role of Insurance in the Allocation ofRisk."_TheAmerican Economic Review61 (May, 1971) 371-9.

King, B. (1966). "Market and Industry Factors in Stock Price Behavior." The Journal ofBusiness 39 (January, 1966) 139-90.

Knight, F. (1933). The Economic Organization. Chicago: The University of Chicago* Itii), iJiiO.

Kuhn, T. S. (1970). The Structure of Scientific Revolutions. 2nd Edition, EnlargedChicago: TheUniversity ofChicagoPress, 1970.

Lakatos, I., and A. Musgrave, eds. (1970). Criticism and the Growth of Knowledge New1ork: Cambridge University Press, 1970.

Ley B., andS. Kunitzky (1974'). "On the Association Between Smoothing Measures andtneRisk ofCommon Stocks." The Accounting Review 49 (April, 1974) 259-70.

Libby, R. (1975a). "The Use ofSimulated Decision Makers in Information Evaluation."TheAccounting Review. 50 (July, 1975) 475-89.

Libby, R. (1975b). "Accounting Ratios and the Prediction of Failure: Some BehavioralEvidence." Journala/Accounting Research 13 (Spring, 1975) 150-61.

Littleton. A. C. (1953). Structure ofAccounting Theory. Monograph No. 5. American Accounting Association, 1953.

Luce, R. D., and II. Raiffa (1957). Games and Decisions: Introduction and Critical Survey. New York: John Wiley &Sons, Inc., 1957.

MacNeal, K. (1939). Truth in Accounting. Philadelphia: University of PennsylvaniaPress, 1939.

Magee. R. P. (1974). "Industry-wide Commonalities in Earnings." Journal ofAccountingResearch 12 (Autumn, 1974) 270-87.

Malinvaud. E. (1972). Lectures in Microeconomic Theory. Amsterdam: North-HollandPublishing Company, 1972.

Marschak, J., and R. Radner (1972). Economic Theory ofTeams. New Haven and London: Yale University Press, 1972.

Marshall, J. (1974). "Private Incentives and Public Information." The AmericanEconomic Review64 (June, 1974) 373-90.

Marshall. R. M. (1975). "Interpreting the API." The Accounting Review 50 (January,1975)99-111.

May, G.0. (1943). Financial Accounting. New York: The Macmillan Co.. 1943.May, R. G. (1971). "The Influence of Quarterly Earnings Announcements on Investor

Decisions as Reflected in Common Stock Price Changes." Journal ofAccounting Research 9 (Supplement. 1971) 119-63.

May, R. G., and G. L. Sundem (1973). "Costof Information and Security Prices: MarketAssociation Tests for Accounting Policy Decisions." The Accounting Review 48(January, 1973)80-91.

McDonald, D.L.(1967). "Feasibility Criteria forAccounting Measures." The AccountingReview 42 (October, 1967) 662-79.

Bibliography 59

Miller, D., and L. A. Gordon (1975). "Conceptual Levels and the Design of Accounting Information Systems." Decision Sciences 6 (April. 1975) 259-69.

Moonitz, M. (1961). The Basic Postulates ofAccounting. Accounting Research Study No.1. New York: American Institute of Certified Public Accountants. 1961.

Nagel, E. (1963). "Assumptions in Economic Theory." The American Economic Review53 (May, 1963)211-9.

Nelson, K.. and R. H. Strawser (1970). "A Note on APB Opinion No. 16." Journal of Accounting Research 8 (Autumn, 1970) 284-9.

Ng, D. (1975). "Information Accuracy andSocial Welfare Under Honogeneous Beliefs."Journal of Financial Economics 2 (March. 1975)53-70.

Paton, W.A. (1922). Accounting Theory. New York: The Ronald Press Company, 1922.Paton, W. A., and A. C, Littleton (1940). An Introduction to Corporate Accounting Stan

dards. Monograph No.3. American Accounting Association, 19-10.Pattillo. J. W. (1975). "Materiality: the(Formerly) Elusive Standard." Financial Execu

tive 43 (August, 1975) 20-7.Pattillo, J. W.,andJ.D. Siebel(1973). "Quantitative Measures of Materiality." Financial

Executive41 (October, 1973) 31-2.37-8.Pattillo, J. W. (1975). "Materiality: the (Formerly) Elusive Standard." Financial Execu

tive 43 (August, 1975) 20-7.Quirk, J., and R.Saposnik (1968). Introduction toGeneral Equilibrium Theory and Welfare

Economics. New York: McGraw-Hill Book Company, 1968.Radner, R. (1974). "Market Equilibrium and Uncertainty: Concepts and Problems" in

Frontiers of Quantitative Economics. Vol. II. M. I). Intriligator and D. A. Kcndrick.eds. Amsterdam: North-Holland Publishing Company. 1974.

Revsine, L. (1970a). "On the Correspondence Between Replacement Cost Income andEconomic Income." The Accounting Review 45 (July, 1970) 513-23.

Revsine, L. (1970b). "Data Expansion and Conceptual Structure." The Accounting Review 45 (October. 1970) 704-11.

Revsine. L. (1970c). "Change in Budget Pressure and Its Impact on SupervisorBehavior." Journal of Accounting Research 8 (Autumn. 1970)90-2.

Revsine, L. (1973). Replacement Cost Accounting. Englewood Cliffs, N. J.: Prentice-Hall. Inc., 1973.

Revsine, L. (1975). "Towards Greater Comparability in Accounting Reports." FinancialAnalysts Journal 31(January-February, 1975)45-51.

Rhode. J. G. (1972). "Behavioral Science Methodologies with Application for AccountingResearch: References and Source Materials." Chapter VII. Report of the Committeeon Research Methodology in Accounting [.AAA. 19721. TheAccounting Review47(Supplement, 1972)494-504.

Rose, J., W. Beaver, S. Beckerand G. Sorter (1970). "Toward an Empirical Measure ofMateriality." Journal o/Accounting Research 8 (Supplement, 1970) 138-56.

Samuelson, P. A. (1963). "Problems of Methodology: Discussion." The AmericanEconomic Review 53 (May. 1963) 231-6.

Sanders, T. H.. H. R. Hatfield, and U. Moore (1938). A .Statement of Accounting Principles. New York: American Institute of (Certified Public) Accountants, 1938.

Savage, L. J. (1954). The Foundations of Statistics. New York: John Wiley&S(Sons, Inc.,1954.

ScitOVSky, T. (1971). Welfare and Competition. Revised Edition. Homewood, III.: RichardD. Irwin, Inc., 1971.

Schroder, N. M., M. J. Driver and S. Streufert (1967). Human Information Processing.New York: Holt, Rinchartand Winston, Inc., 1967.

Securities and Exchange Commission (1976). "Notice of Adoption of Amendments toRegulation S-X RequiringDisclosure of Certain Replacement Cost Data." AccountingSeries Release No. I90(March23,1976).


Sho?i!e«w£; ..J963'- "A SimPlifi(-'d Model for Portfolio Analysis." Management ScienceiJ \ 1 .M). ift,/ I'Jo.

Singhvi, S. Sand H. B. Desai (1971). "An Empirical Analysis of the Quality of Corporate,. '̂nancial Disclosure." The AccoimtingReview46 (January, 1971) 129-38.Smith, J. E., and N. P.Smith (1971). "Readability: AMeasure of the Performance of the

h>7?™? " ,,unct,on of financial Reporting." The Accounting Review 46 (July,

S°U& ?957(S32ACCOUI,Ung Inf°rmati0n Crileria" T'«-- Accounting Review 42Super, F. J., and R. Dolphin, Jr. (1964). "Readability and Corporate Annual Reports."

I He Accounting Review 39(April, 1964) 358-62.Sorter G. II (1969). "An 'Events' Approach to Basic Accounting Theory." The Ac

counting Review 44 (January, 1969) 12-19.

^(August' Smmfi* MarkCl Signa'ing" The Q»arter'y Jo^nal of Economics 87Spence M.. and R. Zeckhauser (1971). "Insurance, Information, and Individual Action "

TheAmerican Economic Review61 (May, 1971) 380-7Sprouse, R. T., and M. Moonitz (1962). ATentative Set o/Broad Accounting Principles for

Business Enterprises. Accounting Research Study No. 3. New York: American Institute of Certified Public Accountants, 1962.

Staubus, G. J (1954). "An Accounting Concept of Revenue." Unpublished Ph.D. dissertation.Graduate School ofBusiness, University ofChicago, 1954.

Staubus G. J. (1961). ATheory of Accounting to Investors. Berkeley. Calif.: Universityof California Press, 1961.

Staubus.G. J. (1967). "Current Cash Equivalents for Assets.A Dissent." The AccountingReview42 (October, 1967)650-61.

Staubus, G J. (1970). "Determinants of the Value of Accounting Procedures." Abacus(December, 1970) 105-19.

Staubus, G. J. (1971). "The Relevance of Evidence on Cash Flows." In Asset Valuationand Income Determination. R. R. Sterling, ed. Lawrence, Kan.: Scholars Book Company, 1971. pp. 42-69.

Staubus, G. J. (1976). "The Multiple-Criteria Approach to Making Accountingueusions. Accountingand BusinessResearch 6 (Autumn, 1976) 276-88

Sterling R. <1967). "AStatement of Basic Accounting Theory: AReview Article." Journal o/Accounting Research 5 (Spring, 1967) 95-112.

"'"niyisVpres^Sterling, R. (1970b). "On Theory Construction and Verification." The Accounting Review

45 (July, 1970)444-57Sterling, R. (1972a). "Introduction." Research Methodology in Accounting. R. Sterling

ice. Kan.; Scholars Book Co.. 1972. pp. 1-7. [Note: papers which appear iiicd., Lawrence.A \ \ ,,o-o, "I i"u " ",*"",'•—'.'. ""*" i'p' '"'' ""'"-'• F"l"-'':' which appear ina.a.A (1972), citedabove - including Beaver (1972), and Rhode (1972) - appear inthis volume as well, together with discussions. ]

Sterling. R. (1972b). "Decision Oriented Financial Accounting." Accounting and Business Research 2 (Summer. 1972) 198-208.

Sterling. R., and W. Harrison (1974). "Discussion of Capital Market Equilibrium In-formation Production, andSelecting Accounting Techniques: Theoretical Frameworkand Review of Empirical Work." Journal ofAccounting Research 12 (Supplement1974)142-57.

Stiglitz, J. E. (1975). "Incentive, Risk and Information: Notes Toward a Theory ofHierarchy." The BellJournal of Economics 6(Autumn, 1975)552-79

Streufert, S. (1970). "Complexity and Complex Decision Making: Convergences BetweenDifferentiation and Integration Approaches to the Predictionof Task Performance "Journal of Experimental Social Psychology 6 (October. 1970)494-509.

Bibliography fil

Streufert, S., and Schroder, H. M. (1965). "Conceptual Structure, Environmental Complexity and Task Performance." Journal of Experimental Research in Personality 1(October, 1965) 132-7.

Sunder, S. (1973). "Relations Between Accounting Changes and Stock Prices: Problemsof Measurement and Some Empirical Evidence." Journal of Accounting Research 11(Supplement, 1973) 1-45.

Sweeney,H. W.(1930). "Maintenance of Capital." TheAccounting Review5 (December,1930) 277-87.

Sweeney, II. W. (1932). "Stabilized Appreciation." The Accounting Review 7 (June, 1932)115-21.

Sweeney, H. W. (1933a). "Capital." The Accounting Review 8 (September, 1933) 185-99.Sweeney, II. W. (1933b)."Income." TheAccountingReview8 (December, 1933) 323-35.Sweeney, H. W. (1936). StabilizedAccounting. New York: Harper &Brothers. 1936.Thomas, A. L. (1969). The Allocation Problem in Financial Accounting. Studies in Ac

counting Research «3. American Accounting Association, 1969.Thomas, A. L. (1974). The Allocation Problem: Part Two. Studies in Accounting Re

search #9. American Accounting Association, 1974.Vattcr, W. J. (1947). The Fund Theory of Accounting and Its Implications for Financial

Report-;. Studies in Business Administration XVII, 2, Chicago: The University ofChicago Press, 1947.

von Neumann, J., and O. Morgenstern (1947). Theory of Games and Economic Behavior.Princeton, N. J.: Princeton University Press, 1947.

Wells, M. C. (1976). "A Revolution in Accounting Thought?" The Accounting Review 51(July, 1976)471-82.

Wilcox, E. B. (1941). "Comments on 'An Introduction to Corporate Accounting Standards.' " The Accounting Review 16 (March, 1941)75-81.

Wilson, R. (1974). "Information in an Equilibrium Model." Working Paper No. 42. Institute for Mathematical Studies in the Social Sciences, Stanford University, 1974.

Woolsey. S. M. (1973). "Materiality Survey." The Journal of Accountancy 136 (September, 1973) 91-2.

Statement on Accounting Theory and Theory Acceptance

Documents