Accounting and Busine,<:s Research, Vol. 40. No. 3 2010 International Accounting Policy Forum, pp. 287-299 287

Conceptual frameworks of accounting froman information perspectiveJohn Christensen

Abstract — This paper analyses the benefits of accounting regulation and a conceptual framework using an informationeconomics approach that allows consideration of uncertainty, multiple agents, demand for information, and multipleinfonnation sources. It also allows private infonnation to enter the analysis. The analysis leads to a set of fundamentalproperties of accounting information. It is argued that the set of qualitative characteristics typically contained in conceptualframeworks does not adequately aggregate the infonnation demands of users of accounting information. For example, theIASB's conceptual fi^mework contains no guidelines for the trade-off between relevance and reliability. Furthermore,neutrality might not be part of an optimal regulation. The statistical bias introduced by the stewardship use of accountinginformation is not necessarily undesirable and will always remain; stewardship is the characteristic of accounting informationthat provides incentives for management to act in the desired way. Accounting infonnation is inherently late compared toother information sources but influences and constrains the content of more timely sources. The accounting system does notexist in a vacuum. Other information sources are present and the purpose of the accounting system cannot be analysedwithout considering the existence of other infonnation sources. Finally, financial statements are audited by an independentauditor. This implies that accounting data are hard to manipulate.

Keywords: accounting regulation; conceptual fi^mework; qualitative characteristics; information economics

1. IntroductionThe question I have been asked to address is howconceptual frameworks contribute to the quality ofcorporate reporting regulation. This is by no meansan easy task. In the paper I shall attempt to show thatan answer requires identification of the concept ofquality of corporate reporting, of the purpose of theconceptual fi^amework, and of the benefits ofreporting regulation. In order to understand theconcept of the quality of corporate reporting it isimportant to analyse the fiindamental characteristicsof accounting information and its limitations.

The idea of the conceptual fi-amework is toprovide a set of consistent principles to guideregulation and reporting of financial information aspart of the political decision process. The IASB'scurrent conceptual fi-amework (IASC, 1989) givesequal ranking to information that is usefiil to a widerange of users in making economic decisions(para. 12) and infonnation that shows the resultsof stewardship of management (para. 14). TheDiscussion Paper that sought to bring together theIASB and FASB conceptual fi-ameworks (FASB/IASB, 2006) asked whether stewardship had a

*The author is Professor of Accounting at the Department ofBusiness and Economics, University of Southern Denmark,Campusvej 55, DK-5230 Odense M, Denmark. Tel: +45 65503244. E-mail: jcn@sam.sdu.dk.

He is grateful to the editor Pauline Weetman, an anonymousreferee, Paul Boyle, Robert Hodgkinson, Richard Macve,Mogens Nielsen, Brian Singleton-Green, Alfi'ed Wagenhofer,and the participants at the 2009 Information for Better MarketsConference for helpful comments.

continuing role in the objective and indicated apreference to focus solely on decision usefijlness(para. BC1.32 to BC1.41). In the proposed concep-tual fi-amework (FASB/IASB, 2008) the mainobjective of decision usefialness is expanded toinclude information about 'management's ability toprotect and enhance the capital providers' invest-ments' (para. OB 9).

Previous work has shown that, in a single firmsetting, the accounting system has to be finely tunedto the specifics of the organisation and its environ-ment, including the economics of the firm, thedecision problems at hand, the private informationof the-parties involved, the public information, andthe moral hazard problems of the organisation.Furthermore, the world contains many firms andmany decision-makers.

It is impossible to construct an income measurethat reflects true income as defined by Hicks (1946)when markets are not perfect and complete (Beaverand Demski, 1979). Such a measure does not exist.Rather, accounting should be viewed as an infor-mation system as acknowledged by both FASB andIASB in their original conceptual fi-ameworks(;FASB, 1978; IASC, 1989). Unfortunately, thereis no universal ranking of information systems(Christensen and Demski, 2003). In addition, it iswell known that no rational preference relationdescribes the decision process of society (Arrow,1951). The accounting system is the result of adelicate balancing of the possibilities imbedded inthe accounting system and the demands of the users.

288 ACCOUNTING AND BUSINESS RESEARCH

This balancing is not entirely of a technical nature asit calls for balancing of preferences of the partiesinvolved. Such balancing cannot be achieved bytechnical rule-making and is inevitably the result ofa political decision process.

The accounting income number reports firm-specific financial information to the market and thusreduces the information asymmetry in the market.The paper considers how information is simultan-eously used by investors to make decisions and toinduce or infiuence management to behave opti-mally or to use the entity's resources efficiently. Thedecision-inñuencing role distorts the reportingincentives. Once the accounting information isused for performance evaluation (or for decisionsregarding replacement of management), incentivesfor earnings management arise (Burgstahler andDichev, 1997; Graham et al., 2005). The reason isthat the financial statements include reporting ofprivate information by management as part of theaccruals. Auditing reduces this problem to someextent. Managers often have an informationaladvantage over the auditors and this prevents theproblem from being completely eliminated.

A current trend in financial reporting is towardadoption of a fair value approach. It is highlyquestionable whether this is a viable path. The firmhas an information advantage compared to the usersof financial information and this advantage is usedstrategically in reporting. Fair value accountingrelies even more on the private information ofmanagement, and enhances the possibilities forearnings management and leaves auditing lessefficient. In addition, it is not obvious whatinformation should be included into the financialstatements. One question is whether the accountingsystem or the users are better at performing theaggregation of various information sources(Christensen and Frimor, 2007).

A related question is how the accounting systembest complements other information sources. Thefinancial statements will always be published latecompared to other information sources. This is dueto the nature of financial statements as all transac-tions must be processed and audited before thestatements are released. In contrast, management'sforecast might be timely. The prime purpose offinancial statements might be to provide incentivesfor reporting of other types of information.

The conceptual frameworks of both IASB andFASB identify sets of the qualitative characteristicsof financial information. The origin of the qualita-tive characteristics is related to the decision orien-tation of accounting and was stated in the ASOBATreport on accounting theory (AAA, 1966). These

qualitative characteristics certainly describe theattribute for a usefial information system when it isused for decision-making purposes in a one-personworld. This might be very different in a multi-person world. For example, one of the characteris-tics calls for unbiased reporting standards, yetrecent research finds that the introduction of biasmight lead to welfare improvements (Christensenand Demski, 2007). Furthermore, it is impossible tomaximise all qualitative characteristics simultan-eously and consequently there is a demand fortrade-offs. However, the frameworks are silent onhow to do this. In a multi-person world it is notpossible to replace the individual preferences with aset of qualitative characteristics.

It is obvious that there is a demand for regulationof financial reporting and that a conceptual frame-work includes the objective and basic principles ofreporting regulation. Given the multi-person natureof the problem it might take the form of aconstitution (in the sense of fiindamental laws andprinciples). The benefit will be that the conceptualframework forces the regulators to constantly seeksolutions that are maintaining and enhancing thecomparative advantage of the accounting systemcompared to other information systems such aspress releases and web-based information sources.

An information economics framework allowsconsideration of uncertainty, multiple agents,demand for information, and multiple informationsources. It also allows asymmetry in the knowledgeof different stakeholders in the market such that thefirm knows something which is of value to themarket participants. Finally, it allows incentiveissues to be part of the analysis.

My analysis of how conceptual frameworkscontribute to the quality of corporate reportingwill fall into four parts. Section 2 will analyse thesupply of accounting information for decision andcontrol. Accounting accruals are seen as the primaryvehicle for management to report their privateinformation. The incentives for such reporting arereviewed. Section 3 will deal with the demand forfinancial information. It is argued that the demandfor decision purposes and control purposes leadsto different ranking of accounting systems. Thesuggestion to have different accounting systemsfor different purposes is analysed and rejected.Section 4 will discuss the fundamental properties ofaccounting information. The qualitative character-istics and other fiandamental properties will be partof this discussion. Section 5 will show the impli-cations of the analysis for the accounting regulationand the conceptual framework. Conclusions areoffered in Section 6.

Vol. 40, No. 3. 2010 International Accounting Policy Fomm 289

2. The reporting organisation2.1. Income measurementOne ofthe prime targets for accounting informationis the investor group and, according to commonwisdom, the main interest for this user group is thefuture cash flows ofthe company. First assume thatthe firm is placed in a perfect capital market undercertainty. The cash fiow series for the lifetime ofthefirm is given by:

Given perfect capital markets, a no arbitrageargument leads to the well-known result that theA^PKof investing in a firm is zero. The value ofthefirm at any given date t is given by:

PV,= >'-J

With this the ineome definition easily follows:

(2)

(3)

This is economic income and coincides with theclassical income as defined by Hicks (1946) and it isequal to cash flow minus economic depreciation.Again, a no arbitrage argument leads to this incomebeing equal to the interest eamed on the investedcapital or:

/, = iPV,-\ (4)

In a perfect world income measurement is notinteresting (Beaver and Demski, 1979). Everybodyknows everything and information adds nothingnew.

2.2. Decision informationImperfection might take on many forms. The first tobe introduced here is adding some of the detailsleading to the given cash flows. The firm isproducing one product and the basis for this is aninitial capital investment and labour in each period.Both are acquired in perfect markets and theproduction fiinction describes a feasible relation-ship between inputs and output. The demand isexogenously given for each period. The realisedcash flows are the consequence of optimal produc-tion during the lifetime ofthe firm. If inventories arepossible and if the production fianction exhibitseconomics of scale or scope, production smoothingwill be part of the optimal production schedule.Thus, the total assets of the firm will both containthe inventory of finished products and the fixedassets. Despite the fact that the value of the firm isuniquely determined, it is not possible to findindividual values ofthe two assets that add up to thetotal value of the firm. The non-separability of the

cost function combined with the non-perfect mar-kets (for the finished product and the fixed assets)leads to this result (Christensen and Demski, 2003).The result points to the difficulty there is in definingappropriate and descriptive accounting measureseven when faced with lots of regularity andcertainty. The analysis by Brom wich et al. (2009)reinforces this point.

Now, no uncertainty leads to no demand forinformation (or there is no such thing as informationin such a world). Formal introduction of uncertaintyinto the model calls for a definition of the errorterms that have an influence on the cash flows ineach period. The simplest model of this typeincludes the following stochastic cash flow series:

CF = [CFo + 60, CF, + E,,..., CFr + «r) (5)

Assume that the e,'s are identically and independ-ently distributed. The accounting system reportsroutinely the realised cash flows, but the realisedcash flow ñ"om period j will have no predictiveability with respect to fijture cash flows. Thus,accounting information is only keeping track oftherealised cash flows, but it is hardly useful.

The introduction of a correlated error structurechanges this. Actions or decisions often have amulti-period effect and this feeds into the stochasticdescription of the fixture cash flows. Now observa-tion/reporting of the cash flows will provideinformation that enables the user to update theexpectation ofthe future cash flows. This estimationuses the correlation structure. This is informationfor valuation purposes as in Peasnell (1982). For thepurpose of facilitating this estimation the account-ing system might be useful. This will be the casewhen the accounting system, together with thereporting of realised cash flows, provides moreinsight into the error structure, thus enabling theuser to better form expectations about future cashflows. More accounting variables might improvethe estimation. The key to finding valuable infor-mation is to get information about the fiindamentaltime processes or the components which character-ise the evolution ofthe cash flow series. Any bias inthe accounting variables does not matter as long asthe user is able to inverse the bias and decipher thecontent of the accounts (Demski and Sappington,1990). The important component remains theunexpected error which is used to form expect-ations. Any systematic bias in the accounting modelis easily countered through a balanced use of theinformation.

Furthermore, the double entry accounting systemsatisfying the clean surplus relations or the com-prehensive income contains counterbalancing

290 ACCOUNTING AND BUSINESS RESEARCH

errors. Over the lifetime of the firm the accountingincome will always equal the total cash flows of thefirm. That is, whichever errors the present account-ing valuation includes, these are balanced by theerror in the ftiture expected accounting incomenumbers (Feltham and Ohlson, 1995).

2.3. Control informationThe control use of accounting numbers is oftenmodelled using the principal-agent model. Theprincipal hires an agent to perform a task for somereason. The principal is unable to observe the actionselected by the agent and at the same time the act is asource of disutility for the agent as he does not wantto work hard. The market for labour of the type ofthe agent determines the level of the salary. Theagent has to be offered at least the utility he can getfrom working elsewhere to accept working for theprincipal. This is the classical moral hazard problem(Hobnstrom, 1979). The naïve interpretation of thismodel is cast in the form of a simple workingrelationship. However, it is descriptive of muchmore complex relationships. Managerial actionchoices are hardly observable in a classic sense. Inthe managerial context the issue of goal congruenceis also predominant in the management accountingtexts (Homgren et al., 2003; Antle and Demski,1988). The bottom line is the agent (manager) wantsto select a different action from the one desired bythe principal (owner). This is the basic moral hazardproblem.

In response to this problem the owner introducesan incentive scheme to make the manager select thedesired action. The infonnation supplied by theaccounting system becomes important here as thepayment is a ñinction of the available accountinginformation and the additional contractible infor-mation signal. The owner's outcome might or mightnot be included in the accounting information. Onereason for this asymmetry is that the owner's timehorizon might extend beyond the accounting andother information available. This is in line with thenet present value focus of the stockholders.

From an accounting perspective it is interestingthat the infonnation of value in this type of model isinfonnation about the act taken. That is, wheneveran additional infonnation source is available, it isuseftal or of value if it provides more informationabout the act selected (Holmstrom, 1979). Theinteresting infonnation in this context is informa-tion that informs the parties about the source of themarket imperfection, in this case the non-observa-bility of the act selected. The ftandamental goalconflict between the manager and the owner isessential for this result. If that were not present, the

manager would simply choose the first best action.Thus, the problem disappears when there is perfectgoal congruence in the organisation such that thereis no demand for incentive pay to promote theactions desired by the owners. This includes thetension between long-term and short-term profitmeasures. Control problems are important toaccounting (Sunder, 1997).

The demand for infonnation for control purposesis closely tied to the act selection. Thus, it dependsupon the set of available actions how these differ interms of the manager's preferences and in terms ofthe owner's preferences. One information system ispreferred to another information system if it is betterat providing incentives for the manager to select thedesired action. Intuitively it translates into how theinfonnation systems are able to distinguish amongthe available actions. When the focus is on theaccounting system it is also important to acknow-ledge the presence of altemative infonnationsources. The value of a particular accounting systemdepends upon which other information sources arepresent. This is, however, only part of the story.

2.4. Reporting incentivesManagers are employed to make decisions. Theyare also supposed to collect and process informa-tion. And finally, they are employed to reportinfonnation, for example, through the accountingsystem. Thus, part of the managerial job is toacquire infonnation and this information is, unlessdisclosed, private to the manager. Some managersare even hired because they possess special know-ledge, which is also private infonnation. In all casesthis adds to the imperfection of the relationshipbetween the owner and the manager. Consequently,the contractual arrangements between the twobecome more complex in response to this privateinfonnation. More interestingly, the timing of howthe events unfold becomes part of the problem.

If the private infonnation gets into the hands ofthe manager after the actions have been selected,there is no immediate control problem. The contractthat controlled the actions of the manager withoutconsidering the new infonnation will continue to doits job and induce the same action choice. However,a new option arises as it might be possible to allowthe manager to communicate his private informa-tion to the public and thus make the informationavailable for contracting purposes. Accountinginfonnation is often of this type as accruals areconstructed at the end of a period.

Given the late arrival of this information it cannotbe used for selecting the action. The communicatedinfonnation can only be used for control purposes

Vol. 40, No. 3. 2010 International Accounting Policy Forum 291

as the information might be used to induce theefficient operation of the firm through incentive pay.The communication of such private information isconstrained as the manager will only communicatesignals that are in the interest of the messenger.Only the manager knows the specifics of theinformation and the communication is impossibleto control directly. Thus there is yet anotherpossibility for gaming by the manager. One wayto resolve this is to make the owner offer a set ofcontracts to the manager. The manager is thensupposed to choose among these. This is a revela-tion game and through his choice the managerreveals his private information. The use of theinformation for contracting purposes is limited tothe use which is specified in the chosen contract.What is more important is that the owner has tocommit to refraining fi-om using the informationotherwise. The communication is controlled by theother more primitive observable variables such ascash fiows and by the potential use of the informa-tion for control purposes (Christensen, 1982). Theaccruals are used for communicating the privateinformation of the managers, and reporting incen-tives have to be taken into consideration. Otherinformation is used to control the reporting incen-tives.

If the information is available to the managerbefore the decision process is finished, the infor-mation has the potential of informing and thusdirectly influencing the decision. This might also bethe very purpose of hiring a manager in the firstplace. There is then a potential for using theinformation in the evaluation process as in theprevious case, but in addition to that there is also apossibility for the manager to use the information inthe decision-making process. The double use of theprivate information has an impact on the controlproblem as that might be improved or made worse.In extreme cases, the private information might beof negative value to the firm. The information fi-omthe manager can be extracted in the same way asabove but more complicated incentive issues haveto be taken into consideration (Christensen, 1981).

Communication of private information is pos-sible both inside and outside the accounting system.Communication within the accounting system islimited to financial information. The initial record-ing of a transaction takes place inside the account-ing system. This information is then oñen combinedwith the manager's private information fi-om outsidethe accounting system to form accruals. Theaccounting system handles this combination usingconsistency as a controlling device. Depreciation isa good example of how management's expectation

is entered into the accounting system as an accrual.The initial recording is historical cost and thedepreciation follows a predetermined plan accord-ing to the expectations of management. During thelifetime of the investment the managers might learnmore about the profitability of the investment. Thenormal accounting treatment will not allow suchinformation to enter the accounting system. Onlyhard evidence is accepted as an excuse for changingthe depreciation plan. Modem times call for fairvalue to enter the accounting system. Fair valueaccounting constitutes another example of account-ing control and here it is historical cost combinedwith market data that forms the accrual.Management has private information concerningthe market value when considering firm-specificassets. Market value is not always exogenouslygiven and users of financial statements haveconcems about the completeness of the marketsearch performed by management. The controlproblem is easy to solve when there is a well-functioning market for the asset in question. Then itis routine to report a market-based value of theasset. When the market is less well-functioning,evidence has to be present to defend the accountingtreatment.

The communication might also take place outsidethe accounting system and then the communicationis fi"ee of the rules, regulation, and conventions thatgovern the accounting system. This informationchannel is heavily used by modem corporations,and security regulation is in place to regulate thesharing of infonnation among market participants.The content of the communication is subject tomarket control, and the published financial state-ments are certainly part of the set of controls.

2.5. AuditingAuditing is an important part of the controls thatallow private information to be communicated tothe decision-makers. The auditor might have twofunctions: that of a quality control and that of anindependent actor who provides credibility to thereport (Kinney, 2000). Given the regulation whichsurrounds the auditing profession, the latter taskmust be very important. The first task could easilybe carried out by a person who is directly employedby the firm. The latter task calls for independence(Antle, 1984).

The auditor usually has a disadvantage comparedto the manager of the firm when it comes toinformation about the firm. If reporting incentiveswere trivial, the manager's self-reporting of firm-specific information would clearly dominate anyinformation that the auditor could provide. It is also

292 ACCOUNTING AND BUSINESS RESEARCH

noteworthy that the auditor only provides anopinion of the published financial statements.Thus, he is not producing the information himselfbut only verifying the content. Given the problem-atic reporting incentives of management, this fiinc-tion increases the infonnation content of thepublished statements. (Christensen and Demski,2003). A consequence of this finding is that it isvery important that the information contained in theaccounting system can be audited. In that way, auditability also becomes a constraint on the accountingsystem.

3. Demand for informationS.I. The users of financial statementsThe demand for accounting information, even for asimple firm and only considering the owners (andpotential owners) and the manager, is quite com-plicated. The present and potential owners haveinvestment decisions as well as control decisions.The demand for infonnation depends upon thefiiture cash flows, the control problem that is facedby the organisation, the access of the two parties toinfonnation, whether there is any private informa-tion, and the possible observables to be used incontracting. The usefiilness of communicated pri-vate information is very sensitive to all of thesefactors. Consequently, even at the firm level, thedemand for information and the optimal choice ofaccounting infonnation system wiU be very specificto the firm characteristics. The demand for infor-mation is partly a response to the fiictions in themarkets faced by the firm and the intenelationshipamong the available sources of information, thegoal congruence of management, and the constitu-ents of the firm and the incentives. A minor changein one information source might have dramaticconsequences for the information content and theuse of accounting accruals. The reporting incentivesare hard to control. Thus there is no universally bestway to manage the reporting of the firm. The choicemust refiect a cost benefit comparison in order toreach an optimal system (Christensen and Demski,2003).

The information content of the accounting sys-tem is mainly firm-specific information providingthe investors with input for their investment deci-sions. Most of the information contained in theaccounting system is endogenous to the firm butsome pieces of infonnation are the consequence ofthe mixing of endogenous and exogenous informa-tion. The prime example of this is fair valuevaluation, which includes market infonnation.

Modem finance has taught us that a rational, risk-averse investor invests in a diversified portfolio of

assets. The Capital Asset Pricing Model (CAPM)has infonned us that the main ingredient in thepricing of a security is the association of the securityand the market portfolio such that the correlationbetween the security and the market portfolio, i.e.the beta, accounts for the pricing of the security(Beaver, 1998). The investor will demand a riskpremium for the market risk that is associated withinvesting in a specific security. The firm-specificrisk will be diversified and the result is that this partof the firm-specific risk disappears fi-om the equa-tion. The investor may also want information aboutfirm-specific risk in order to diversify, particularly ifhe has non-diversifiable endowments (e.g. propertyinvestments or skills). Consequently, the investor'sdemand for information concems the risk and theconelation of the firm retum and the market retum.This is not the type of information which is giventhe highest priorify in the accounting system.

On the supply side, the accounting systemcontains financial information about the activitiesof the firm. The data are initially collected within thefirm as transactions. Later, revaluations andaccounting accruals are added to the system. As aresult, some data are hard data in tJie form ofrealised cash flows, and other data are of a softernature as the accmals are based upon the expect-ations of management, perhaps inspired by exo-genous events like price changes. A generalcharacteristic of the accounting information is thestamp which is provided by the auditor of the firm.The accounting system has the comparative advan-tage that it produces firm-specific informationprimarily about the finn's financial position.

Thus the accounting information is not indemand by the general investor who follows theadvice to invest in well diversified portfolios.Rather, accounting infonnation is usefiil for personswho are placed (for some reason) in a speculativeposition. Some investors look for information aboutfixture cash flows to identify when it is optimal toexchange the investment for cash. Another group ofinvestors look for stewardship information toinduce efficient operation of the firm. There is nota generally best accounting system across firms(Christensen and Demski, 2003). The optimalinfonnation system is unique to each relationship,and the accounting system has to compromiseamong the users and producers. A choice also has toreflect the preferences of the stakeholders of thefirms and it has to balance the possible uses of theaccounting system. Furthermore, as Arrow's the-orem suggests the non-existence of a social welfarechoice fiinction, the choice must be the result of apolitical decision process (Anow, 1951, and

Vol. 40, No. 3. 20\0 International Accounting Policy Forum 293

Demski, 1973). The exposure draft, (FASB/IASB,2008), acknowledges this dilemma more specific-ally than the extant IASB and FASB conceptualframeworks, as it identifies the capital providers asthe primary users (para. 0B5 to 0B8). Capitalproviders' demand for information includes boththe abilify to generate fiiture cash flow and theabilify to protect and enhance the investments. Thefocus on reporting incentives could be stronger.

This is fijrther mixed with a public goodsproblem in the sense that once the information isproduced and the cost for that is incurred, theinformation is a free good. Consequently, if leftalone, the market would end up with an under-supply of accounting information. This creates alemons problem (Akerlof, 1970) in the market forthe assets of the firm and there is a demand forregulation of the supply of firm-specific informa-tion. If the signalling behaviour is important theresult might be an oversupply of information.

3.2. Multiple uses of financial informationAccording to the discussion of the reporting firm,accounting serves two purposes: decision andcontrol. Information for decision purposes is infor-mation that enables the decision-maker (aninvestor) to estimate the fiiture cash flows forinvestment decisions. This means information thatfeeds into the net present value calculation.Different signals lead to different decisions.Information for control purposes informs the deci-sion-maker about the 'act' selected by the managerof the firm. Here the important characteristic is theability of the information to provide informationthat enables the owner to distinguish the desirablefrom the undesirable action. The purpose of this isthen to allow the owner to provide incentives forselection of the desirable action. The two purposesare not identical and the rankings of informationsystems according to these two purposes do notnecessarily coincide. This means that when facedwith a choice among a set of information systems,one information system might be preferred for onepurpose and another might be preferred for the otherpurpose. This implies that there is not one univer-sally optimal accounting system independent of theuse of the information (Gjesdal, 1981; Christensenand Demski, 2003).

One way to proceed is to consider severalaccounting systems - one for each purpose or usergroup. Generically that could be one for controlpurposes and one for decision purposes serving thestockholders of the firm. In this way, the accountingsystem could overcome the incentive issues raisedpreviously as management will only communicate

information which is in its own best interest forcontrol purposes. Separating the two sets of reportswill remove this conflict. Unfortunately this is not aviable option. Management will only communicateeverything if the users are able to and will commitnot to use the information too aggressively. Giventhe separation of management and stockholders, itis hardly possible for the owners to commit to such apolicy. Bad news in the decision informationdomain will at some point spill over to bad newsin the control domain and thus the incentives forcomplete and truthfiil communication in the deci-sion domain break down. Therefore, we cannotexpect that a separation of the user groups and theirreports will lead to an accounting system fordecision-making which is free from the bias intro-duced by the incentives of management.

4. The properties of accounting information4.1. The qualitative characteristics of financialstatementsThe qualitative characteristics of the conceptualframework are the attributes which make theinformation usefiil to users according to theconceptual frameworks of the IASB (IASC, 1989)and FASB (FASB, 1980a, 1980b, 1984, 1985). Thisis the important link between the information sourceand the users. The qualitative characteristics thenfiinction as a proxy for the users. The details of thedecision problem are replaced by the qualitativecharacteristics. As noted earlier, the origin of thequalitative characteristics is related to the decisionorientation of accounting and was stated in theASOBAT report (AAA, 1966). These qualitativecharacteristics certainly describe the attribute for auseful information system when it is used fordecision-making purposes in a one-person world.The inherent multi-person nature of most account-ing issues is ignored. The question to be analysedhere is whether the qualitative characteristics canreplace the users when the regulators are decidingupon accounting standards.

Focus first on the pair 'relevance' and 'reliability'as these attributes have hitherto been identified asthe most important ones (although 'reliability' isreplaced by 'faithfiil representation' in FASB/IASB,2008). The IASB conceptual framework (IASC,1989: para. 45) calls for a balancing betweenqualitative characteristics but offers little assistancebeyond a reference to 'professional judgement'. Inthe wording of the Exposure Draft (FASB/IASB,2008), 'Enhancing qualitative characteristicsimprove the usefulness of financial informationand should be maximised to the extent possible'(para. QC 25). In order to analyse this balance.

294 ACCOUNTING AND BUSINESS RESEARCH

briefly consider the model of Feltham and Xie(1994). They consider a multi-task agency model.The manager faces a two-dimensional task and issupposed to select an action pair a = (a^, 02). Theowner wants to maximise expected profit U = biai +02^2- The accountant has to choose between twodifferent information systems. The first will report aprofit of;?! = ¿la, + ¿202 + ei and the second willreport a different profit ofp2 = Cjûi + C2a2. The firstinformation system weighs the two actions accord-ing to the objective function of the firm but hasnoise as well. The second has no noise but gives abiased profit compared to the objective ofthe firm.Thus the first information system scores high onrelevance and low on reliability. The second is veryreliable but less relevant. Using the two infonnationsystems for contracting purposes illustrates theconsequences of a second best world. Using the firstprofit measure leads to a deficiency. Providingincentives to work imposes a non-trivial risk uponthe manager and consequently he requires a riskpremium in his pay for performance contract. Thesecond information system leads to a decision fromthe manager which is not aligned with the first bestchoice as the mix ofthe two actions is skewed. Thisis also inefficient. Consequently, there will bedifferent costs associated with the two infonnationsystems. The first will include a risk premium andthe second will refiect the unbalanced weighteddecision. Now perform a comparative static analy-sis. If b is small (compared to ci) the secondinformation system will be optimal. If b is large(compared to ^i) the first information system will beoptimal. The information systems including therelevance or reliability characteristics are notchanged,' yet the optimal information systemchanges as a consequence of the difference in theunderlying decision problem. The optimal choice isnot accurately described by the pair relevance andreliability. Wagenhofer (2009) makes a similarpoint.

To take the analysis one step fiarther, the conceptof faithful representation or neutrality is considered.Faithful representation means that the transactionsand other events should be represented in thefinancial statements in a way it purports to be. Thistakes away any consideration of managerial incen-tives to control the information system. Yet it iswidely acknowledged and documented that there isa phenomenon called eamings management andthat this takes on many forms. This has been

' i t is suggested that reliability is replaced by faithfulrepresentation in FASB (2008). This replacement does notchange the argument as the interpretations of reliability andfaithful representation are identical in the proposed setting.

documented in many and very different ways (Dye,2002; Nelson et al., 2002; Demski, 1998). The nextlevel of eamings management has led to theemergence of something that might be labelleddesigner transactions, which stands for transactionsthat just satisfy a set of conditions to qualify forbeing accounted for in a specific way. The Enroncase was a huge system of such designer transac-tions. The point is that the focus on qualitativecharacteristics skips over the finer details of thedecision problem and in this case over the reactionof those making the reporting system once aregulation is in place. They might design transac-tions to circumvent the regulation. The regulatorsmust consider the incentives of the informationproducers to maximise the result of the regulationeffort. Furthermore, this might lead to optimality ofnon-neutral standards and non-neutral accounts assuggested by Dye (2002) and Christensen andDemski (2007).

The qualitative characteristics work as a way ofsimplifying the decision problem faced by theregulators as the finer details of the accountingdecision problem including preferences, decisionproblems and information environment are simpli-fied into only viewing the qualitative characteristicsof accounting infonnation. This appears to be toosimplistic as it blinds the regulators to incentivesthat are inherent in the system producing accountinginformation and to the more delicate trade-offs thatthe regulators (and information producers) arefacing.

4.2. The fundamental properties of financialstatementsIn very general terms, the purpose of financialstatements is to provide information for the con-stituents of the firm. This is a very broad purpose,yet financial statements have some very fundamen-tal properties which will remain no matter theregulation. Some of these will be discussed in thesix points below. First, the optimal reporting for thefirm is unique to the specifics of the firm. Second,the general purpose of accounting information isusually cast in the wording of decision informationand stewardship information. The bias introducedby the stewardship use of accounting informationwill always remain. Third, the accounting informa-tion specialises in firm-specific information, andmainly investors holding a speculative positionbenefit from financial reporting. Fourth, theaccounting information is inherently late comparedto other information sources. Fifth, the accountingsystem does not exist in a vacuum. Other informa-tion sources are present and the purpose of the

Vol. 40, No. 3. 2010 Intemational Accounting Policy Forum 295

accounting system cannot be analysed withoutconsidering the existence of other infonnationsources. Sixth, the financial statements are auditedby an independent auditor. This implies thataccounting data are hard to manipulate.Eventually a synthesis of these points will lead tothe identification of the role of the conceptualframework.

Following the earlier discussion of the reportingfirm, the main content in financial statements isfinancial infonnation about the firm. This followsfrom the historical development and it also reflectsthe comparative advantage of financial reporting.Furthennore, it was concluded that the optimalinformation system, which balances cost and bene-fits of the information system, is highly specific tothe details of the specific reporting situation. Thisincludes the decision problems faced, the informa-tion present, and the distribution of this informationamong involved parties. There is no universallyoptimal infonnation system independent of thespecifics of the reporting situation.

Furthennore, it is well established that therankings of infonnation systems for decision pur-poses and stewardship purposes are not aligned(Gjesdal, 1981). As argued earlier, it follows fromthe institutional setting that it is impossible to havetwo different financial reporting systems - one forstewardship purposes and one for decision pur-poses. It is impossible for the users to commit to notusing the decision-relevant infonnation for stew-ardship purposes as the use of the informationsystem is decoupled from the production.Consequently, the situation-specific optimalaccounting system will balance the pros and consof the information system for the different purposes.This suggests that there is always a bias in theaccruals which is related to the stewardship use ofthe infonnation.

Another point is how the accounting system bestsupplements the other, perhaps more timely, infor-mation sources that are found in the infonnationsociety. The famous Ball and Brown (1968)diagram suggests financial accounting indeed pro-vides information which is used by the marketparticipants to value the securities in the market.Unfortunately, the diagram also suggests that only asmall fraction (8% according to Lev, 1989) of thetotal infonnation released to the market stems fromthe published financial statements. The marketreacts to all kinds of infonnation and this informa-tion is clearly timelier than accounting infonnation.Yet most infonnation sources are not regulated.Accounting infonnation is heavily regulated andhas an important effect on other information

sources. The financial report has a potential forcontrolling the infonnation content in other perhapsmore timely information sources.

The conceptual framework implicitly assumesthat financial statements should carry all relevantinformation and thus it disregards the existence ofother information sources. This view does not allowfor a specialisation of the different types ofinfonnation sources. The accounting system con-structed in one way might be a better supplement toexisting infonnation sources than another account-ing system that is supposed to stand on its own. Thebig question remains as to who is best able toaggregate the financial information with othersources of infonnation which are available to themarket participant. It appears to be too naive toassume or conclude that this aggregation is bestperformed by the accountants. The famous Roll(1984) paper suggests that this should be takenseriously. Who would, at the outset, have expectedthe financial markets to outperform a set ofmeteorologists when it comes to forecasting wea-ther in a small region of Central Florida? Themarket mechanism is an extremely strong informa-tion aggregator and it might, despite the fact thateach of the market participants only has very noisyinformation compared to the firm, be very efficientat performing such an aggregation

The final observation is that financial statementsare unique in the sense that they are audited. That is,the private infonnation of the firm is verified by anindependent auditor before being entered into thefinancial statements. Consequently, only informa-tion that passes this filter is included (Kinney,2000). This implies that financial statements arehard to manipulate and produce hard infonnationthat is useftil in repairing inefficiencies in markets.

Where do these six observations lead? First of all,financial statements are not particularly suited toserve the diversified investor. Mainly to investorsholding speculative positions will the financialstatements be of value. Next, is the balancing ofinformation to be included in and excluded from thefinancial statements? The first observation is thatthis is an empirical question as it is highlycontingent upon the situation whether the account-ant or the market is best at aggregating information.Routinely, the accounting system steps away fromincluding investment in the ftiture into the assetsbecause the benefits are uncertain both with regardto timing and amount. The market has no problemin including such information into the valuation ofsecurities. The second observation is that account-ing information is inherently late by construction.Events have to take place and the entry into the

296 ACCOUNTING AND BUSINESS RESEARCH

published statements has to be verified by anauditor. Numerous sources of information step inand fill the gap. This is evidenced by Ball andBrown's (1968) diagram and Beaver (1998).

This leads to the following question: whyregulate accounting information when most of theinformation action is going on in the non-regulatedregime of other information sources? These sourcesare perhaps regulated with respect to timing due tosecurities regulation but not with respect to content.One possible answer to that question is that theregulated accounting information serves as theinformation source which 'controls' the otherinformation sources. Then all free voluntary infor-mation disclosures are at the time of financialreporting compared to the published financialstatements. If they are consistent it is viewed asgood news, whereas inconsistency is regarded asbad news. The information content of the voluntarydisclosure is a fiinction of the control that is builtinto the accounting system. The financial statementsserve an important role in controlling such infor-mation.

5. Accounting regulation5.1. The purpose of a conceptual frameworkBefore going into the specifics of the present andproposed conceptual framework it might be usefulto consider the stated purpose and scope of theconceptual framework. The purpose of the presentIASB conceptual fi-amework is to assist the Board indeveloping future accounting standards, to assistthe Board in promoting harmonisation of regula-tions and accounting standards, to assist nationalstandard-setters, to assist auditors in formulatingopinions, and to assist users (IASC, 1989). TheFASB states a similar purpose (FASB, 1980b).Thus, the purpose is twofold. One is to help thestandard-setter to develop future standards, and theother is to help those producing and using thefinancial statements. A framework could beregarded as a constitution defining the generalprinciples for the development of accountingstandards in the regulatory domain and for theinfonnation content of financial statements in theusers' domain. To fulfil this purpose a frameworkshould be invariant over a long period and formu-late the general rules which constitute the core offinancial reporting.

As already indicated, the IASB and the FASB arejointly participating in a project that is intended tolead to a new conceptual framework which unitesthe two fi-ameworks of the two institutions. Thiswork is in progress and many preliminary workingpapers have been released for comment. In the

exposure drañ of the joint Conceptual Framework(FASB/IASB, 2008) the purpose is reformulated asestablishing concepts that underlie financial report-ing. The framework is thought of as a coherent set ofconcepts that fiows from an objective. Manyquestions are being asked and not many have animmediate answer.

5.2. Coordination of the financial statementsThe research activities of the universities serve theimplicit regulation of accounting. Most of the ideaswhich form the basis for our thinking of accountingissues stem from the research community. Thenotable contribution of Paton and Littleton (1940)on corporate accounting standards provides a deepinsight into the fundamental and problematic issuesof income measurement. Also the AmericanAccounting Association's Committees onAccounting Reporting have had some infiuence,most notably the ASOBAT (AAA, 1966) report.

The research industry is not well coordinated.Consequently, a set of definitions of the elements offinancial statements is part of the conceptualfi-amework. The elements are the assets, the liabil-ities, the equity, the income, and the expenses. Theframework also provides definitions of recognitionrules related to the basic elements. Finally, thegeneral rules of accounting measurement areincluded in the conceptual framework. Takentogether, the definition of elements of financialstatements is thought to govern the inclusion andexclusion of information in the financial statements.

Analysing the demand for accounting informa-tion for a specific entity leads to a specific optimalinformation structure. The accounting system has tobe finely tuned to the specifics of the organisationand its environment. The flexibility of the account-ing system is a key to its success as an informationsystem. Within the general framework of accrualaccounting there is room for many variations. Thisallows feeding the expectations of management intothe accounting information in a controlled way. Toomuch regulation would destroy this flexibility andleave the accounting system useless (Christensenand Demski, 2003).

The optimal accounting information system willfill the gap lefi between the private and publicinfonnation to induce optimal decision-making inthe most general sense. Provided there is a well-defined social preference relationship this informa-tion will possibly be unique except for the repre-sentation or the scaling of the information system.Sending a message fi-om one individual to anothermight take on many equivalent forms, e.g. usingdifferent languages. The important function of

Vol. 40, No. 3. 20\0 International Accounting Policy Forum 297

financial statements as an infonnation source is thatthe users are able to invert the mapping thatproduces the information in the first place suchthat the user leams the primitive underlying staterealisation or event (set of events) (Antle andDemski, 1989; Demski and Sappington, 1990).

As the world contains many firms and manydecision-makers, there is a demand for coordinationof the scaling or representation of the infonnation.This would serve the purpose of coordinationamong different users of the information such thatthe infonnation might be understood by a broaderaudience and not coded for a specific user. In theframework this is equivalent to tiie set of definitionsand the set of eiements which constitute thefinancial statements. No doubt this demand forcoordination is real. The demand could be satisfiedin the financial statements as a description of theapplied accounting methods. The textbooks orprofessors are other candidates for taking care ofthis coordination. Given the anarchistic and innova-tive nature of both, this is probably not a good placeto do the coordination. National regulation will notdo the job either, given the open society.Consequently, this job is best served by theintemational accounting regulation.

As noted previously, the current view is thatfinancial statements provide financial informationabout the firm. The format of the informationsystem is defined as accmal accounting anduniversally agreed upon. Pointing this out belongsto a long-term valid conceptual fi-amework. Thespecific definitions of what constitutes the elementsof the subsection of the financial statements aresubject to change as part of an evolution and mightbe better placed in the standards.

5.3. A conceptual frameworkThere is certainly a demand for accounting regula-tion as a market failure can be observed in themarket for information supply. Those who possessthe information might have poor incentives todisclose such information and the market is alsohaunted by a lemons problem as suggested byAkerlof (1970). The literature on the demand forregulation of financial statements is vast and arecent summary of the arguments is given byBushman and Landsman (2010).

Now to the initial question of how do conceptualfi-ameworks contribute to the qualify of corporatereporting regulation? The conceptual fi-ameworkcan be viewed as the constitution (statement offundamental laws and principles) that keeps controlover the process of accounting regulation. Aconstitution should have long-term validify and it

should not be changed in response to small changesin the workings of sociefy. This calls for a ratherrobust wording of the conceptual fi-amework.However, this is not the way the present conceptualframework is set up. It is far too detailed and mightconsequently fail in its purpose, as observed by theAAA committee on Financial Reporting (AAA,2009). I share their view.

A conceptual fi-amework should fiinction as aconstitution to remind the regulators of the overallgoal of financial reporting such that the details ofindividual standards are kept in line with that.Therefore a conceptual fi-amework should containthat goal. As found earlier there is not unanimifyamong stakeholders on this issue and it is part of apolitical process. Thus it is impossible to definewhat is meant by qualify of corporate reportingobjectively and often we are reducing this questionto one of measuring the cost and benefits ofregulation as in Schipper (2010). The regulatorsare supposed to balance the pros and cons ofintroducing or revising a standard. The overall goalis to find a socially optimal level of disclosure offirm-specific financial information which leads towell fiinctioning capital markets and to efficientfirms. Rather than providing a set of qualitativecharacteristics which does not guide the regulatoryprocess, as noted earlier, it would be more usefial tostate the perceived comparative advantages and theperceived limitations of financial statements.

One of the advantages of the accounting systemis that it is audited, which makes the informationhard to manipulate. This is important given the roleaccounting plays in reporting otherwise undisclosedinformation and in controlling other sources ofinfonnation. It might also flag that some pieces ofinformation are hard evidence, whereas other piecesare softer, such as accruals. The latter are reportedby management but have an accounting stamp asthe procedure for producing them lends itself toauditing. The usefiiiness of the pieces of accountinginformation depends critically upon the hardness ofthe data.

The limitations include the potential bias offinancial statements. The word bias can be inter-preted in two ways. One meaning implies that thestatements do not represent the expected value ofthe asset, i.e. the reported value is not equal to themean. It is admittedly a nice property that ameasurement is fi-ee fi-om bias, but any deviationfi-om this norm does not constitute a major problemas that is easily resolved once the source of the biasis known. The troublesome part is the biasintroduced by an involved parfy. Identifying andresolving this fype of bias could point to the

298 ACCOUNTING AND BUSINESS RESEARCH

importance of the incentives of management for thepreparation and interpretation of financial state-ments. This would put these incentives in a centralrole when it comes to the development of newstandards and when it comes to interpreting thefinancial statements of firms. Along these lines, theAAA committee has developed a few ground rulesfor a conceptual framework (AAA, 2009). Thefiiture development of the standards might benefitfrom allowing competition among standards andallowing firms to decide which set of standards theywant to satisfy (Dye and Sunder, 2001).

6. Summary and conclusionsThe conceptual framework has been with us for along time and the regulation of accounting evenlonger. Much regulation has been a consequence ofobserved business failures (Clikeman, 2009). Theglobalisation of business has led to a call forharmonisation of accounting standards around theglobe and as a result the FASB and IASB havejoined forc'es to make one set of accountingstandards. One of their joint projects is a commonconceptual framework ^ASB/IASB, 2006, 2008).Consequently, the present interest in the develop-ment of a new framework is seen.

In this paper I have reviewed the demand for aconceptual framework from an information eco-nomics perspective. This has been a broad analysis.The point of departure was supposed to be theconsequences for the quality of financial reportingof a conceptual framework. This is a very difficultquestion as the notion of quality is very hard todescribe in the first place and secondly because theroles of allocation between a conceptual frame-work, the actual accounting regulation, and theinformation content in financial statements are notpredetermined. Therefore, I have taken the path ofanalysing the role and content of a conceptualframework as a set of ground rules that is usefiil inthe regulation of financial reporting.

The focal point of a conceptual framework mustbe the comparative advantage of accounting.Accounting is an information source which isalways produced late in a decision process. Thatstems from the fact that a main characteristic ofaccounting information is that accounting data: (1)are based upon the financial relationship of the firmwith outside parties and upon formal recognitionrules; (2) are subject to auditing; (3) acknowledgethe role of other, perhaps more timely, informationsources; (4) aggregate and allocate information overtime and units; and (5) are hard to manipulate. It isan important role for the conceptual framework to

help the accounting system in maintaining thecomparative advantage.

ReferencesAAA (1966). A Statement of Basic Accounting Theory.

(ASOBAT). American Accounting Association.AAA (2009). Alternative Conceptual Frameworks forFinancial Reporting. American Accounting Association.

Akerlof, G. (1970). 'The market for "lemons": quality,uncertainty, and the market mechanism'. Quarterly Journalof Economics, 84(3): 488-500.

Antle, R. (1984). 'Auditor independence'. Journal ofAccounting Research, 22(1): 1-20.

Antle, R. and Demski, J. (1988). 'The controllabilityprinciple in responsibility accounting'. The AccountingReview, 63(4): 700-718.

Antle, R. and Demski, J. (1989). 'Revenue recognition'.Contemporary Accounting Research, 5(2): 423-451.

Arrow, K. (1951). Social Choice and Individual Values. NewHaven, CT: Yale University Press.

Ball, R. and Brown, P. (1968). 'An empirical evaluation ofaccounting income numbers'. Reprinted in Journal ofAccounting Research, (1989) 27: 202-217.

Beaver, W. (1998). Financial Reporting: An AccountingRevolution. Englewood Cliffs, NJ: Prentice-Hall.

Beaver, W. and Demski, J. (1979). 'The nature of incomemeasurement'. The Accounting Review, 54(1): 38-46.

Bromwich, M., Macve, R. and Sunder, S. (2009). 'Hicksianincome in the conceptual framework'. Unpublished manu-script, London School of Economics (LSE).

Burgstahler, D. and Dichev, I. (1997). 'Earnings manage-ment to avoid earnings decreases and losses'. Journal ofAccounting & Economics, 24(1): 99-126.

Bushman, R. and Landsman, W.R. (2010). 'The pros andcons of regulating corporate reporting: a critical review ofthe arguments'. Accounting and Business Research, 40(3)(Special Issue): 259-273.

Christensen, J. (1981). 'Communication in agencies'. BellJournal of Economics, 12(2): 661-674.

Christensen, J. (1982). 'The determination of performancestandards and participation. Journal of AccountingResearch, 20(2): 589-603.

Christensen, J. and Demski, J. (2003). Accounting Theory.An Information Content Perspective. New York, NY:McGraw-Hill/Irwin.

Christensen, J. and Demski, J. (2007). 'Anticipatory report-ing standards'. Accounting Horizons 21(4): 351-370.

Christensen, J. and Frimor, H. (2007). 'Fair value, account-ing aggregation and multiple sources of information'. In R.Antle, P.J. Liang and F. Gjesdal (eds.). Essays onAccounting Theory in Honour of Joel S. Demski, pp. 1-15. Kluwer.

Clikeman, P.M. (2009). Called to Account. FourteenFinancial Frauds that Shaped the American AccountingProfession. Routledge.

Demski, J. (1973). 'The general impossibility of normativeaccounting standards'. The Accounting Review, 48(4): 718-723.

Demski, J. (1998). 'Performance measure manipulation'.Contemporary Accounting Research, 15(3): 261-285.

Demski, J. and Sappington, D. (1990). 'Fully revealingincome measurement'. The Accounting Review, 6(2)- 363-383.

Dye, R.A. (2002). 'Classifications manipulation and Nash

Vol. 40, No. 3. 2010 International Accounting Policy Forum 299

accounting standards'. Journal of Accounting Research,40(4): 1125-1162.

Dye, R.A. and Sunder, S. (2001). 'Why not allow FASB atidIASB standards to compete in the U.S.?'. AccountingHorizons, 15(3): 257-271.

FASB (1978). Concepts Statement No. 1: Objectives ofFinancial Reporting by Business Enterprises. FinancialAccounting Standards Board.

FASB (1980a). Concepts Statement No. 2: QualitativeCharacteristics of Accounting Information. FinancialAccounting Standards Board.

FASB (1980b). Concepts Statement No. 4: Objectives ofFinancial Reporting by Nonbusiness Organizations.Financial Accounting Standards Board.

FASB (1984). Concepts Statement No. 5: Recognition andMeasurement in Financial Statements of BusinessEnterprises. Financial Accounting Standards Board.

FASB (1985). Concepts Statement No. 6: Elements ofFinancial Statements. Financial Accounting StandardsBoard.

FASB/IASB (2006). Preliminary Views on an ImprovedConceptual Framework for Financial Reporting: TheObjective of Financial Reporting and QualitativeCharacteristics of Decision-useful Financial ReportingInformation. Issued for comment by the FinancialAccounting Standards Board and by the InternationalAccounting Standards Board.

FASB/IASB (2008). Conceptual Framework for FinancialReporting: The Objective of Financial Reporting andQualitative Characteristics and Constraints of Decision-Useful Financial Reporting Information. Exposure Draft.Issued by the Financial Accounting Standards Board andby the International Accounting Standards Board.

Feltham, G. and Ohlson, J. (1995). 'Valuation and cleansurplus accounting for operating and financial activities'.Contemporary Accounting Research 11(2): 689-731.

Feltham, G. and Xie, J. (1994). 'Performance measurecongruity and diversity in multi-task principal-agent rela-tions'. The Accounting Review, 69(3): 429-453.

Gjesdal, F. ( 1981 ). 'Accounting for stewardship'. Journal ofAccounting Research, 19(1): 208-231.

Graham, J.R., Harvey, C.R. and Rajgopal, S. (2005). 'Theeconomic implications of corporate financial reporting'.Journal of Accounting & Economics, 40(1-3): 3-73.

Hicks, J. (1946). Value and Capital. Oxford: ClarendonPress.

Holmstrom, B. (1979). 'Moral hazard and observability'.Bell Journal of Economics, 10(1): 74-91.

Homgren, C , Foster, G. and Datar, S. (2003). CostAccounting: A Managerial Emphasis. New Jersey, NJ:Prentice-Hall.

IASC (1989). Framework for the Preparation andPresentation of Financial Statements. Published by theInternational Accounting Standards Committee Board(IASC Board) in July 1989 and adopted by theInternational Accounting standards Board (IASB) inApril 2001. Currently available in the annual volume ofIFRS published by the IASB.

Kinney, W. (2000). Information Quality Assurance andInternal Control for Management Decision Making. NewYork, NY: McGraw-Hill.

Lev, B. (1989). 'On the usefulness of eamings and eamingsresearch: lessons and directions fi-om two decades ofempirical Kitarçh'. Journal of Accounting Research, 27(3)Supplement: 153-192.

Nelson, M.W., Elliot, J.A. and Tarpley, R.L. (2002).'Evidence fi-om auditors about managers' and auditors'eamings management decisions'. The Accounting Review,77(4) Supplement: 175-202.

Paton, W. and Littleton, A. (1940). An Introduction toCorporate Accounting Standards. American AccountingAssociation.

Peasnell, K.V. (1982). 'Some formal connections betweeneconomic values and yields and accounting numbers'.Journal of Business Finance & Accounting, 9(3): 361-381.

Roll, R. (1984). 'Orange juice and weather'. AmericanEconomic Review, 74(5): 861-880.

Schipper, K. (2010). 'How can we measure the cost andbenefits of changes in financial reporting standards?'.Accounting and Business Research, 40(3) (Special Issue):309-327.

Sunder, S. (1997). Theory of Accounting and Control.Cincinnati, OH: South-Westem Publishing.

Wagenhofer, A. (2009). 'Global accounting standards:reality and ambitions'. Accounting Research Journal,22(1): 68-80.

Copyright of Accounting & Business Research is the property of Croner.CCH Group Limited and its content

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

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