A major change to the tax treatment of R&D came into effect on January 1, which surprisingly received little media attention. The change, a provision of the Tax Cuts and Jobs Act of 2017 (TCJA), no longer allows companies to deduct their R&D expenses in the year they are incurred (R&D expenses). Instead, these costs must be amortized over five years (or 15 years in the case of foreign expenditure) (R&D capitalization). Note that there is no corresponding change in generally accepted accounting principles (GAAP), so R&D will continue to be expensed in companies’ financial statements.
Why has this change received so little attention, especially given the focus on the R&D tax credit (which also affects taxes, but not financial reporting)? One of the reasons could be that the change does not matter, that companies will continue to invest in R&D as they have done in recent years. This is plausible, since five years from now, when this year’s R&D has been fully amortized, R&D spending and R&D amortization should follow each other. Often, however, seemingly innocuous changes produce unintended consequences.
I found two “natural tax experiments” that provide insight into how businesses might respond to change. The first is the 1954 American transition in the opposite direction – from capitalization to R&D spending. The second compares the recent experience of US companies with that of UK companies, which are governed by International Financial Reporting Standards (IFRS), where development (the D of R&D) is already capitalized, as long as development is judged viable.
The US transition to R&D spending
Data from S&P Global Compustat allowed me to examine how R&D investment responded to the 1954 tax shift from capitalization to R&D expenditure. Unfortunately, few companies reporting R&D were publicly traded at the time – just 50 in 1950, but tripling to 145 in 1960. R&D investment leading to change rose from 1.6% of revenue in 1950 to 2.6% in 1954 Figure 1). After the change, it continued to rise, reaching 3.7% in 1962.
Since the set of companies changed quite dramatically over the period, I also looked at how R&D intensity changed relative to a company’s baseline each year. . The orange line in Figure 1 represents the average percentage change in firms’ R&D intensity relative to their own benchmark. The figure is consistent with, but less dramatic than the overall change. While the immediate effect in 1955 was a 0.4 percentage point decrease in R&D intensity, firms then resumed the trend of increasing R&D intensity by around 0.12 percentage points per year. . Because this trend started before 1954, it is difficult to attribute the rise to accounting change, but it certainly did not inhibit the trend.
US R&D versus UK R&D in 2000
The second “natural experiment” to test the impact of expenditure versus capitalization of R&D compares the United States to Great Britain (United Kingdom) in 2000. I chose the United Kingdom because it uses IFRS accounting rules and has many public R&D companies. I chose 2000 because after that a number of companies went public before their revenues exceeded their R&D expenditures. Since these are “firms” rather than viable firms, they distort R&D intensity.
Figure 2 reveals that while development capitalization (D) offers benefits to UK companies, they are overwhelmed by other disadvantages. The R&D intensity and R&D productivity of UK companies is on average lower than that of US companies. As a caveat, just as a lot happened in the US between 1951 and 1961 other than accounting change, there are many differences between the US and the UK other than the rules accountants.
A final consideration for the capitalization debate concerns earnings manipulation. A handful of academic studies have shown that pressure to meet quarterly earnings forecasts can cause companies to cut R&D. Indeed, cuts in R&D immediately boost profits, while the consequences on growth are postponed for a year or more. How does capitalization affect this temptation?
While intuition suggests capitalization reduces the temptation to cut R&D under earnings pressure, it actually increases the temptation. Indeed, capitalization reduces the “fiscal cost” of cuts in R&D during the year of investment. Table 1 shows why. Consider a company that planned to invest $100 million in R&D, but due to earnings pressure plans to cut R&D by $20 million. In 2021, this reduction would have increased after-tax profits by $15.8 million ($20 million less $4.2 million in tax). Under the 2022 tax law, the same reduction in R&D would generate after-tax profits of $19.2 million ($20 million less $0.8 million in tax).
In short, there do not appear to be any clear advantages to capitalizing on R&D, and there are indications that there will be disadvantages. American companies (which spend their R&D) already invest more in R&D and are more productive with this R&D than their British counterparts (who capitalize their D). Additionally, investment in R&D (as a share of company revenue) increased after the United States shifted to R&D spending in 1954, suggesting that it may decline by moving in the opposite direction. . Finally, capitalization increases the temptation to reduce R&D to meet profit forecasts. So, if anything, we might expect R&D investments to decline with the return to capitalization.