New Evidence from Canada: Tax Policy Does Affect Research Spending

In a new study published by the National Bureau of Economic Research, three economists study the effect of a recent change in Canada’s research and development (R&D) tax credit on subsequent spending by small companies. The question is especially interesting because small firms may lack sophisticated tax advisors, earn few profits and thus have a lower tax liability against which to deduct tax credits, and have a harder time financing the fixed costs that come with additional research.

In “Do Tax Credits Affect R&D Expenditures for Small Firms? Evidence from Canada,” the authors find that firms that qualified for a larger tax credit did spend more on R&D in the following years compared to firms of similar income whose tax situation did not change. They also find evidence that the refundable nature of the credit made a significant difference.

According to the paper, Canadian tax law allows all countries to deduct 100 percent of research performed in Canada from their taxable income. It also provides all firms with a non-refundable tax credit of 20 percent of qualifying expenditures. However, for small- and medium-size companies (determined by the previous year’s taxable income) the credit is increased to 35 percent up to a certain level of research expenditures. Credits earned at the 35 percent rate are refundable. In 2004 Canada increased the maximum amount of expenditures that could qualify for the credit and also increased the taxable income thresholds for qualifying for the credit. The authors used this reform to measure the change in research spending in following years.

The paper’s main result is that companies that became eligible for additional tax credits increased their research spending by 15 percent. Given the size of the tax reduction, this implied an elasticity of -1.5. In other words, a 1 percent fall in taxes caused companies to increase their research spending by 1.5 percent. Their estimates indicated that small firms may be more responsive to R&D tax incentives, perhaps because they face greater liquidity constraints.

The authors also found indications that liquidity constraints might be dampening the effect. In a perfect market, firms would be able to increase their spending by incremental amounts until the marginal benefits of additional research exactly equaled the marginal cost. In reality, companies might fall short of this level because more research might require them to invest in additional resources such as engineers or capital that would largely be unemployed. The elasticity for spending on contract research, whose fixed costs are generally low, was less than the elasticity for research wages or capital. However, companies that had recently invested in research capital showed a higher response to the improved tax credit than other companies, presumably because the marginal cost of using these facilities was now very low. Finally, the elasticity of the wage bill for contract research companies was much larger than that of other firms. These firms probably have a much easier time ensuring that additional researchers are fully employed.

The paper noted that, unlike in the United States, the relevant portion of the Canadian tax credit is refundable. This allows companies to benefit from it immediately even if they do not owe taxes. Although the United States allows companies to carry unused credits forward for use in future years, the time lag and uncertainty about future profits makes our system less valuable to business.

The main rationale for the R&D tax credit is that companies are usually unable to capture the full benefit of their own R&D spending. This implies that the additional research motivated by a more generous tax credit delivered significant social benefits to the economy. The authors did not attempt to measure these benefits, however.

The United States is currently in the midst of a debate on whether its R&D credit should be made permanent. The fact that one of the authors is currently working as a Senior Economist for the White House Council of Economic Advisors is a promising sign that a broad spectrum of economists and policymakers now recognizes the credit’s worth. Yet over time, as ITIF research has shown, the generosity of the U.S. credit has slipped in comparison to that of other countries, harming our relative advantage as a place to conduct R&D. For instance, the U.S. credit is not refundable, applies only to research spending above a moving average baseline of past spending and is relatively low (14 percent for the Alternative Simplified Credit). In contrast, the Canadian law provides a significant amount of refundable credits for small and medium-sized companies and all expenditures up to a cap qualify for it. Given evidence of its effectiveness and the important role of research in enhancing productivity and living standards, Congress should act soon to make an improved R&D credit a permanent part of a reformed tax code.