Ten years ago, while I was acting chief accountant at the Securities and Exchange Commission, one of our biggest concerns was how to reduce the complexity in financial reporting. While other issues have come and gone in the time since, the idea that U.S. Generally Accepted Accounting Principles should be simpler has continued to resonate with companies, auditors, regulators, standard setters, and even financial statement users.

A decade ago, we thought that complexity could be addressed best in the development of new standards, by ensuring that standards were principles-based and that we didn’t develop a special accounting model for every new transaction. And the Financial Accounting Standards Board has generally done that, with some success. The new revenue recognition standard, for example, replaces more than 450 pages of guidance in the Accounting Standards Codification with around 275.

More recently, FASB has been attacking complexity directly, through the simplification initiative that began just about a year ago. The initiative targets little things that make reporting more difficult without providing much (or any) benefit to financial statement users. It differs from previous efforts to simplify accounting standards in that simplification is the driving force behind the projects in the initiative.

There have already been some noticeable achievements and a few more that are likely to be finalized soon, including:

Elimination of extraordinary items.

Classification of all deferred tax assets and liabilities as long-term instead of based on the related financial statement items.

Presenting debt issue costs as a reduction of the debt instead of a separate asset.

Allowing share-based payment expense to be recorded without a forfeiture assumption, with forfeitures recognized only when they occur.

Continuing to allow stock options to be classified as equity awards even if the company allows employees to request that the company withhold shares to cover estimated withholding taxes at any marginal tax rate, instead of just the minimum rate as currently allowed.

Eliminating the requirement that the equity method of accounting be applied retroactively when the conditions for its application are met sometime after the investor first invested in the investee.

These are all useful steps toward reducing complexity without diminishing the usefulness of financial statements. FASB deserves credit for undertaking this initiative and pursuing it effectively, as well as for the simplification achieved in revenue recognition and other standard-setting projects.

Despite all of this progress, however, hurdles to simplification are apparent, and they aren’t new at all. I’ve written about them before, in particular in columns published here in November 2007 and March 2012. That they continue to arise shows how we wound up with complex GAAP in the first place, and why simplifying it is anything but simple.

Sometimes We Like Complexity

Sometimes, reporting companies view a standard that, on its face, looks more complex than the alternative, actually as the simpler one to use. I was personally reminded of this late last year after one of my columns ran in this space. I questioned in that column how it could be that both FASB and the International Accounting Standards Board were touting their tentative conclusions in the lease project to be simpler than the other’s. I wondered how FASB’s proposal, which has two models to account for leases rather than a single model, could be simpler than the single model pursued by IASB.

Several groups interested in the project contacted Compliance Week or me directly to argue that simplicity in this situation should be measured based on the difficulty of the change. That IASB’s model would be simpler did not, in these groups’ view, overcome the fact that FASB’s tentative conclusions will have lower transition costs.

The same issue has been cited with respect to potential changes that would simplify the accounting for tax benefits from stock option exercises and transfers of assets between companies in the same consolidated group. While the potential new guidance is easier to understand for users, and should be easier to apply for companies, many are against them, stating that they have already developed the systems needed to implement current GAAP and would rather not incur the cost of change for such small items.

Despite 10 years of focusing on complexity, we’re still dealing with the same underlying causes.

All three of the items I mention above that have garnered objections would also result in more volatility in net income if they were adopted. Of course, companies generally prefer less volatility in earnings and are often willing to deal with complex accounting to achieve that. As I noted in this space eight years ago, hedge accounting is one of the most complicated accounting models there is, and it is completely optional. Yet companies do it because they believe the matching and smoothing that hedge accounting provides results in more faithful presentation of their results. In other areas, such as pension accounting, companies choose complex accounting models that smooth earnings over time, under the theory that the effects of long-term decisions are more fairly recognized over that long-term, even though the resulting accounting does not result in a better presentation of the entity’s financial position.

Sometimes We Can’t Help Ourselves

As I noted earlier, the new standard on revenue recognition achieved a good deal of simplification, largely by stating principles and then generally sticking with them throughout the document. But now FASB is proposing to pursue a few changes to the standard that show one of the most common ways that complexity gets into GAAP.

With respect to “minor” promised goods or services, FASB intends to provide specific guidance on when they can be disregarded, instead of just relying on materiality. As I’ve written about before, there is concern that without a special exception, companies and auditors will need to perform non-value-added work to quantify the extent of these immaterial promises that aren’t accounted for.

With respect to situations where the seller arranges shipping of goods to the customer, FASB is intending to provide specific guidance on the accounting, rather than letting the general principles apply. Many asked for specific guidance on this matter, as they believe the standard as issued might result in difficult accounting or diversity in practice.

And with respect to determining whether sales taxes are an obligation of the company or the customer, FASB intends to provide an exception so that companies could just treat them all like pass-through items. This would be done to alleviate concerns that determining who the tax was really on in the many jurisdictions in which such taxes exist might be time-consuming and require judgment that not be justified because the matter is not all that important.

Each of these little changes is understandable and justifiable when considered in isolation. But so were all of the other exceptions, specific models, alternatives, and accommodations ever put into GAAP. If we keep adding these things every time we get a principles-based standard, we won’t solve the problem of complexity.

The revenue examples are particularly troublesome, as the standard is converged with International Financial Reporting Standards, and constituents outside of the United States do not appear to have the same problems as we do. As I wrote three years ago, if we can’t apply standards in a reasonable manner without specific guidance, we’re destined to continue to ask for, and to get, far more detail in standards than we really need. This won’t serve us well in the long run.

What’s Next

FASB has some difficult decisions to make about reducing complexity and keeping it from creeping back in. From one angle, the progress made by the simplification initiative to date has been (and will be) substantive even if those proposals that are getting some pushback are abandoned, so it might well be reasonable to leave those projects for consideration another time. But that feels to me like it would be giving up too easily. When the problem is transition costs, it seems that the long-term benefits would usually outweigh the short-term pain.

With respect to volatility, the considerations are more difficult. If the volatility that would result reflects real economic events, it seems like that potential couldn’t justify not making the change. If, however, the volatility would result from an accounting mismatch or anomaly, then the existing complexity would appear to be serving a purpose, and a broader project seems appropriate. Of course, deciphering which of those is really at play in a given situation is not easy.

Also not easy for FASB is figuring out when to give in to calls for exceptions, specific guidance, and accommodations in the name of reducing diversity, making the application of the standard easier, or just saving time, at the cost of making the standards more complex, risking further problems down the road.

While these issues aren’t easy, they also aren’t new. Despite 10 years or more of being focused on the issue of complexity, we’re still dealing with the same underlying causes. Simplification clearly isn’t simple.