Response to the GASB's Invitation to Comment on Revenue and Expense Recognition: Project No. 4-6I

The American Accounting Association's Accounting and Auditing Standard's Committee of the Governmental and Nonprofit Accounting Section is charged with responding to requests for comment from standard setters on issues related to governmental and nonprofit financial reporting. This commentary provides the Committee's response to the GASB's Invitation to Comment on Revenue and Expense Recognition: Project No. 4-6I (hereafter, the ITC).

The Invitation to Comment includes a discussion of the project's scope and discusses the components of a comprehensive revenue and expense recognition framework introducing two models—an exchange/nonexchange model and a performance obligation/no performance obligation model. The ITC provides a discussion of the benefits and costs of the two models and includes definitions provided in existing guidance for exchange and nonexchange transactions in addition to the existing recognition approach for nonexchange transactions in Statement 33, as amended. The ITC includes the potential further expansion of the earnings recognition approach that would be applicable to the exchange/nonexchange transactions model.

The ITC follows with a description of the performance obligation/no performance obligation model. Again, the discussion includes a potential definition of a performance obligation for the governmental environment, as well as a potential recognition approach that employs the performance obligation model. Appendices with illustrative examples and background information pertinent to the project are provided for both models.

In addition, the ITC suggests that an alternative model may be suitable for governments. The GASB seeks stakeholder input to develop a comprehensive framework for revenue and expense recognition to develop a model that would provide robust guidance to address a wider range of transactions and improve comparability of revenue and expense recognition practices across various types of governments for more useful information to users of financial reports.

The Committee's response submitted to the GASB's request for comment letters is provided below. Changes to the original manuscript are noted in square brackets to designate text that was added to improve clarity.

Our comments are based on perspectives from teaching governmental accounting, interfacing with accounting professionals in the public sector and governmental employers of our students, and researching governmental accounting issues for academic publications. We speak as individuals serving on the Accounting and Auditing Standards Committee for the Governmental and Nonprofit Accounting Section and do not represent an official position of the American Accounting Association. In addition, the comments reflect the overall consensus view [majority opinion] of the Committee, not necessarily the view of every individual member.

We recognize that complexities of governments present a challenge in developing a comprehensive revenue and expense model that could be applied in all real-world situations. However, we applaud the spirit of the GASB to question the base orientation for recognition of these complicated items and to work toward establishing a framework to improve comparability and provide more useful information. The following presents specific comments organized by the questions posed in the GASB Invitation to Comment, Project No. 4-6I.

The current exchange/nonexchange model provides a distinction between transactions in which essentially equal values are exchanged and those transactions that lack a reciprocal exchange. This model demonstrates that the activities of governmental entities do not always entail quid pro quo transactions (e.g., selling a service for a commensurate fee). It therefore helps emphasize the difference in the goals and objectives in the governmental sector, compared to the goals and objectives in other sectors. While readers of the current guidelines may sometimes interpret the accounting guidance differently, it is not clear that the conceptual method of classification has anything to do with the differences in interpretation. An example was provided in Appendix B of the ITC (page 28, B10) in which some survey respondents reported recognizing utility fees when a good or service was “sold” and others when an obligation to the customer was fulfilled. These areas of inconsistency in reporting could be handled by providing more detailed implementation guidance rather than adopting a completely different model.

The current model of revenue and expense follows an earnings recognition approach for governmental accounting. While Altamuro, Beatty, and Weber (2005) find evidence consistent with the FASB's assertion that reporting earnings based on performance obligations prior to completion of the earning process contains value-relevant information about future performance, this type of revenue recognition process increases transactions related to unearned revenue. In the governmental setting, this may be problematic and result in increased future adjustments and, thus, decreasing the informativeness [decision usefulness] of financial reports.

Developing a comprehensive revenue and expense model based on the existing exchange/nonexchange model has many advantages over a new performance obligation model. First of all, the comprehensive model would encompass guidance for exchange transactions that is presented in multiple pronouncements and expand the implementation guidance of the existing model. In addition, there would be less need to re-educate users and preparers in terms of the reporting structure in governmental units, minimal cost for implementation of a comprehensive model for revenue and expense transactions, and enhanced consistency and comparability to provide more useful information.

The exposure draft mentions leaving Statement 33 provisions intact for nonexchange transactions, while the discussion during the GASB webcast on March 14, 2018 suggested that the language in Statement 33 could be modified. We suggest that the language in Statement 33, with regard to grants and contracts, is the very area in which preparers have different interpretations of when revenue should be recognized and, therefore, needs to be further clarified. It is actually Statement 33 that is more in need of revisiting than the conceptual language of the exchange versus nonexchange compared to the performance versus no performance obligation. We address this in more detail under 3.1. and 4.2. below.

While other standard-setting bodies, in particular the FASB, are moving toward a performance obligation model, it is unclear that this move will provide more consistent, comparable data than the previous rules-based standards. The flexibility inherent in how contracts and revenue models are conceptualized by for-profit firms may allow firms to depict their revenue streams in a way that they think best shows the relationships between different performance obligations. Adopting a performance obligation/no performance obligation model for governmental accounting, and thereby basing the revenue recognition on the language in legal contracts rather than the overall intent of the agreement, is likely to lead to more gamesmanship of revenue recognition than the current exchange/nonexchange model. Even under the current model, governments often insert clauses that the contract could be rendered void if the legislature refuses to appropriate money for subsequent years. Current standards sometimes override this language if the likelihood of non-appropriation is low.

The performance/no performance model in the for-profit sector does not have a proven track record that the benefits will outweigh the cost to implement or that the potential lack of comparability will be improved across firms based on how contracts are constructed. As academic researchers, we hesitate to suggest that this new model would be value added since there has not been sufficient time to gather evidence to support a claim. In addition, shifting to a performance/no performance model would cause significant need for re-education of preparers, auditors, and users, which would likely add significant cost with no assurance of added benefit. There is no evidence at this time to suggest that a performance/no performance model would result in a more uniform interpretation of factual situations and some risk of significantly more variation. Given the potential risks and uncertainties that arise from applying the performance obligation/no performance obligation model in governmental entities, we do not recommend this basis to be used for classifying transactions and recognizing revenue and expense.

We would also express a concern with the terminology applied in a governmental setting. There are language nuances associated with the “no performance obligation” that could be perceived quite negatively in the governmental sector. Various types of taxes are now classified as nonexchange transactions and, as noted in the ITC (page 27, B9), the majority of state and local government revenues are generated from nonexchange transactions. When voters assent to an increase in their property taxes or sales taxes for a certain purpose, they see the government as having a duty to spend the money as specified or to perform certain activities with the funds collected. Under the current model, taxes are treated as revenues as soon as they are measurable and available for use in accordance with the budget. Categorizing these tax collections as having “no performance” obligation would be confusing, but categorizing the taxes collected as having a performance obligation might suggest their recognition should be deferred until spent. This kind of conceptualization could (1) pose some rather thorny issues of cost allocation that are not inherent in current practice, and (2) defeat the benefits of the dual reporting model wherein the modified accrual basis matches tax inflow to the appropriate budget year, rather than matching the inflows with the associated expenditure. In addition, as noted by Marton and Wagenhofer (2010), the legal existence of a claim may differ across jurisdictions, thereby creating inconsistencies in revenue reporting and diminishing the decision usefulness of information.

As we mentioned in our answer to Question 2.1, we believe that the language in Statement 33 with regard to grants and contracts could lead to different interpretations of when revenue should be recognized and needs to be further clarified. We think that the language in Statement 33 is more in need of being revisited than the conceptual language of the exchange versus nonexchange compared to the performance versus no performance obligation.

Based on our understanding, under this alternative model the exchange/nonexchange classification scheme would be retained, but with the details of the revenue recognition conforming conceptually with those of the FASB and other groups' performance/no performance models.

Even though using the alternative model can perhaps provide data that would be more similar to for-profit businesses for the exchange-like transactions, we would reiterate that the performance/no performance model is as yet still an unproven model. The profession has not had enough experience with this model to determine whether exchange-like transactions would be treated more or less similarly under the new principles-based guidance. Consequently, a change to the performance/no performance model, either in totality or piecemeal, would have similar risks and costs of implementation with no clear evidence of providing more comparable or meaningful information. Given the short-term nature of political actors in the governmental sector, any gains in flexibility might be perceived as providing less accountability to citizens.

We are not convinced that the current exchange/nonexchange characterization is broken. What needs more attention are the classifications in GASB Statement 33. The technical provisions for distinguishing reimbursement grants, purpose grants, and multiyear time-restricted grants lead to wide variation in interpretation based on slight differences in legal language rather than the underlying substance of these arrangements. Further, in most cases the voluntary and involuntary government mandate transactions are handled the same and do not warrant separate distinctions in GASB Statement 33. We suggest that the GASB further clarify the nonexchange standards and classifications rather than the exchange ones.

Specific alternative models have been explored in other revenue recognition forums. The FASB and the IASB have explored revenue recognition models including the customer consideration (similar to the performance obligation model) and the measurement model (fair value model) (Schipper, Schrand, Shevlin, and Wilks 2009). The fair value model would be extremely challenging to implement in a governmental setting but does bring to light some key considerations when it comes to measurement, a topic left unanswered with this ITC but an integral one to the final considerations when choosing a final model. Antle and Demski (1989) carefully delineate that any revenue recognition model with an accrual piece introduces variation in information reporting that requires others to be educated about the impact of these timing adjustments. Thus, as salient as the current debate between the two main options discussed in the ITC, carefully thinking about additional options may help the GASB determine if the strengths of the current model are relevant to the fundamental debate about needing to change the revenue recognition model for governments.

Throughout our letter, we express a preference toward enhancing the exchange/nonexchange model. However, we cannot truly give a conclusive opinion due to the lack of evidence that would indicate the superiority of the performance obligation/no performance obligation model over other revenue recognition models. In addition, the performance obligation/no performance obligation model still leaves unresolved some of the motivations behind the GASB's efforts in this revenue and expense recognition project. Transactions classified as no performance obligation would still be subject to the specialized guidance in Statement 33, which does not address the issues raised in ¶5 of Chapter 1 in the ITC document (this is also acknowledged in ¶25.c. in Chapter 3).

Another motivation for this project, subjectivity in the assessment of value, as described in ¶6 in Chapter 1 of the ITC, is also not resolved by the performance obligation/no performance obligation model. The determination of a transaction as a performance obligation includes the concept of a binding arrangement that “includes identifiable rights and obligations of the parties that articulate in equivalent terms” (¶3 and ¶5, Chapter 3, ITC; https://www.gasb.org/jsp/GASB/Document_C/DocumentPage?cid=1176169978401&acceptedDisclaimer=true). This component of the definition of a performance obligation is analogous to the component of the definition of an exchange transaction. The concept of “articulate in equivalent terms” is very similar to the concept of equal or like value, thus users would again face subjectivity in the assessment of value in addition to determining the other components that define a performance obligation. [The performance obligation model does not provide guidance on the concept “articulate in like terms.”] Because the performance obligation/no performance obligation model seems to address the absence of guidance for exchange transactions while not adding to the specialized guidance for nonexchange transactions, [nor resolving the “subjectivity in the assessment of value” or the “articulate in like terms” issue] a comprehensive model may include multiple models. As such, additional analysis of the value added by other revenue recognition models (including the performance obligation/no performance obligation model) is needed for our committee and the board to work toward improved transparency and consistency in reporting through the current revenue recognition model or a new framework comprised of several models.

We appreciate the GASB's desire to develop a comprehensive revenue and expense model for state and local governments. Thank you for your dedication to the standard-setting process and for allowing interested parties to provide valuable insights to create the best solution for future reporting.

Subsequent to the comment letter period closing, the GASB invited the Committee to discuss the comment letter at one of its public meetings. A committee representative participated in a hearing on May 18, 2018. The standard-setting process continues with GASB re-deliberations to consider definitions related to the classification process. The next phase will include deliberations related to revenue and measurement issues. A first draft of preliminary views is scheduled for February 2020.

Future research is needed to consider the relevant issues and help to inform the GASB as the board works toward establishing and improving standards for state and local governments to provide useful information to users of financial reports. The following suggestions for future research were developed by the Committee in response to the GASB's request for more information on practical issues to inform the standard-setting process:

• Can the exchange/nonexchange model be written in a conceptual way that is not rules based, but rather principles based?

• Do the differences in environment between accounting for governmental units and the private sector warrant a conceptual model for revenue and expense recognition that differs from the performance obligation model?

• Are the services that governmental units provide their citizens performance obligations?

• What types of governmental entity transactions contain a performance obligation? What is the estimated frequency and magnitude of performance obligation transactions within governmental entities?

• What are the characteristics of governmental entities that might have the greatest impact of a change to a performance obligation model?

• Governmental entities appropriate funds before incurring an expense/expenditure, thus there is a legal view to expense/expenditure recognition in addition to an economic view. How do governmental entities recognize expenses/expenditures from an economic versus legal standpoint?

• Exploratory research is needed to evaluate the potential advantages and disadvantages of the exchange/nonexchange and the performance obligation/no performance obligation models to inform the GASB regarding the perceptions and expectations of financial statement users.

• What disclosures regarding revenue and expenditure/expense recognition would be useful to provide clarity and comparability of financial reports?

With additional information and discussion that might arise from future research, the norms are challenged, and new ideas are allowed more space to develop. This is the best time to consider how the conceptual framework could be strengthened to develop a new model of revenue and expense recognition that can be applied to transactions today and provide a basis for the accounting needs of tomorrow in an ever-changing business and governmental environment.

More broadly, it is a privilege to have the opportunity to be involved in the standard-setting process. The Committee encourages researchers, educators, and all users of financial statements to get involved in the standard-setting process and provide input regarding potential impacts of changes to the standards.