For each year the taxpayer pays or incurs $100 in domestic R&E spending, a comparison of the full deduction in Table 1 with the total amortization deductions allowed for each year in Table 2 shows a significant discrepancy in the early years, with the totals converging only in Year 6 (and beyond). In the case of foreign R&E spending, for which the amortization period is 15 years rather than five, the difference between Tables 1 and 2 for the early years will be even more pronounced. And if the taxpayer’s R&E spending fluctuates significantly from year to year, the totals may never converge. Also, in the case of fluctuating expenditures, the amount of the amortization deduction in any given year may bear no relationship to the amount of R&E spending in that year, meaning that, in any given year, a taxpayer with a high out-of-pocket cost might have a relatively small amortization deduction, and vice versa. In contrast, under the deduction model of old section 174, the taxpayer’s annual deduction exactly matched the taxpayer’s R&E spending for that year.
As this illustration shows, if a company’s domestic R&E spending is fairly consistent over a six-year period, the near-term impact of the switch to amortization will be significant, but over the longer term the company’s amortization deductions will eventually approximate their annual deductions under prior law. For taxpayers with significant foreign R&E spending, the near-term tax effect will be even greater, and even if that spending is fairly consistent, it will take 16 years, rather than six years, before that effect diminishes. With respect to both domestic and foreign R&E expenditures, if a taxpayer’s R&E spending fluctuates significantly from one year to the next, the new amortization regime will have a flattening effect, as the tax impact of one year’s higher-than-usual R&E spending will be spread out over six or 16 years. Thus, in any given year, the taxpayer’s amortization deduction will bear little relationship to its actual R&E spending for that year. The difference between actual R&E expenses and allowable deductions will produce “feast” years and “famine” years, and businesses will have to plan accordingly.
1. Comparing the Amortization Periods
For taxpayers that chose to amortize rather than deduct their R&E expenses under prior law, the TCJA has introduced changes as well. Specifically, the TCJA has altered the section 174 amortization period. These alterations affect (1) the length of the amortization period, and (2) the date on which the amortization period begins.
Under old section 174, a taxpayer that amortized its R&E spending could choose any amortization period equal to or greater than 60 months. Compared to an immediate deduction or a fixed five-year amortization period, the flexibility to spread amortization deductions over a shorter or longer period enabled companies to choose the method best suited to their anticipated income stream This could be especially advantageous for a startup that did not anticipate having significant income in its early years. Under the new law, however, the amortization periods are fixed – five years for domestic R&E, and 15 years for foreign R&E. This lack of flexibility creates fewer opportunities for tax planning.
On its face, the new fixed five-year amortization period might seem comparable to the minimum 60-month period under prior law. However, that is not the case. As discussed below, there is a significant difference in the starting date of each amortization period.
Under old section 174, the amortization period of 60 months or more began when the taxpayer first began to realize benefits from the R&E expenditures. Generally speaking, this was deemed to occur when “the taxpayer first put the process, product, invention or similar property to which the expenditures relate to an income-producing use.” Because this event might occur years after the date of the R&E expenditure, even a taxpayer that opted for the shortest amortization period of 60 months might not be able to begin taking these deductions for several years. However, a taxpayer that preferred to take deductions sooner could simply choose to deduct rather than amortize its R&E costs. Here, again, the flexibility of the old version of section 174 could be especially beneficial to startups.
In contrast, under the TCJA, the fixed 5- and 15-year amortization periods begin at the midpoint of the year in which the taxpayer pays or incurs the R&E expense. In other words, the amortization period begins immediately. On the one hand, for companies that otherwise would have enjoyed a full deduction under old section 174, this earlier commencement for the amortization period arguably softens the blow of losing the deduction (although this mitigating effect would be negligible for taxpayers with foreign R&E expenses that were once deductible in full, but which are now subject to 15-year amortization). This “quick start” could be a small bit of good news for established companies with significant income streams. The quick start would not be advantageous, however, for a startup or other business that would benefit from deferring its amortization deductions into future years.
Thus, the combined effect of losing the ability to choose an amortization period greater than five years, and being forced to begin taking amortization deductions as soon as the R&E expense is paid or incurred, means that the new amortization schedule is significantly accelerated compared to that of prior law. For a startup or other business that has little income in the early years of its R&E undertakings, the TCJA substantially impairs the company’s ability to postpone amortization deductions until it has significant income to offset.
2. Extended Depreciation Schedules
Under new section 174, even though the cost of depreciable property is excluded from the statutory definition of “specified research and experimentation expenditures,” the actual amount of a business’s depreciation or depletion allowances for the taxable year is expressly included. This was also true under prior law. However, there is a significant difference. Under old section 174, a taxpayer could deduct its depreciation or depletion allowances according to the same schedule that would apply if the allowances were unrelated to R&E. However, new section 174 requires each annual depreciation or depletion allowance to be amortized rather than deducted in full. Thus, if a business has purchased property for use in its R&E activities that is subject to an allowance for depreciation under section 167, under new section 174 the business can no longer take annual depreciation deductions based on the equipment’s useful life, as it could under prior law. Instead, the annual depreciation deduction that would otherwise be allowed must now itself be amortized over the fixed 5- (or 15-) year term.
To take a simplified example, suppose that a taxpayer spends $1,000 to purchase equipment for use in its domestic R&E activities, and the equipment has a useful life of 10 years and zero salvage value. Under old section 174, if the taxpayer chose to amortize rather than deduct this expense, then using the straight-line method (and disregarding, for simplicity, the half-year convention), the taxpayer would be entitled to the normal depreciation deduction of $100 per year under section 167 (the general depreciation provision); the amortization schedule of 60 months or more under old section 174 would not apply. Under new section 174, however, this $100 of annual depreciation under section 167 is no longer deductible in full, but must itself be spread over 5 years, just like the taxpayer’s other R&E expenses for the year. As a result of this extended depreciation schedule, the taxpayer’s annual depreciation deduction is effectively reduced to $20 in the first year, $40 in the second year, $60 in the third year, topping out at $100 per year in years 5 through 10, and winding down again to $20 in year 14, turning what otherwise would be 10 years of straight-line depreciation into 15 years of extended and “curved” depreciation. Under new section 174, therefore, it will take this taxpayer 14 years to recoup its investment in the equipment, even if the equipment must be replaced at the end of its 10-year useful life. The “long tail” on each end of this curve—like the depreciation period itself—will be even longer if the capital expenditure is attributable to foreign research and thereby subject to 15-year amortization; in that case, the depreciation period for this 10-year property will be 24 years.
It is not clear whether Congress intended this result. The statutory language that now has the effect of requiring taxpayers to amortize their annual depreciation and depletion allowances was copied verbatim from the old version of section 174. In that context, the language functioned to make clear that the taxpayer’s annual R&E deduction could not include the entire cost of acquiring depreciable or depletable property, but could include the amount of the annual depreciation or depletion allowance for that property. Thus, the 1954 Congress made clear that the deduction for R&E expenses was not a free pass to take an immediate deduction for that $1,000 piece of equipment (to use the same example as above) just because it was going to be used in R&E activities. Old section 174 thereby prevented taxpayers from mismatching income and expense; because that $1,000 piece of equipment could be expected to benefit the taxpayer’s R&E activities throughout its 10-year useful life, it would conflict with fundamental tax principles to allow a deduction for the entire amount in the year it was acquired. While the 1954 Congress intended to encourage R&E activities, its generosity was not unlimited, and it apparently drew the line at ignoring the economic reality of durable property. In addition, if a business acquired depreciable property for use in its R&E activities, but opted to amortize rather than deduct its R&E expenses, it was not even allowed to amortize the property’s cost over the section 174 term of 60 months or more; instead, the property was subject to the normal depreciation rules of section 167. Under old section 174, therefore, the amortization schedule for R&E expenditures had no impact on the depreciation schedule for property acquired for use in R&E activities.
The opposite is true under the TCJA amendment. As demonstrated by the example above, the new rules have a significant effect on the treatment of depreciable property, compelling taxpayers to spread the costs of such property over a period that is significantly longer than its normal depreciation period, and which can therefore far exceed the property’s useful life. Did Congress actually intend to create this extended depreciation schedule? It can be justified, in principle, by the argument that even a normal annual depreciation deduction under section 167, if attributable to R&E activities, produces a benefit to the taxpayer over the same period of time as other R&E expenses, such as wages paid to R&E employees. Yet the complete absence of any discussion in the legislative history, combined with Congress’s verbatim copying from the pre-TCJA statute, suggests that Congress may not have thought deeply about whether this result was desirable, and may not even have realized that the new statute would have this effect.
Whatever Congress intended, the Service has embraced the extended depreciation schedule. The interim guidance issued in 2023 treats “cost recovery allowances” as R&E expenses that must themselves be amortized, expressly including “depreciation, amortization, or depletion allowances with respect to property used in the performance of SRE activities or in the direct support of SRE activities,” and providing multiple examples.
Ironically, then, the new rules mean that businesses are penalized for devoting depreciable assets to R&E activities. How might businesses respond to the extended depreciation schedule? Some may be reluctant to purchase costly new equipment that might facilitate their R&E activities, since this equipment will effectively cost them more on an after-tax basis when compared to prior law. Such companies may forego or postpone investing in new equipment, even if this makes it more difficult to achieve their R&E goals. Alternatively, they may choose to lease rather than purchase such equipment; although the lease payments themselves would have to be amortized, the company’s initial non-deductible expenditure would be less than the cost of purchasing the same equipment. A company that acquires such equipment to lease to other companies will, of course, be able to depreciate the equipment according to the normal section 167 depreciation rules, unimpeded by section 174, because it is using the equipment in the trade or business of leasing. A corporation engaging in R&E activities might even have its own corporate affiliate purchase such equipment and then lease it to the entity conducting the R&E.
An additional complication arises if a business uses depreciable property in both R&E and non-R&E activities. Under old section 174, this would not affect the depreciation schedule. Under new section 174, however, using the property in R&E activities will be subject to the extended depreciation schedule, while using the property in other trade or business activities will be subject to the normal depreciation schedule under section 167. Although the statute does not address this situation, the interim guidance indicates that taxpayers must allocate their allowances for depreciation, amortization, and depletion between the R&E and non-R&E activities in which they are using the property. Accordingly, the extended depreciation schedule will apply only to the portion allocable to R&E activities. As noted earlier, the interim guidance allows taxpayers to use any cost allocation method that “reasonably relates the costs to the benefits provided to the SRE activities,” provided that they adopt a consistent method within each category of costs. The only example that the guidance provides involves allocating facility cost recovery allowances for a building, where the allocation reflects the ratio of the square footage used in R&E activities to the square footage used in non-R&E activities. However, this relatively simple example glosses over potential complexities. For example, it assumes that distinct areas of the building are devoted exclusively to R&E activities while others are devoted exclusively to non-R&E activities. This may not always be the case, especially for newer enterprises working in cramped facilities. In addition, other types of cost recovery allowances—such as amortization deductions for intellectual property or depreciation deductions for equipment—do not lend themselves to an allocation based on square footage. Taxpayers will therefore need to find an alternative method for allocating the use of these assets, such as one that reflects the amount of time the asset is used in each type of activity. As in the case of labor costs, taxpayers will need to keep records to substantiate their allocations. Faced with such records, however, the Service will have difficulty challenging the taxpayer’s allocations.
3. Failure and Abandonment
Failure and abandonment are not unheard of in the world of R&E. If a taxpayer’s R&E efforts do not yield distinct economic benefits, or if those benefits are short-lived, the inflexible amortization period under new section 174 conflicts with longstanding tax principles governing loss or abandonment of assets, and therefore can create hardships for the taxpayer. Some R&E efforts may yield only short-lived economic benefits, such as an innovation that is quickly rendered obsolete by some other emerging technology. Alternatively, the R&E may reveal that an initially promising idea does not work at all, or that it works, but in a manner that is economically unfeasible for commercialization in the near term, so that, at best, the idea must be shelved indefinitely pending future developments that may or may not occur. In these situations, the difference in tax treatment under the old and new rules is striking.
Prior to the TCJA, R&E costs could be deducted or amortized without regard to their productivity. In addition, general tax principles allow taxpayers to deduct the unrecovered basis of any property that is abandoned or disposed of, or which becomes obsolete. For example, under section 165, a taxpayer may deduct an asset’s remaining basis in the year the asset is abandoned or becomes obsolete. Likewise, the regulations under section 167 (governing depreciation for patents and several other types of intellectual property) permit taxpayers to accelerate an asset’s depreciation schedule if it becomes obsolete sooner than expected, or to take a loss deduction if the asset is retired or abandoned.
Although the TCJA did not alter these general tax principles, it rendered them inapplicable to R&E expenditures. Even if the products of the taxpayer’s research expenditures are disposed of or abandoned before the costs have been fully amortized, new section 174(d) dictates that the inflexible 5- or 15-year amortization period still applies:
If any property with respect to which specified research or experimental expenditures are paid or incurred is disposed, retired, or abandoned during the period during which such expenditures are allowed as an amortization deduction under this section, no deduction shall be allowed with respect to such expenditures on account of such disposition, retirement, or abandonment and such amortization deduction shall continue with respect to such expenditures.
The result is a mismatch of income and expense, requiring the taxpayer to spread the amortization deductions over the fixed period, even though the R&E investment will not produce income during this period.
The interim guidance only addresses the application of this rule to situations in which a corporation ceases to exist for tax purposes. The tax treatment depends on the nature of the transaction (or series of transactions). In the case of a subsidiary liquidation or reorganization under section 381 (generally speaking, a carryover transaction), the acquiring corporation must continue to amortize the transferor’s unamortized R&E expenses for the remainder of the transferor’s section 174 amortization period, essentially stepping into the shoes of the transferor. In all other cases, the transferor can deduct the unamortized amounts in its final taxable year.
Notably, section 174(d) does not expressly address situations in which the R&E expenditure produces no “property” at all. Some research efforts will not produce any commercially useful asset. For example, the taxpayer may decide to abandon a research project before it produces anything of value. Alternatively, the research may produce information that has no immediate value, but may be useful at some indefinite point in the future. It may produce information of a negative type—for example, evidence that an initially promising idea does not work. In some cases, this “negative” information may have value as a trade secret, thus qualifying as “property” within the meaning of section 174(d). In other situations, it may have no value at all—for example, if key competitors already possess this information, or if the jurisdiction in question does not protect “negative” information as a trade secret.
Because section 174 does not recognize any exemptions from fixed-year amortization, it appears that even the R&E expenditures arising from research efforts that lead to a “dead end” must be amortized over the statutory period, even though the research may never produce any income at all. This, too, results in a mismatch of income and expense.
C. Unreasonable R&E Expenditures
Prior to amendment by the TCJA, section 174(e) subjected R&E expenses to a rule of “reasonableness,” stating: “This section shall apply to a research or experimental expenditure only to the extent that the amount thereof is reasonable under the circumstances.” According to the regulations, an R&E expenditure was reasonable “if the amount would ordinarily be paid for like activities by like enterprises under like circumstances.” Only if the amount was reasonable could it be deducted or amortized as an R&E expense. Amounts in excess of what was reasonable were subject to recharacterization as “disguised dividends, gifts, loans, or similar payments.”
However, the TCJA removed the reasonableness requirement. The legislative history does not indicate why. Without a statutory requirement of reasonableness, it is not clear whether the Treasury still has the authority to impose such a requirement by regulation. The absence of any expression of congressional intent in this regard will make this determination even more difficult.
Some experts have suggested that Congress may have affirmatively chosen to abandon the reasonableness requirement. According to this theory, because full deduction is no longer an option under section 174, Congress may have intended to compel taxpayers to capitalize even those R&E expenses that are unreasonable in amount.
It is difficult to understand, however, what rational tax policy goals would be served by permitting capitalization of unreasonable R&E expenditures. The only way this would generate a public benefit would be if the taxpayers were also prohibited from amortizing those capitalized amounts. In that case, the taxpayer could not recoup the capitalized amounts until it disposed of the research project or the entire line of business. By delaying or denying any tax benefit from the unreasonable expenditures, this approach would remove the motivation to incur such expenditures. However, nothing in the statute indicates that taxpayers are prohibited from amortizing their unreasonable R&E expenses. Thus, because section 174 requires taxpayers to amortize all of the capitalized amounts within five (or 15) years from the year the expenses are paid or incurred, permitting amortization of excessive R&E expenses will lead to excessive amortization deductions, which will eventually reduce the Treasury’s tax revenues.
This result might be avoided, however, if the Service invokes the doctrine of substance over form in order to recharacterize unreasonable R&E expenses. The Service’s past practice, under old section 174 and the accompanying regulations, was to recharacterize excessive R&E expenditures as dividends, gifts, loans, or other payments, depending on the circumstances. If the Service is permitted to continue applying this well-established doctrine to unreasonable R&E expenditures under new section 174, this would lead to results that better reflect the economic reality of the excess payments.
However, federal courts have held that the Service cannot invoke the substance-over-form doctrine when Congress, through its legislation, has expressly authorized the elevation of form over substance. Of course, new section 174 does not expressly authorize amortization of unreasonable R&E expenditures. On the other hand, because Congress did not explain why it removed the reasonableness requirement of prior law, and the reason for this deletion is not otherwise apparent, a taxpayer might take the position that Congress made an affirmative decision in the TCJA to abandon the reasonableness requirement for R&E spending, an argument that the above-mentioned experts have already suggested. If courts find this argument persuasive, then the Service will no longer be able to recharacterize unreasonable R&E expenditures according to their economic substance, and taxpayers will indeed be able to amortize unreasonable R&E expenses. This result could undermine the revenue-generating potential of new section 174.
The interim guidance issued by the Service in 2023 expressly acknowledges that the TCJA removed the reasonableness restriction. Nothing in the guidance suggests that the Service will use its administrative authority to reinstate such a restriction.
D. Interaction with Other Capitalization Rules
The TCJA’s requirement that all R&E expenditures must be capitalized and amortized over fixed periods also raises new questions about the relationship between new section 174 and the existing capitalization rules under sections 263 and 263A. These questions arise whenever the R&E expenses are incurred in connection with producing tangible or intangible property.
Section 263 requires taxpayers to capitalize the costs of making “improvements” to increase the value of any property, including intellectual property and other intangibles. However, under section 263(a)(1)(B), this requirement does not apply to “research and experimental expenditures deductible under section 174.” Section 263A, which requires capitalization (or treatment as inventory costs) of the direct and indirect costs of producing tangible property (known as the UNICAP rules), contains a similar restriction, exempting “any amount allowable as a deduction under section 174.” Prior to the TCJA, the exemption language in sections 263 and 263A avoided conflict with section 174, allowing R&E expenses to remain fully deductible even when they produced tangible property for the taxpayer. The TCJA did not amend section 263 or 263A, thus leaving the “deduction” language in place. Because R&E expenditures are no longer “deductible,” this raises the question of how to interpret this hold-over language in the legacy capitalization provisions. If Congress meant to exclude all R&E expenditures from section 263 and 263A capitalization, not merely the actual amortization deductions allowed by section 174, why did it not make conforming amendments to these sections?
One interpretation that is consistent with the statutory language is that sections 263 and 263A no longer exempt R&E expenditures at all. However, if capitalization is required under section 174 as well as section 263 or 263A, this raises the question of the appropriate amortization period: Is it five (or 15) years under section 174, or the useful life of the property (for example, a patent or tangible property), under the normal rules for property subject to section 263 or 263A?
An alternative interpretation is that only the amount of the actual amortization deduction allowed to the taxpayer under section 174 is exempt from capitalization under the legacy rules. The rest of the R&E expenses attributable to creating or improving the property would therefore be added to the basis of the property under section 263 or 263A. Under this approach, the basis of the property would decrease each year to reflect that year’s amortization deduction under section 174. But if the property produced by the taxpayer is itself amortizable, then it will still be necessary to determine which amortization period applies. Although this interpretation is consistent with the statutory language, the resulting approach would be unnecessarily cumbersome.
Fortunately, instead of trying to make sense of the mismatch between post-TCJA section 174 and the literal language of sections 263 and 263A, the Treasury has resolved the conflict by focusing on what Congress probably meant to do, rather than what it actually did. According to the post-TCJA regulations interpreting section 263A, the UNICAP rules apply to “preproduction costs, such as costs attributable to research, experimental, engineering, and design activities,” but only “to the extent that such amounts are not research and experimental expenditures as described in section 174.” In contrast to the statutory language, the regulations exempt R&E “expenditures” from the UNICAP rules rather than the amount “allowable as a deduction.” Thus, under the Treasury’s interpretation of section 263A, R&E expenditures are completely excluded from the UNICAP rules. Therefore, their amortization is governed exclusively by section 174.
Although the Treasury has not updated the section 263 regulations since the TCJA’s enactment, the 2023 interim guidance reveals that the Service is taking the same approach that it adopted in the new section 263A regulations. The guidance states that R&E expenditures may not be capitalized under either section 263 or 263A. Thus, the capitalization requirements of section 263 do not apply to any R&E expenditures governed by section 174. Even if those expenditures produce tangible property, their amortization period will be governed exclusively by section 174.
E. Research Tax Credit
Businesses that must now amortize their R&E expenditures can still utilize section 41—the research tax credit—for some degree of tax relief. The research tax credit was first enacted in the Economic Recovery Tax Act of 1981, and while it was initially scheduled to expire at the end of 1985, it was repeatedly extended until it finally became permanent in 2015. Explaining its motivation for extending the credit in 1986, Congress cited its “serious concern about the then substantial relative decline in total U.S. expenditures for research and experimentation.” By repeatedly extending the credit, and finally making it permanent, Congress has manifested a continued desire to encourage R&E activity, in sharp contrast to its decision to eliminate R&E tax deductions in the TCJA.
However, for most taxpayers the tax credit will do little to blunt the impact of losing the R&E deduction.As discussed below, there are two reasons why the deduction has generally been more valuable than the credit. First, the credit is subject to a number of limitations. Second, it is an incremental credit, applicable only if and to the extent that a business increases its qualifying research expenditures above a baseline amount established by its qualifying expenditures in prior tax years.
The research tax credit applies to a much narrower range of R&E expenses than section 174. Although they must meet the definition of “specified research or experimentation expenditures” under section 174, not all such expenses are eligible for the credit. Specifically, the tax credit applies only to a limited subset of research-related expenses: wages, supply costs, fees paid to third parties for computer use, and fees paid to outside contractors. Thus, it is not available for such R&E expenses as rent, utilities, depreciation, attorneys’ fees, and the cost of obtaining patents. In addition, no credit is allowed for research conducted after beginning commercial production of the product/invention, or for research related to adapting an existing product/invention for a particular customer or reverse engineering an existing product/invention. In most cases, only a partial tax credit is available for the costs of research conducted by contractors. In addition, the credit applies only to the cost of domestic research, a significant limitation for companies with foreign research activities. Many software development costs are excluded as well. Thus, many of the expenses that were formerly deductible under section 174 will fail to qualify for the section 41 tax credit.
In contrast to the former R&E deduction, which applied to a taxpayer’s total annual R&E expenses, the research tax credit is an incremental credit. Generally speaking, a business can obtain the credit only if and to the extent that it increases its qualifying research expenditures above a baseline amount established by its qualifying expenditures in prior tax years. Thus, the more a business has spent on research in previous years, the more difficult it will be to claim the section 41 credit in a subsequent year. Therefore, businesses that routinely incur high research expenditures—such as pharmaceutical and software companies, as well as other technology-intensive businesses—may be able to claim little or no credit in any given year.
While the nominal tax credit is 20 percent of qualified expenses, in most cases the maximum available credit is only 10% of those expenses. Even at the newly reduced corporate tax rate of 21%, full deduction under section 174 would be significantly more valuable; generally speaking, that deduction would reduce a business’s tax liability by 21 percent of the much larger base of expenditures that qualify under section 174 compared to section 41.
As noted earlier, with immediate deductions no longer available for R&E expenditures under section 174, some taxpayers may attempt to categorize fewer of their expenses as R&E expenditures, and more of them as “ordinary and necessary” expenses that are deductible under section 162. However, such taxpayers should consider the impact that this allocation might have on their research tax credit. At a minimum, an expense must qualify as an R&E expenditure under section 174 in order to qualify for the tax credit. Therefore, any expenses that the taxpayer deducts under section 162 will be ineligible for the tax credit. Because section 41 takes account of only a subset of the R&E expenditures that qualify under section 174, taxpayers should be especially careful in allocating expenditures that fall into this subset.
F. Foreign Research Activities
The TCJA imposed a longer 15-year amortization period for R&E expenses arising from foreign research activities in order to incentivize taxpayers to spend their research budgets in the United States, which would permit them to amortize those expenses over five years. However, some companies may need to conduct their research in multiple locations, both within the United States and abroad. Under the new rules, such companies will have to find some way to allocate the R&E expenses that benefit both their domestic and foreign activities. For example, these could include shared resources such as certain hardware and software costs, compensation and benefits for employees who work in multiple locations, and the cost of using services that perform data analysis, such as supercomputers.
The TCJA does not provide a method for performing such allocations. The interim guidance from the Service states only that “taxpayers must look to where the SRE [specified research and experimentation] activities . . . are performed to determine whether the corresponding SRE expenditures are attributable to foreign research.” It does not indicate how to determine the “corresponding” expenditures. Unless future regulations provide more guidance, taxpayers will have to develop their own allocation methods and hope that these are accepted by the Service. One method might be to calculate the total non-shared R&E expenses of their domestic and foreign activities, respectively, and divide their shared R&E expenses accordingly. However, this would not necessarily be an accurate reflection of the degree of benefit each activity received from the shared expenses. For example, one activity might use far more supercomputing time than the other activity, even though its other R&E expenses are much lower. Since different taxpayers might adopt different allocation methods, regulatory guidance could promote uniformity or, at the very least, curb abuse.
For some taxpayers, the TCJA’s attempt to encourage domestic over foreign R&E activities may actually incentivize the opposite result: The loss of the R&E deduction may tilt the balance in favor of relocation. Corporate taxpayers may choose to shift a large part of their R&E activities to foreign affiliates located in countries that provide more generous treatment for those expenses and/or lower corporate tax rates.
In a May 2022 letter to Congress, a large group of businesses and trade organizations observed that the U.S. is now one of only two developed nations that require research and development expenditures to be amortized rather than expensed. This could create a significant incentive for U.S. corporations to move their research activities to their foreign corporate affiliates, which can take advantage of more research-friendly tax regimes. As a result, the United States could lose substantial numbers of high-paying research jobs to other countries.
In Ireland, for example, R&E expenses are eligible for both a refundable 25% tax credit and a tax deduction against the relatively low 12.5% corporate tax rate (compared to the 21% rate in the United States), yielding a total effective tax deduction of 37.5%, combined with a reduced corporate tax rate on income from certain copyrighted or patented assets. Where a foreign country provides favorable tax treatment for royalty income, corporate groups may choose to hold any intellectual property produced by their R&E efforts in these foreign entities as well, thus shifting income streams to the low-taxed foreign corporations. It remains to be seen whether the TCJA’s reduction of overall corporate tax rates from 35% to 21% will discourage such emigration.
IV. Potential Impact on Research and Development
Because the TCJA’s changes to section 174 have only recently taken effect, it is too soon to measure their actual impact on the R&E activities of American companies. However, it is possible to make a number of projections. First, taxpayers that engage in R&E activities are likely to face higher tax liabilities due to the loss of the R&E deduction. A company that engages in a fairly consistent level of R&E activity will experience the largest tax increase in the near term, but as amortization deductions accumulate over the years, the shift from deduction to amortization is likely to even out over time. In contrast, companies with highly variable annual R&E spending will experience a continuous mismatch between their actual expenditures and the tax benefits of those expenditures. Companies that increase their R&E spending over time will suffer a greater and more prolonged economic impact from losing the deduction; however, they can offset some of this negative impact by taking advantage of the research tax credit for their increased spending. Companies with significant spending for foreign research activities will be especially hard hit, because those expenses are now subject to 15-year amortization on foreign expenses and do not qualify for the tax credit.
A number of studies have demonstrated a positive correlation between research tax benefits and research productivity. The correlation appears to be especially strong for younger firms, probably because of their financial constraints. Although many different factors may influence a business to increase or decrease its research and experimentation budget, studies that control for these other factors have concluded that tax incentives have a significant impact on companies’ research and experimentation activities, and that the long-term impact is greater than the short-term impact. Based on these studies, the changes in section 174 may lead to decreased research activity and, therefore, decreased technological productivity; they may also have a disproportionate impact on younger firms.
While the reduction in overall corporate tax rates may offset this effect somewhat, its impact may be limited, for two reasons. First, the younger firms that are likely to be disproportionately affected by the changes in section 174 may not have significant income in their early years, making the corporate tax rate reduction relatively unimportant. Therefore, the reduction in tax rates is more likely to benefit larger, established firms. Second, the rate reduction is not tied to research activity. Therefore, a company that has lost its R&E tax deduction could rationally decide to devote more of its spending to activities that will generate high income in the short term, since all of the company’s income will now be taxed at the same low rate. If the company continues to spend money on research activities that have become more costly on an after-tax basis, it may take many years before the products of this research will actually generate income, at which point the tax rates may have increased. It could be rational, therefore, to divert more of this spending to fully deductible activities such as advertising and distribution of existing products and services that will generate immediate income at the current low tax rate. Rather than investing in new and nascent technologies for its future product pipeline, the company will focus on maintaining its income-maximizing and fully deductible status quo. When research and experimentation becomes too expensive, businesses are less likely to undertake these efforts, especially if competitors can copy the end result, which is likely to be the case with innovations that are unpatented and not amenable to protection under the law of trade secrets. The collective impact of these choices may not be felt immediately, but could lead the United States to lose some of its technological competitiveness over the long term. It could also have a significant impact on employment.
One report ranked the United States 26th out of the 36 OECD nations with respect to research and experimentation incentives as of 2018, and predicts that this ranking will decline further under the new amortization regime. Another notes that requiring amortization of R&E expenses is “highly unusual tax treatment from historical and international perspectives.” The other OECD countries permit, at a minimum, full expensing of R&E costs, and in some cases provide “super deductions” amounting to more than 100% of the company’s R&E expenses.
Ironically, under the new amortization regime many businesses will receive more favorable tax treatment for the cost of manufacturing, shipping, advertising and designing packaging for the products created through their research than for the cost of undertaking the research itself. Section 174, both before and after the TCJA, does not apply to costs incurred for activities other than product development and improvement. Advertising expenses and package design expenses, however, are fully deductible when incurred, even when they are likely to provide benefits significantly beyond a single taxable year. And while the non-R&E costs of manufacturing and/or distributing tangible goods are normally subject to capitalization or inventory accounting under the uniform capitalization rules of section 263A, businesses with annual gross receipts below $25 million are exempt from section 263A, an exemption that was significantly broadened by the TCJA.
At the same time that domestic companies were beginning to feel the impact of losing the R&E deduction, Congress also enacted the bipartisan CHIPS and Science Act of 2022, which directs nearly $280 billion in government spending for research activity largely focused on semiconductor chip production. This spending figure greatly outstrips the anticipated $119.2 billion in tax revenues that the repeal of R&E deductions is projected to generate between 2018 and 2027. The government is therefore spending more to encourage the development of specific technologies (albeit important ones) than it will be generating by reducing tax benefits for all technologies. It remains to be seen whether this targeted spending will be worth its relatively high cost, as well as the relative disadvantaging of other technologies.
V. Conclusion
In contrast to the policy explanations that accompanied Congress’s decision to permit deductions as well as flexible amortization of R&E expenses under the 1954 and 1986 Acts, Congress made little attempt to explain its decision to eliminate both options in favor of inflexible amortization rules under the TCJA, and indeed may never have intended these changes to actually take effect. The delayed effective date of the changes, as well as the numerous bills that have sought—thus far unsuccessfully—to repeal or further delay them, suggest that many members of Congress harbor doubts about the wisdom of the new regime. Even as this article goes to press, yet another effort at repeal has commenced. One source suggests, however, that the impending 2024 elections mean that repeal will not happen until 2025 at the earliest.
Critics of the section 174 amendment have argued that it will have significant negative economic implications, a prediction that is consistent with independent economic studies showing a positive correlation between tax incentives and research activity. Companies may shift their spending away from research activities in order take advantage of the full deductions available for most non-R&E expenditures. If companies curtail their R&E activities, or shift those activities to foreign affiliates operating under more research-friendly tax regimes, this will lead to significant losses of high-paying jobs in the technology sector.
While the elimination of R&E deductions is expected to increase federal tax revenues in the short term, the long-term effect is another matter. If the loss of tax incentives causes companies to reduce their R&E activities, this will have a negative impact on the development of new products and technologies that could have given rise to significant tax revenues in the future. Even in the short term, both taxpayers and the Treasury Department will have to devote significant resources to resolving the many ambiguities in the new rules.
Several factors could limit the negative effects of the section 174 amendments: the reduction in overall corporate tax rates, the continued availability of the research tax credit, and the government’s substantial new investments in the domestic semiconductor industry under the CHIPS and Science Act. It remains to be seen, however, whether the cumulative effect of these new developments will counteract the loss of the R&E tax deduction, or whether the TCJA’s reversal of more than 60 years of tax policy will tap the brakes on American innovation.
