February 26, 2021
Mr. David R. Bean
Director of Research and Technical Activities
Governmental Accounting Standards Board
401 Merritt 7, PO Box 5116
Norwalk, CT 06856-5116
RE: Project No. 4-6P
Dear Mr. Bean:
PricewaterhouseCoopers LLP appreciates the opportunity to comment on the GASB’s Preliminary Views, Revenue and Expense Recognition (the PV). We commend the Board for taking on this important project and in particular, for exploring the “performance obligation” concepts that the FASB, IASB, and the International Public Sector Accounting Standards Board (IPSASB) have incorporated (or are considering incorporating) into their respective guidance. Our primary points are summarized below, and we provide further detail on these and additional observations in the Appendix.
In our view, the first question to determine categorization should be whether a transaction is an exchange of “something-for-something.” If it is, then we believe the performance obligation recognition approach, as proposed in the PV, should be applied; otherwise, GASB 33, Accounting and Financial Reporting for Nonexchangce Transactions, would be applied. We support this approach because we do not believe that nonexchange transactions contain the types of obligations that the performance obligation model is meant to address. For example, we do not view a requirement to spend resources for allowable purposes under expenditure-driven grants, when the resource provider is not receiving any direct benefit, to be a performance obligation. We therefore request that the Board reconsider the proposed Category A/B model. As discussed in the Appendix, we believe that the proposed model results in certain negative outcomes that would not be encountered in our preferred exchange-nonexchange model.
Since the Board has previously rejected categorization based on the exchange-nonexchange model, our comments in the Appendix also address recommendations related to the categorization and recognition approach proposed in the PV. Overall, we believe that the proposed model is workable (although, the generic nature of the naming convention used to distinguish categories is confusing) and we support use of the performance obligation recognition approach for Category A transactions when the resource recipient has a performance obligation for the direct benefit of the resource provider. We do not believe that expenditure-driven grants should be included in Category A or that the proposed recognition model for those grants is sufficient since it ignores eligibility requirements beyond incurring qualifying expenditures.
Over the past year, questions have arisen over the proper recognition of expenditure-driven grants created by the CARES Act. We do not believe the recognition model proposed in the PV would address those questions. We also do not believe that any contributions should be included in Category A because the costs of applying the performance obligation recognition model to these arrangements will outweigh any benefits realized.
The PV also presents an incomplete model for revenue and expense recognition. Based on the published timeline, the first opportunity many stakeholders will have to evaluate the complete model will be when the exposure draft is issued in mid to late 2023. Considering the widespread impact to financial reporting of governmental entities and the complex areas still to be deliberated, we encourage the Board to consider issuing another Preliminary Views document containing the complete model. This would give stakeholders more time to consider the decisions made by the Board, to provide quality feedback, and to raise implementation issues to be addressed in the exposure draft.
* * * * *
Reconsider the categorization approach
We believe that the exchange-nonexchange distinction, with certain modifications, should be the basis for categorization of revenues and expenses. We believe that revenues and expenses deemed to be exchanges should be subject to the performance obligation recognition approach proposed in the PV and those deemed to be nonexchange should be subject to GASB 33. We believe the exchange-nonexchange approach is more intuitive than the proposed Category A/B approach and would ensure consistency in accounting and financial reporting between public-sector and private-sector entities, particularly for exchange transactions.
We believe the current problems with the exchange-nonexchange distinction could be substantially fixed by de-emphasizing the notion of “equal value” being exchanged and instead focusing on the notion of “something for something.” For transactions when the exchange-nonexchange distinction requires judgment, such as expenditure-driven grants, we believe the Board should clarify the identification of an exchange transaction by leveraging the research performed by the FASB in the development of ASU 2018-08
. While application of the exchange-nonexchange model requires judgment, the same is true for any model, including the one proposed in the PV. However, the model proposed in the PV is less intuitive than current guidance. As a result we do not believe the benefits outweigh the costs.
In public comments, the GASB staff have used the example of a bottle of water to explain why the exchange-nonexchange model notion of “equal value” is problematic. In the example, the staff indicated that an individual may be willing to pay different amounts for a bottle of water depending on where they are purchasing it or their circumstances at a particular point in time (i.e., the differences in price are reflective of the value the individual places on the bottle of water at that point in time rather than the objective value of the bottle of water). In our view, it is inappropriate to assume that this transaction could only be classified as an exchange based on the objective value of the bottle of water. Whether a person buys a bottle of water for $1 or $5, an exchange has clearly occurred because both parties willingly received something and sacrificed something. We acknowledge, however, that this example is a market-based transaction and that governments do not always engage in transactions for which pricing is based on the open market.
The public-purpose mission of government further complicates the determination of whether a transaction is an exchange. For example, many citizens receive potable water via governmental entities for prices not based on market rates. Similarly, public transportation is heavily subsidized for public policy purposes so that the price a customer pays for a bus or train ticket is significantly less than the fair market value of the service provided. Again, we believe these examples are only problematic if an exchange is defined within the context of equal value or fair market value. In both of these examples, there is no question that an exchange of “something-for-something” has occurred between the two parties to the transaction.
The exchange-nonexchange approach is currently in use by private-sector and not-for-profit entities (NFPs) and does not result in significant inconsistency in revenue recognition among those entities. While many private sector entities set prices based on open market transactions so that price may be more reflective of fair market value, we do not believe this is an essential aspect of an exchange definition for the public sector; the “something-for-something” notion is sufficient.
Similar to governmental entities, NFPs sometimes engage in exchange transactions at prices that are not established in open markets. While judgment is required to determine whether some of these transactions are exchange or nonexchange, the FASB has provided clarifications to the exchange-nonexchange distinction that have reduced inconsistency among NFPs without significantly changing the basic categorization model.
In ASU 2018-08
, the FASB established guidance that has improved the exchange-nonexchange model, particularly for grants. We encourage the GASB to consider whether similar guidance would improve its exchange-nonexchange guidance. Consistent with the PV’s third model assumption (“for accounting and financial reporting purposes, the government is an economic entity and not an agent of the citizenry”), ASU 2018-08
states that a resource provider, such as the federal government, is not synonymous with the general public. We believe this guidance would aid governments in evaluating the exchange-nonexchange distinction for many grants, including many expenditure-driven grants, because the resource provider in many of those arrangements does not receive anything of significance for its direct benefit in exchange for the grant funds; the benefits flow to the resource recipient, the general public, or both. In keeping with the notion of a direct exchange of value occurring between two parties, we suggest that the model clearly indicate that in situations when the federal government funds activities that will benefit the general public, a direct exchange has not occurred. In other words, the federal government’s “outsourcing” of activities that will benefit the general public does not equate to a benefit received by the federal government itself.
There is legitimate debate about whether certain expenditure-driven grants (for example, sponsored research grants from the federal government to a public university) are exchange transactions. Some view these grants as exchange transactions because the federal government has essentially outsourced its responsibility to conduct research. Others view them as nonexchange transactions because the federal government usually does not retain any intellectual property or other rights to the results of the research. Ultimately, the guidance in ASU 2018-08
classified many of these grants as nonexchange, and therefore they are not recognized using the performance obligation approach in ASC 606
. We support a similar conclusion for governmental entities because of the difficulty in applying the performance obligation approach to most grants (a result similar to the GASB’s decision to exempt expenditure-driven grants from the performance obligation approach despite being classified in Category A).
Judgment will always be required to determine whether certain transactions are exchange or nonexchange. The GASB seems to have pointed to the judgment required in the exchange-nonexchange model as justification for moving to the Category A/B approach (PV chapter 1, paragraph 11); however, application of the A/B approach also requires judgment. For example, in step 3 of the model, determining whether rights and obligations are substantive will require judgment. The PV’s illustration of a lapel pin being provided in exchange for a $100,000 contribution has an objectively clear answer, but there are real-world government transactions when the answer is much less obvious.
To better understand the judgment required in step 3, we looked to some of the nonauthoritative examples within GASB 33 to stress-test the model proposed in the PV. In example 20 in GASB 33, a corporation makes a $5 million donation to build a classroom building at a community college but requires its name to be prominently displayed on the building. The question is how to determine if the naming opportunity is a substantive obligation in relation to the right to the donation. Consider an alternative fact pattern in which $5 million is donated for construction of a new football stadium at a public university in exchange for which the corporate donor requires that its name be prominently displayed on the stadium facing a major highway. Because of the greater prominence, some may consider the naming related to the stadium to be more substantive than the community college example. This might change if the donation amount was $50 million instead of $5 million.
In these fact patterns, determining whether the rights and obligations are substantive in relation to each other is at least as complex as determining whether these are exchange or nonexchange transactions. Therefore, we do not believe that the A/B approach will necessarily lead to less judgment or less diversity in practice. In particular, for transactions for which the exchange-nonexchange distinction under current guidance is unclear, we believe the PV is replacing one area of significant judgment (i.e., is the arrangement exchange-like?) for another (i.e., are the rights and obligations substantive?).
If the Board decides to retain the exchange-nonexchange model, we recommend removing the “exchange-like” concept. Private-sector entities are required to bifurcate transactions that have both an exchange and nonexchange component and account for each component under the applicable guidance (for example, the exchange component is accounted for under ASC 606
and the nonexchange component is accounted for under IAS 20 or ASC 958-605
). The entity first determines the fair value of the exchange component and assigns any remaining value to the nonexchange component. Judgment is required in order to assign value to the exchange, but we think this model results in an accounting answer that better reflects the economics of the transaction.
As a result of rejecting the exchange-nonexchange distinction, the proposed categorization model in the PV results in expenditure-driven grants and certain types of contributions (we refer to these as “conditional contributions” throughout this Appendix) being classified within Category A. We do not support this result. For conditional contributions, classification within Category A results in the application of the performance obligation approach; as discussed below, we believe the costs of recognizing conditional contributions under the performance obligation approach far outweigh the benefits. Retention of the exchange-nonexchange model would not have led to this undesirable result since all contributions are inherently nonexchange transactions.
In addition, although the Board has not fully concluded on all of the elements of the Category A/B model, we are curious about its future direction relating to the allocation of transaction price. Chapter 6, paragraph 15 of the PV indicates that the Board intends to require that the transaction price be allocated to identified performance obligations. In the private sector, the allocation is based on the value of each obligation. However, the Board rejected the exchange-nonexchange categorization model due to the perceived difficulty in assessing value. It would appear difficult to make a reasonable allocation based on something other than value. While we cannot anticipate the model the Board is considering, we question if the costs of creating a model based on something other than value, as well as the costs of applying and auditing that model, will be justified.
If the Board ultimately endorses the categorization and recognition model proposed in the PV, please see our comments below.
Categorization model is workable, but requires clarification
We believe the categorization model set forth in the PV is workable, but that steps 3 and 4 will require significant judgment for certain transactions. Therefore, if the Board moves forward with the Category A/B model, we recommend the Board provide further guidance about how to determine if rights and obligations are substantive and whether they are interdependent. We provided an example above illustrating how it may be difficult to determine if rights and obligations are substantive in relation to each other. In Chapter 1 of the PV, paragraph 9 states that “...the Board’s preliminary view...is that relying on value as a categorization tool is unreliable because value reflects the subjective perceptions of the parties to the transaction.” We are unclear on what basis could be used to determine whether rights and obligations are substantive in relation to one another without relying on some kind of value. We recommend the Board provide further guidance on how to think about step 3. Otherwise, we believe preparers and practitioners will default to value because it is the most intuitive way to perform the assessment.
Similarly, we believe step 4 needs further clarification. In the PV, interdependence is discussed in the context of “providing goods and services” (Chapter 4, paragraph 28). Unlike the example in the PV (a grant to a transit authority to provide services to the elderly), many expenditure-driven grants do not involve the provision of goods and services to third parties. Many such grants are for the governmental recipients' own activities. A grant to a public safety agency to acquire equipment for its employees or a grant to a public university to perform research on climate change does not result in the specific “provision of goods and services.” When a grant is to be used for the recipient’s own benefit, it may be difficult for some to conclude that there is any obligation in the arrangement, and therefore may conclude that the grant is purpose-restricted rather than expenditure-driven. We do not view the simple act of spending money in accordance with the terms of the grant as an obligation, and it is unclear how a research grant to a public university (which are classified as expenditure-driven grants) differs from the purpose-restricted donation in Case 13 in Appendix C. We recommend additional clarity from the Board, especially if the Board believes that entitlement to grant funds is dependent on the conduct of a specific activity, regardless of whether that activity involves the “provision of goods and services.” Without this clarification, we believe there will be diversity in practice.
Support for the performance obligation recognition approach
We support the Board’s decision to apply the performance obligation recognition approach to most Category A revenues (see our comments below on the proposed practical expedient for expenditure-driven grants and application of the performance obligation recognition approach to conditional contributions). We believe most Category A revenue streams of governmental entities are similar to the revenue streams of private-sector business entities and NFPs accounted for as exchange transactions, which are now recognized using the performance obligation approach. When there are no significant differences between governmental and private-sector transactions, our view is that the accounting for those transactions should be the same. We acknowledge that the Board made some changes to the model applied by private sector entities (for example, use of the term “binding arrangement” instead of “contract” and removing the evaluation of collectibility from the categorization model), but we believe these modifications are necessary and appropriate for the governmental environment and do not represent significant departures from the model applied by the private sector.
With the exception of the comments below on expenditure-driven grants and conditional contributions, we recommend the Board continue to develop a recognition and measurement model similar to ASC 606
. This is particularly important to governmental entities that compete directly with their private-sector counterparts for resources. For example, governmental and not-for-profit health care entities compete for capital in the municipal bond market. Health care entities typically issue revenue bonds, rather than general obligation bonds, and revenue metrics are a primary focus of both investors and analysts. Today, governmental and not-for-profit health care entities use the same revenue recognition model, so the revenues reported in their financial statements are comparable. We believe this consistency should be maintained.
Categorization and recognition for expenditure-driven grants
We do not support the categorization of expenditure-driven grants as Category A. As discussed above, we do not believe that all expenditure-driven grants contain performance obligations in the traditional sense. Many of these grants provide resources that directly benefit the resource recipient rather than requiring the recipient to provide goods or services to others. We do not view a requirement to spend money, without benefit to the counterparty, to be a performance obligation. The fact that the Board proposed an exception to the performance obligation recognition approach for expenditure-driven grants seems to indicate that the Board also felt that the performance obligations in those grants differ significantly from obligations in other Category A revenue transactions.
Rather than classifying expenditure-driven grants in Category A and then providing a practical expedient to exempt them from Category A recognition, we recommend that the Board allow all expenditure-driven grants to be included in Category B as a practical expedient. Ultimately, we believe that GASB 33, with some improvements, provides a suitable recognition model for those transactions. Alternatively, the Board could indicate that the act of spending money is not deemed to be a performance obligation, in which case expenditure-driven grants would fail step 4 of the model and be included in Category B.
In the PV, the Board proposed that governments recognize expenditure-driven grants as expenses are incurred. While this guidance exists in GASB 33 today, under GASB 33, arrangements may be subject to multiple eligibility requirements that all must be met prior to recognition. We believe the proposed practical expedient could lead governments to recognize grant revenue before they are entitled to that revenue. A grant may be reimbursable as expenses are incurred but may also contain other contingencies that must be met before the government is entitled to the funds; the model proposed in the PV would allow recognition based solely on incurring eligible expenses. We do not support a model that fails to take into account the need to satisfy the conditions of the grant prior to recognition.
Over the past year, a number of questions have arisen regarding the recognition of grants made under the CARES Act. While many of these programs are unusually complex, they serve as good examples to stress-test the proposed model. The Higher Education Emergency Relief Fund is an expenditure-based grant, but certain tranches contain other contingencies that must be met before the institution is entitled to the grant. For example, institutions not only must have incurred eligible expenditures for pandemic-related activities, but also must have disbursed a similar amount of funds to their students. The Provider Relief Fund reimburses for certain lost revenues caused by the pandemic; the Board was clear in Technical Bulletin 2020-1 that it does not view lost revenue as akin to incurring eligible expenditures. Additionally, the federal government imposed a specific measurement period over which healthcare entities must evaluate eligible expenditures and lost revenues, which some have viewed as a contingency under GASB 33. We do not believe that the practical expedient for expenditure-driven grants proposed in the PV results in an accounting answer that reflects the nature of these grants.
Categorization and recognition for conditional contributions
Although not clearly addressed in the PV, our understanding is that certain conditional contributions could be considered Category A if the rights and obligations are substantive and interdependent. We oppose this result. Contributions are inherently nonexchange transactions and therefore it is counterintuitive to think that any contribution would be subject to a performance obligation model based on the provisions of goods and services. Attempting to account for a contribution under the performance obligation recognition approach would be significantly more costly than under GASB 33 today and would yield no added benefit. If the Board moves forward with the Category A/B approach, we recommend that the Board exempt all contributions from Category A. At the very least, all contributions must be exempted from the performance obligation recognition approach.
If conditional contributions are Category A, there will be significant costs to preparers to determine the performance obligations in the binding arrangement. The PV indicates that for Category A transactions, governments need to determine the number of performance obligations in a binding arrangement and then allocate transaction amounts to each obligation; however, the GASB has not proposed an allocation methodology. We believe the determination of performance obligations and the relative allocation to each in a contribution would require significant judgment. While a government may indeed need to perform certain activities before they are entitled to a contribution, the binding arrangement rarely assigns specific contribution amounts to those individual activities. Donors and recipients may have vastly different views about the relative importance of, or effort required for, each activity. Even in cases when two identical contributions are received with identical performance obligations, the two donors may have different views about the relative value of the activities stipulated in the agreements.
While this is just one potential difficulty in applying the performance obligation recognition model to conditional contributions, we believe it illustrates the additional costs governments would incur to account for contributions. We think GASB 33 is a more efficient and intuitive model for conditional contributions.
Suggested improvements to GASB 33
Because we believe that expenditure-driven grants should be classified in Category B, we recommend that the Board maintain the current guidance in GASB 33, paragraph 20c on reimbursements. It is unclear in the PV if the Board intends to retain GASB 33, paragraph 20d as currently written or if the Board plans to amend that guidance as the project progresses. We feel that clarifying contingencies would help governments identify eligibility requirements in grants and contributions.
The distinction between eligibility requirements and purpose restrictions has long been problematic. While we do not believe that problem can be fully solved, we recommend that the Board amend GASB 33 to focus on the notion of “entitlement” as the differentiator between eligibility requirements and purpose restrictions. If a government must do something before being entitled to the grant, that is an eligibility requirement. If the government is entitled to the funds but must use them for a specific purpose in the future, that is a purpose restriction. Again, we encourage the Board to leverage the research performed by the FASB in developing ASU 2018-08
, which helped clarify the distinction between conditions and restrictions.
As the GASB moves forward with its deliberations on variable consideration in Category A transactions, we recommend the Board consider whether similar guidance should be added to GASB 33. The CARES Act contained grants for which a recipient could reasonably assess whether they met eligibility requirements at a point in time, but there was uncertainty in how much of the grant they would receive for having satisfied those requirements. Although grants of this nature are rare, we believe adding guidance on variable consideration to GASB 33 would provide governments direction in accounting for such arrangements in the future.
Expense categorization and recognition
We have focused our comments on revenue throughout this letter because we strongly believe that governments need enhanced guidance in that area, particularly as other standard setters have adopted the performance obligation approach. While we do not feel there is as much of a need for guidance in the area of expense recognition, we think application of the performance obligation recognition approach to Category A expenses is feasible. As indicated above, and for similar reasons, we would prefer the model be based on an exchange-nonexchange distinction rather than Category A/B.
Because there is not as much of a need for additional guidance on expense recognition, we suggest the Board consider whether the costs of adopting this model outweigh the benefits. A government’s expense transactions tend to be significantly more varied than their revenue streams, which would require governments to perform numerous assessments for expense categorization and recognition. Additional time will be incurred by preparers to document these assessments and for auditors to test them. Since we don’t expect application of the recognition model in the PV to significantly change expense recognition, we question whether these additional costs are justified.
Linkage to the Financial Reporting Model
We will file a separate comment letter on the GASB’s Exposure Draft, Financial Reporting Model Improvements (FRM). Many governments are just as concerned about the presentation of their flows statement as they are about recognition and measurement of the activities in that statement. Governments involved in business-type activities, many of which compete for resources with private sector entities, are particularly concerned about whether revenues and expenses are presented as operating or nonoperating. It is difficult to assess the changes proposed in FRM without a clear understanding of the results of this project, particularly since there are significant issues still to be discussed on revenue and expense recognition. We also do not believe that there is a pressing need for issuance of FRM and therefore recommend that the Board combine the two projects. We think a single project that provides a holistic view of revenue and expense recognition, measurement, and presentation would be beneficial to many stakeholders.
We encourage the Board to include as much implementation guidance in the standard as possible, including illustrative examples (similar to the examples provided in GASB 33). In recent years, the Board has issued standards on complex topics, such as fiduciary activities and leases, but has not provided implementation guidance or examples in the standards. Subsequently, the Board has issued implementation guides to help governments apply the standards. We believe this has the following unintended consequences:
- It reduces the number of governments willing to early adopt the standard. Expecting that an implementation guide will eventually be issued makes governments reluctant to early adopt. Management may make a decision that is subsequently contradicted by a future implementation guide, leading to possible restatement of the government’s financial statements.
- It indirectly shortens the transition period that governments have to adopt the new guidance. A standard that requires significant implementation guidance may indicate that it was not fully operational as issued. Since it can take a year or more for an implementation guide to be issued after the issuance of the standard, this inherently provides less time for the government to evaluate the impact of the guidance before the required adoption date.
Considering the extended timeline of this project and the abundance of lessons learned from other standard setters that have already issued new revenue recognition guidance, we believe that implementation guidance and examples can efficiently be included as part of this project, rather than as a separate future project.