Extensive international research shows that offering research and development (R&D) tax credits to businesses with an objective of encouraging them to spend more on R&D, a popular policy among governments in the world nowadays for promoting innovation and economic growth, does not encourage them to engage in additional research activities.
It simply makes no sense for a government to adopt a policy that effectively is a subsidy of public money to private businesses to subsidize their operational costs.
The evidence is that businesses may simply use the tax credits to pay for R&D that would have been done anyway and not do what the policy aims to encourage them to do, that is, spend more on R&D than what they are currently spending.
This issue must not be taken lightly because the cost of the policy in terms of forgone tax revenue is difficult to determine in advance and can potentially be very high. In Ireland, for example, the cost of its R&D tax credit scheme doubled from €282m in 2012 to €553m in 2014, reaching 0.3 percent of its GDP. Ireland’s scheme is the most costly among the OCED countries that provide tax relief on private R&D expenditure.
Two economists in an important research article on the topic have commented on the importance of evaluating the policy: “The large amounts of public money it involves, under the form of forgone tax revenue, make it an obvious object of public-policy evaluations. Asking ‘Does the R&D tax credit work?’ is an important and legitimate question … .”
Unintended consequences of such an R&D tax policy could occur when tax credits are granted to all R&D spending of a business. Under this so-called “volume-based” or “level-based” R&D tax credit scheme, suppose company B normally spends $10 on R&D a year and suppose a 10 percent tax credit is introduced, company B stands to receive a $1 tax credit even without spending extra money on R&D.
If company B has no intention of spending extra money on R&D, the $1 tax credit simply becomes a windfall subsidy by the government to subsidize part of the company’s current R&D spending.
Under this scenario, not only does the tax scheme not increase total R&D spending but it may actually lead to a decrease in total private R&D spending since part of the original private R&D spending is now paid by tax money.
Even if company B increases its R&D spending in response to the tax credit scheme, unless the increase is greater than $1, any increased spending less than or equal to one dollar is actually subsidized in full by the $1 tax credit and is not paid by the company’s own money.
Furthermore, it should be noted that in this case, not only is the increased spending wholly subsidized by the $1 tax credit but the remaining value of the tax credit becomes subsidy for the company’s original R&D spending too.
For example, if a company increases its R&D spending by 0.5 dollar, half of the one-dollar tax credit will subsidize this increase in whole while the remaining 0.5 dollar will become subsidy to the company’s original $10 R&D spending.
Thus, under the “volume-based” R&D tax credit scheme, if private firms’ increase in R&D spending is less than the tax credit granted (following the example above, $0.5 new spending<$1 tax credit), even though the total R&D spending might have increased ($10 original spending plus $0.5 increase), there is in fact a net reduction in total private R&D spending ($10 original spending minus $0.5 tax credit).
It is only when the increase in R&D spending by private firms is greater than the tax credit granted that there will be an increase in total private R&D spending.
In principle, an incremental R&D tax credit scheme is a better arrangement for bringing about more private R&D spending because such a scheme grants tax credit to additional R&D spending only. In this case, unless the tax credit is 100%, a one-dollar tax credit can always generate more than a dollar’s worth of additional R&D spending by firms.
In practice, however, there are ways for firms to gain tax credits while avoiding the requirement of spending additional R&D money. One way is to defer R&D expenditure originally earmarked for a particular year.
For example, to claim the tax credit, company B mentioned above can defer $5 that it would otherwise spend in the current year to the following year. By deferring the spending, on the books, the company’s R&D spending in the current year becomes $5 while that of the following year will increase to $15. From the perspective of tax authorities, the latter represents a $10 increase in R&D spending by company B, which is eligible for claiming tax credits. For company B, however, there is actually not any real increase in R&D spending; its total R&D spending over two years, tax credit or not, is the same – $20 in total.
As such, it may turn out that tax credits granted under an incremental scheme in effect similarly subsidize businesses for the R&D that they would otherwise be doing anyway. In the end, it seems the volume-based scheme and the incremental scheme are not dissimilar.
Over the years, a vast volume of research has been done to ascertain the effectiveness of the R&D tax credit policy in boosting R&D spending by private firms. The following is what is learnt from the current state of research regarding the policy’s effectiveness.
– Evidence on whether incremental schemes are more effective than volume-based schemes is mixed; but it tends to suggest that the effectiveness of the two schemes does not differ significantly.
– Studies that are more sophisticated and rigorous indicate that a dollar of foregone tax revenue on R&D tax credits raises private firms’ R&D expenditure by less than a dollar (i.e. tax credits cannot increase total private R&D spending).
– Meta-analyses attempting to verify and sum up findings from various research studies by using statistical methods find evidence of publication bias in the related literature. Publication bias is bias arising from “the fact that authors may have a natural tendency to look for positive results and that it is easier to publish articles showing a significant impact of a policy than those showing no effect.” Because of the bias, a policy’s impact reported in the literature appears stronger than its actual impact in the real world. In other words, the actual effectiveness of the R&D tax credit policy may be weaker than that reported in the literature.
–One meta-analysis finds that after correction for the publication bias, a dollar of tax credit increases R&D spending by 0.03 dollar only.
— Another meta-analysis finds that if uncorrected for the publication bias, the presence of R&D tax incentives is associated with an increase of 58 percent in R&D expenditure. But, after correction for the bias, the corresponding value drops significantly to 7 percent only. The implication of this study is that the actual impact of the R&D tax policy can be a lot weaker than the impact reported in published research studies.
— Comparing private R&D spending in different countries, it is found that in countries with a strong emphasis on tax incentives, private R&D spending has not increased any faster than in countries with limited tax incentives or even no tax breaks at all.
— Qualitative studies on the decision-making processes of private firms provide further support to the quantitative research findings above. They find that R&D credits have little if any effect on decisions to conduct individual pieces of R&D projects.
— One study even points out that most of the firms studied consider R&D tax credits simply as a “bonus”.
The research findings presented above allow us to conclude that the impact of R&D tax credits on the growth of private R&D spending is in fact very limited, and that private firms do not respond to the credits in the ways that policy-makers expect them to.
Naturally, not all policy-makers are the same. There is skepticism. Former US Treasury Secretary, Paul O’ Neill, once said in 2001: “Go talk to people who make practical business decisions about how much those credits influence the level of R&D that they invest in. … You find somebody who says I do more R&D because I get a tax credit for it, you’ll find a fool.”
Likewise, Germany, albeit being one of the most innovative countries, does not have any tax policy aimed at stimulating private R&D.
According to the available evidence above, the amount of additional R&D generated by a dollar of tax credit can be very small and can be significantly less than a dollar. It is likely that a substantial proportion of tax credits granted is in effect unconditional public subsidies to the R&D that private firms would do anyway regardless of the tax credits. Under this scenario, not only is the tax credit policy unable to bring about additional private R&D money, it actually reduces the amount that would have been spent.
Meanwhile, since a dollar of tax credit can only generate less than a dollar of R&D spending at most, R&D tax credits are at best merely unconditional public R&D grants that fully fund private firms’ new R&D projects. Due to the tax credits, firms may be willing to do extra R&D, but they are certainly not going to spend extra own money. The tax credit policy thus appears unable to bring about additional private R&D money even at its best.
The research findings and observations above raise the issue of the rationality of using R&D tax credits as a policy instrument to boost private R&D spending.
If the objective is to fund private firms to do additional R&D, it would be a lot cheaper to achieve this through handing out tax money as research grants directly to private firms than indirectly as tax credits. For a same amount of tax money, direct grants can support more R&D than tax credits can because no tax money will be wasted on R&D that would have been done anyway.
If however the objective is to increase total private R&D spending, an appropriate policy instrument remains to be found.
– Contact us at [email protected]