In this paper I use a principal-agent framework to explore the relation between the hierarchical structure of firms and the accounting information technologies available to them. My analysis is related to that in Melumad, Mookherjee, and Reichelstein [1992] and Ziv [1993]. Melumad, Mookherjee, and Reichelstein model a principal who employs two privately informed agents and chooses either a flat structure where both agents contract and communicate with the principal, or a hierarchical structure in which the principal contracts with only one agent, who subsequently writes a subcontract with a second agent, creating a twolayer organizational form.' Melumad, Mookherjee, and Reichelstein use the revelation principal to prove the general superiority of the flat structure. They add exogenous restrictions on communication (with respect to dimensionality and complexity of the message space) to demonstrate a demand for hierarchy. Ziv [1993], in a moral hazard setting, solves for the optimal number of agents in a one-layer firm, under different exogenously given information structures. In this paper, I take an approach that allows the principal to choose the number of layers in the firm, the

Discretionary disclosures, SFAS No. 106, Accounting standard adoption, Private information

Auditing, Qualified opinion, Market reaction, Litigation exposure

This article examines the tax‐compliance game between taxpayers, a tax‐collecting agency, and third‐party tax‐return preparers. In our model, taxpayers are uncertain about their taxable income and may hire tax practitioners to reduce tax uncertainty. We examine the viability of tax practitioners as a signaling device (taking into account the effects on the behavior of the tax‐collecting agency) and investigate the desirability of encouraging (or discouraging) the use of tax practitioners via the use of alternative tax‐crediting rules. Our study establishes that tax crediting enables legislators to deal better with the consequences of taxpayers' strategic reporting. We show that the effects of changes in crediting rates cannot be replicated by changes in tax rates or penalties; the government would generally like to “price discriminate” in subsidizing tax practitioners' involvement in the tax‐compliance process. It is suboptimal to permit all taxpayers to take a tax credit for preparers' fees; some taxpayers should be denied such a subsidy. Further, if the government is constrained to adopt an identical credit schedule for all taxpayers, it will often find that a policy allowing no tax credit Pareto dominates any uniform crediting policy. Résumé. Les auteurs examinent le jeu auquel se livrent, en matière d'observation de la législation fiscale, les contribuables, les agences de perception de l'impôt et les tiers chargés de la préparation des déclarations de revenus. Dans le modèle retenu par les auteurs, les contribuables sont incertains de leur revenu imposable et peuvent recourir aux services de fiscalistes dans le but de réduire cette incertitude. Les auteurs se penchent sur le caractère indicatif du choix du fiscaliste (en tenant compte de l'influence qu'exerce l'agence de perception sur le comportement) et se demandent s'il convient d'encourager les contribuables à recourir aux fiscalistes ou de les en dissuader au moyen de différents mécanismes de dégrèvement fiscal. L'étude démontre que le dégrèvement fiscal permet au législateur de mieux faire face aux conséquences des stratégies de déclaration des contribuables. L'on constate que la modification des taux d'imposition ou des pénalités ne livre pas les mêmes résultats que la modification des taux de dégrèvement; l'État vise généralement la ≪ discrimination en fonction du prix ≫ en subventionnant le recours aux fiscalistes dans le processus d'observation fiscale. Il ne serait pas optimal de permettre à tous les contribuables de se prévaloir d'un crédit d'impôt pour les honoraires des auteurs de leurs déclarations; une telle subvention devrait être refusée à certains contribuables. En outre, si l'État se voit contraint d'adopter un programme de dégrèvement identique pour tous les contribuables, il constatera dans bien des cas qu'une politique ne permettant aucun dégrèvement est supérieure, au sens de Pareto, à toute politique de dégrèvement uniforme.

While empirical evidence alludes to the intertemporal nature of corporate voluntary disclosures, most of the existing theory analyzes firms' voluntary disclosure decisions within single‐period settings. Introducing a repeated, multiperiod, disclosure setting, we study the extent to which firms' strategic disclosure behavior in the past affects their prosperity to provide voluntary disclosures in the future. Our analysis demonstrates that by voluntarily disclosing private information firms make an implicit commitment to provide similar disclosures in the future, and therefore are less willing to voluntarily disclose information in the first place. This effect is expected to be of larger magnitude for firms (1) with a long history of absence of voluntary disclosures and an impressive past operating performance, or (2) that operate in a relatively stable and predictable business and information environment, or (3) whose managers have a long time horizon and a high degree of risk aversion.

As different activities cannot be measured or communicated with the same precision, accounting information is often only a partial and unbalanced reflection of the fundamental economics, emphasizing certain aspects of the underlying operations while disregarding others. We highlight this inherent imbalance in information as the source of an interaction between corporate operating and discretionary disclosure strategies, and thereby also as an important determinant of the information acquisition strategy. We demonstrate that information imbalance, via its distorting effect on operating activities, leads to a reduction in the propensity of managers to acquire information and provide voluntary disclosures.