Printer Friendly

Are U.S. tax incentives for corporate R&D likely to motivate American firms to perform research abroad?

Current U.S. tax policy toward research and development (R&D) is based on the notion that the United States is the dominant competitor in the world for technology-intensive products. (1) While there is some truth to this notion, U.S. technological dominance is eroding at a rapid pace. Table 1 demonstrates that the U.S. share of global sales of technology-intensive products has declined steadily as compared with the major trading partners. Table 2 shows that since the early 1980s the U.S. has spent a constant share of gross domestic product (GDP) on R&D, whereas Japan has increased the proportion of GDP spent on R&D.

At the same time, U.S. firms have increased investments in overseas non-defense R&D at a much faster pace than investments in domestic R&D. Overseas R&D performed by U.S. firms now accounts for well more than 10 percent of total R&D performed by U.S. firms. (2) In addition, data from the Bureau of Economic Analysis (BEA) relating to operations of foreign affiliates of U.S. companies reveal that foreign affiliates of U.S. multinational firms are accounting for an increasing proportion of R&D performed by U.S. firms globally. (3)

More recently, Japan revised its tax-based incentives for R&D investments, and countries such as the United Kingdom, India, and Canada have enacted generous tax and non-tax incentives to attract additional R&D investments. To be sure, R&D investments are expected to follow growth in foreign production, but rapidly increasing R&D costs, declining product life cycles, and inter-country differences in the cost of performing R&D may also influence where firms perform R&D. A key question for U.S. tax policymakers is whether the level of tax incentives for the performance of R&D in the United States is competitive with incentives offered by foreign counterparts.

This article investigates whether the level of tax-based incentives influences where U.S. firms perform R&D investments. It compares annual percentage increases in R&D performed by foreign affiliates of U.S. multinationals in 12 countries for 7 industrial groups over the 1990-2000 period. The data are based on private-sector financed R&D investments (non-governmental sources) by U.S. affiliates as reported by the BEA. The reported results indicate that when compared with the companion industry in the United States, the growth rate in R&D performed by U.S. foreign affiliates was higher in countries with tax-based R&D incentives than in countries without tax-based incentives. Moreover, non-financial foreign affiliates of U.S. firms in Japan, Mexico, and Ireland had annual increases in R&D at rates higher than the remaining countries. When compared with each of the 12 countries on a relative basis, the growth rate in R&D performed by foreign affiliates of U.S. firms in Japan, Mexico, Ireland, and Brazil were also higher than the growth rate of R&D investments by foreign affiliates of U.S. firms in the remaining 8 countries, including investments in the United States.

These results have implications for U.S. tax policy regarding attracting R&D investments of foreign affiliates of U.S. firms to the United States and for keeping existing R&D investments of U.S. interests in the United States. The location of R&D facilities in the United States has important benefits for the American economy. Firms performing R&D hire technical personnel and generate manufacturing jobs that produce employment tax receipts and other favorable economic consequences. Moreover, the location of R&D facilities in the United States is conducive to keeping technology ownership in the United States rather than in foreign countries that may offer more generous levels of R&D-related tax incentives. One way of making the United States a more attractive place for performing R&D is to increase the level of tax-based incentives for R&D in line with other nations that are attracting U.S. R&D investments.

This article first discusses export share of technology-intensive products. It next examines the role of tax-based incentives in the location of R&D investment and discusses the research design. Finally, it discusses the results and advances several tax policy recommendations.

I. Export Share of Technology-Intensive Products

With the explosive growth in global competition and the erosion of U.S. market share of technology-intensive products over the last 2 decades, the stimulation of R&D activity, particularly in the private sector, has become a staunchly embraced policy tool of governments worldwide. Table 1 shows that since 1980 the United States has lost about 6 percentage points of global market share in technology-intensive products. The affected industries are the aerospace, computers and machinery, communications equipment, and pharmaceuticals industries. (4) Closer examination of Table 1 shows that lesser-developed countries such as Singapore and Mexico have increased their market share at the expense of other countries such as the United States, United Kingdom, Germany, and Italy.

Concomitant with the observed decline in U.S. market share of technology-intensive products is evidence that U.S. R&D expenditures have remained relatively flat over the 1980-1998 period, as have such expenditures for the major industrialized nations. An increasing amount of R&D financed by U.S. firms, however, is being performed in foreign countries. One reason for choosing locations other than the United States to perform R&D may be differences in the level of incentives offered for encouraging R&D by individual countries. Governmental subsidies are invaluable in stimulating R&D investments because of the uncertainty in the payoff from R&D and market rivalry.

Research has shown that national governments can help private industry increase their global market share of products under imperfectly competitive market conditions. (5) Alternative governmental actions include (1) imposing tariffs on imports, (2) offering export incentives, (3) forming export cartels, and (4) subsidizing technological innovation. The offering of tax incentives to encourage technological innovation may be the most effective of the alternatives because it is an indirect support and, therefore, would not be in violation of the World Trade Organization (WTO) agreement that prohibits direct tax incentives for exports. (6) Apart from affecting the market share of technology-intensive products, the level of tax incentives offered for R&D is expected to influence where firms locate R&D facilities.

A growing number of countries from both the major industrialized nations and smaller emerging economies are enacting tax-related subsidies to attract R&D facilities of foreign multinationals. Of the 7 major industrialized nations, 4 countries along with the United States have some type of tax incentives and/or other non-tax programs for R&D expenditures. France and the United States have a tax credit based on increases in R&D in excess of historical levels. In contrast, Japan is moving from an incremental credit to a non-incremental credit, whereas Canada has an investment tax credit (ITC) of 20 percent of capital expenditures for R&D. Countries such as Singapore, India, Mexico, and Australia also have enacted generous R&D subsidies to attract foreign R&D facilities. A brief summary of the credit provisions for the United States and several other countries is provided in Appendix 1.

II. The Role of Tax-Based Incentives in the Location of R&D Investments

The site of production facilities alone may not explain where firms locate R&D facilities. Firms are expected to choose among alternative sites to perform R&D based on their derived user-cost of in-house R&D. The user-cost of in-house R&D is the amount of gross revenues that must be earned to pay for the amount invested in R&D and to achieve firms' required rate of return (Cordes, et al., 1987). The user-cost of R&D may be expressed as follows (Equation 1):

(1) [C.sub.t] = [P.sub.r]/[P.sub.Y] * (1 - [tau] - k)(r + [delta]/(1 - [tau]

In Equation 1, [P.sub.r]/[P.sub.Y] is the true purchase price of a unit of R&D, and (1 - [tau] - k)(r+[delta]) is the opportunity cost of adding one unit of R&D input. (7) In Equation 1, [tau] is the corporate tax rate on income, k is the tax incentive for R&D, [delta] is the rate of obsolescence, and r is the discount rate or the weighted cost of debt and equity financing.

Of the elements used to derive the user-cost, k (the tax incentive for R&D) may be the most significant element with respect to where firms locate R&D facilities or where they invest the marginal dollar of R&D investment (United States or foreign locations). The rationale is that the uncertain payoff from R&D may lead firms to choose locations with the highest level of R&D incentives to lower the user-cost of R&D.

R&D incentives lower the user-cost and may make the difference in whether or not firms pursue certain risky projects with low expected values or projects with significant lags in revenue streams from product or process research. The uncertainty inherent in the payoff from R&D arises from 2 sources (Huchzermeier and Loch, 2001) (8) First, R&D investments do not guarantee that profitable innovation will result from product or process-related research projects. Because the payoff from R&D lags other investment alternatives, R&D incentives may increase the net present-value of R&D sufficiently to make the incentives economically plausible. Second, rivalry among firms to be the first to innovate may reduce the payoff from successful outcomes because of price competition among rival innovators.

Table 3 provides an illustration of inter-country differences between special tax incentives for R&D, or k, which influence the user-cost of R&D and possibly the location of R&D investments. For purposes of illustrating country-level differences in tax-based R&D incentives, data for 20 research-intensive firms in the U.S. are used to simulate effective rates of R&D subsidies in 7 countries with some form of tax-based R&D-related incentives over the 1998-2000 period. The rates are based on specific national tax-based incentives for R&D investments over and above the normal business deductions provided for R&D expenditures.

As shown in Table 3, the United States ranks among the lowest with respect to the average effective rate of tax-based incentives for corporate R&D. (9) The companies fared best under the Singaporean, Canadian, Indian, and U.K. programs and worst under the French and U.S. R&D programs. Of the companies examined, Microsoft fared best for all countries examined except for in France and Canada, quite possibly because R&D expenditures grew faster than sales during the examined period. This finding concerning the relation between the location of R&D investments and tax-based R&D incentives serves as an organizing principle in the formulation of the following research question:

H[A.sub.1]: On balance, countries with tax-based R&D incentives had higher increases in R&D investments from U.S. sources than countries not offering such incentives.

The research question addresses whether U.S. companies located their R&D activities to a country based on the availability of R&D tax-based incentives. Geopolitical risks aside, the research question presupposes that firms are generally indifferent about the region of the world in which they perform research. The research question also holds economic factors such as the cost of labor, R&D facilities, patent protection, and other legal, social, and economic factors constant across countries.

III. Design of the Study

In conducting this research, aggregate level data from the Bureau of Economic Analysis (U.S. Department of Commerce, various) were used to compare the annual rates of increase in R&D investments among foreign locations by non-financial affiliates of U.S. multinational firms. The annual rates of increase in R&D investments for each of the industries were compared across the 12 countries using Analysis of Variance (GLM Procedure) during the 1990-2000 period.

The test variable, RDP, was measured based on the annual rate of change in the amount of R&D investment reported by the BEA for each of the 14 countries examined. The data for all countries except the United States were based on the dollar amount of R&D investment from the Operations of U.S. Parent Companies and Their Foreign Affiliates files (BEA, various). (10) The data for the United States were based on annual R&D investments for industrial groups reported by the National Science Foundation (NSF, various).

The class variable, COUN, represents the countries compared. All countries except for Germany, Italy, and the United Kingdom were classified as offering special tax-based R&D incentives during the 1990-2000 period. (11) Brazil was coded as offering tax-based R&D incentives for the industrial machinery and computer equipment industry along with electrical equipment industry during the 1990-2000 period. (12) Mexico was classified as offering tax-based incentives for R&D for 1998 through 2000. In addition, India was categorized as offering special tax-based R&D incentives starting in 1997. The remaining countries with tax-based incentives were labeled as offering tax-based R&D incentives for the entire 1990-2000 period. (13) Appendix 1 discusses the tax-based provision for each country, and Appendix 2 presents the computational formulas of the tax-based incentives for each country.

IV. Results and Tax Policy Recommendations

A. Descriptive Statistics

Table 4 presents the average annual rates of change in R&D investments for each of the countries examined relative to the U.S. R&D (Panel A of Table 4) and to total R&D performed in the 12 countries examined (Panel B of Table 4). The raw data as collected from the aforementioned sources had 1,232 observations (14 countries x 11 years x 8 industries) with various missing data points for a number of the industries in each country during the 11-year period (1990-2000).

Because the variable of interest is the rate of change in R&D investments, industrial groups in each country with more than 4 years of missing data were eliminated, thereby leaving 826 observations (industry years). After eliminating industry groups in each year with more than 4 years of missing data, the miscellaneous manufacturing industry was eliminated. Moreover, Singapore and India had too few data points remaining in too few years, which led to the elimination of the data for both countries. After this step, 532 industry years in 7 industries remained during the 1991-2000 period. To eliminate outlier values, the extreme 5 percent of the data on both the upper and lower tail of the dependent variable (RDP) were eliminated, thereby leaving 458 observations (industry years) for 12 countries.

Panel A of Table 4 illustrates the ratio of foreign R&D investment by foreign affiliates of U.S. companies to R&D investments in the United States for the same industries. Furthermore, the table shows the annual rate of change in the R&D ratios. Panel B shows the same information except that the information is based on the ratio of foreign R&D investment of foreign affiliates to total R&D investment worldwide. The mean values for the annual rates of change are for 7 industries averaged over 10 years. (14)

In Table 5, the 458 industry years consisting of 11 (the Table 5 results exclude the U.S. data) countries and 7 industries over 10 years were used to compare the group means for industry years with tax-based incentives to industry years without tax-based incentives. (15) The statistical comparison was done with the General Linear Model (GLM) Procedure on the SAS Statistical Software Package (SAS Institute Inc., 2002). Panels A and B of Table 5 summarize the Analysis of Variance results. Panel A of Table 5 shows that the difference in the group means differed significantly from zero (p=0.0501). (16) Panel B shows that industry years with tax-based incentives had a mean of 9.61 percent as compared with a mean of 2.24 percent for industry years without tax-based incentives. The difference between the group means is significantly different from zero (p<0.0001).

The implication is that countries with tax-based incentives beyond the regular tax deduction for R&D had higher annual increases in R&D investments from U.S. affiliates than other countries. This result supports the stated research expectation that countries with tax-based incentives for R&D had higher increases in R&D investment from foreign affiliates of U.S. firms than countries without tax-based incentives.

In Panel A of Table 6, the inter-country differences in the annual rates of change in R&D investments relative to U.S. investment also were compared by way of the Analysis of Variance technique using the GLM Procedure on the SAS Statistical Software Package (SAS Institute Inc., 2002). Panel A of Table 6 shows that the inter-country differences in the annual rates of change in R&D as a proportion of U.S. R&D for the companion industry did not differ significantly from zero (p=0.1418). (17) Panel B shows, however, that the change in the R&D investment proportion for Japan, Mexico, and Ireland accounts for much of the variation among the 11 countries examined (p values<0.10). In other words, Japan, Mexico, and Ireland had larger increases in R&D investments from U.S. foreign affiliates than the remaining countries.

The results in Tables 5 and 6 were re-estimated by expressing R&D investments in individual countries as a proportion of total investments in all 12 countries examined (11 foreign countries plus the United States) rather than as a proportion of R&D investments in the United States. Panels A and B of Table 7 summarize the Analysis of Variance results. Panel A of Table 7 shows that the difference in the group means is significantly different from zero (p-0.0436). Panel B of Table 7 shows that industry years with tax-based incentives had a mean of 9.02 percent as compared with a mean of 1.73 percent for industry years without tax- based incentives. (18) The implication is that countries with tax-based incentives beyond the regular tax deduction for R&D had larger annual increases in R&D investments from U.S. affiliates than other countries, a finding in line with the Table 5 results and supportive of the research expectation.

Panel A of Table 8 shows that the inter-country differences in the annual rates of change in the R&D investment as a proportion of U.S. investment differed significantly from zero (p=0.0713). Unlike the case when R&D is expressed as a proportion of U.S. R&D for the companion industry, when R&D is divided by total R&D for all 12 countries in the same industry there are significant differences among the countries with respect to the growth rate in R&D. Panel B of Table 8 shows that the change in R&D investment proportion for Japan, Mexico, Ireland, and Brazil accounts for much of the variation among the 12 countries examined. (19) The results in Tables 7 and 8 confirm the results in Tables 5 and 6 that countries with tax-based incentives for R&D had higher rates of increases in R&D investments from foreign affiliates of U.S. firms.

B. Sensitivity of the Results

The reported results were evaluated for 2 primary concerns regarding the data. First, U.S. R&D data changed from the Standard Industrial Classification System to the North American Industrial Classification System starting in 1998, which may cause measurement errors in the data during the changeover years. To assess the sensitivity of the results to the changeover, the results were re-estimated after eliminating data for 1999 and 2000. (Even though the change occurred in 1998, SIC data were available through 1998.) The results did not change meaningfully.

Second, the rate of change in the R&D ratios may vary significantly from one year to another because of periodic increases or contractions in R&D investment. The problem was addressed in the estimation by eliminating the lower and upper tail of the annual change in the R&D ratios and by taking the leg of the ratios used in the analysis. (20) Results based on the log of the ratios, however, were consistent with those reported herein.

C. Tax Policy Recommendations

Considered together, the reported results show that countries with tax-based R&D incentives had higher annual increases in R&D investments by foreign affiliates of U.S. firms than countries without special incentives. Country-level differences in the annual rates of change of U.S. R&D investment in the countries examined also show that countries such as Japan, Mexico, Ireland, and Brazil had higher annual rates of increases in R&D investment by U.S. affiliates than other countries. These results have implications for U.S. tax policy regarding attracting R&D investments of foreign affiliates of U.S. firms to the United States and for keeping existing R&D investments of U.S. firms in the United States.

R&D is the lifeblood for the growth of an economy, a country, or a firm. R&D contributes greatly to how global market share for goods and services is divided and, therefore, directly affects the relative standard of living among trading nations. (21) Every U.S. administration since the early 1960s has recognized this fact and has sought to encourage R&D spending by tax incentives or by govern merit-sponsored R&D programs. (22) Firms performing R&D hire technical personnel that produce employment tax receipts and other favorable economic consequences.

Four policy recommendations are advanced to make the United States more competitive with respect to attracting additional R&D investments from foreign affiliates of U.S. firms and keeping existing R&D investments of U.S. firms in the United States. The proposals are for U.S. tax policymakers to:

* revise the R&D tax credit to make it permanent and to allow a credit based on total R&D;

* adopt a uniform definition of R&D for purposes of both section 41 and section 174 of the Internal Revenue Code;

* allow the R&D credit independent of whether the firm pays the AMT in the year the credit is earned; and

* repeal the requirement of section 863(b) to allocate R&D expenditures between foreign and U.S. sources of income in the determination of firms' foreign tax credit limitation.

The first proposal would revise the credit provisions under section 41 to base the credit on total R&D outlay rather than on incremental R&D spending. Under the U.S. incremental credit formula, firms may have to keep increasing annual R&D spending to get less and less in credit. Wheeler (2001) points out, however, that firms would generally want to cut costs, including research costs, to improve profits. (23) Moreover, under the current credit formula the credit is increased if sales decline and is decreased or eliminated altogether if a firm's sales increase, other things the same. One solution is to base the credit on total R&D outlay in the current period rather than on incremental spending.

In a sense, all R&D spent in the current year is incremental because it allows the exploitation of existing knowledge. Technological innovation is more often made possible from building on prior knowledge of the firm, industry of membership, or unrelated industry. In many instances, current spending is necessary to take advantage of past breakthroughs outside the firm or industry. Through current spending, significant spillovers can also result in related or unrelated products or processes. Past R&D is a sunk cost and, as such, eligibility for the R&D credit in the current period should not be based solely on past spending.

The second proposal would provide a uniform definition for both creditable and deductible R&D expenditures. A number of the countries use the same definition for determining creditable and deductible R&D expenditures. Since its enactment in 1981, the U.S. credit has been burdened by a fluid definition of creditable expenses. Recipients of the credit expend significant resources to determine whether certain costs qualify for the credit but can have such costs disqualified on a later audit by the Internal Revenue Service.

One way to address the problem of determining creditable R&D expenditures is to replace the U.S. incremental credit with a credit based on total R&D-related labor expenses. Under this proposal, the rate of credit can be calibrated with the amount of expenditures that would qualify for a credit based on total R&D-related labor expenses. The obvious benefit from adopting the labor-cost definition is the minimization of IRS-taxpayer disputes regarding the definition of creditable R&D expenditures.

The third proposal would allow the use of the credit in the year earned independent of whether or not recipients pay the AMT in the year the credit is earned. Under current law, if a firm pays the AMT available R&D tax credits cannot be used in that year. In the perverse extreme, R&D credits can actually reduce AMT credits that would otherwise be carried forward to offset regular taxes in future years when the firm pays the regular tax rather than the alternative minimum tax. The alternative mini mum tax is in substance a pre-payment of a firm's taxes and, therefore, its interaction with the R&D credit is questionable.

Under the fourth proposal, the required allocation of R&D between foreign and U.S. source income would be eliminated. While the sourcing rules may be necessary to reduce tax avoidance in respect of certain stewardship expenses subject to allocation, it appears to be counterproductive for R&D. The mandatory allocation of R&D between foreign and U.S. sources could affect the decision of whether to locate R&D facilities in a U.S. or a foreign location because of the effect on creditable foreign taxes.

IV. Conclusion

The reported results provide evidence that countries with tax-based incentives for R&D had higher annual increases in R&D investments by foreign affiliates of U.S. firms than countries without such incentives. The article suggests 4 policy changes to U.S. tax laws to attract R&D investments of U.S. affiliates overseas and to keep R&D investments of U.S. firms in the United States. One way of making the U.S. more competitive is for the United States to enact a non-incremental credit based on total R&D-related labor expenses. An increasing number of countries are basing tax-based incentives on total R&D expenditures rather than on incremental spending.

A credit based on total labor expenses would simplify the determination of the credit and help businesses factor in the credit with some level of certainty when deciding where to locate research facilities. If the credit were based on total R&D related labor costs, the statutory rate of credit can be calibrated with the amount of expenditures that would qualify such a credit to make it fiscally feasible. One obvious benefit from adopting the labor-cost definition is the minimization of IRS and taxpayer disputes regarding the definition of creditable R&D expenditures. Moreover, to the extent additional jobs are created from the labor-based credit, it would generate employment tax receipts that would partly offset its tax expenditure.

Appendix 1 (27)

Inter-Country differences in R&D Tax Incentives U.S. R&D Tax Incentives

The U.S. Congress enacted special tax provisions in 1981 to stimulate R&D spending by U.S. firms (Pub. Law No. 97-34). Qualified R&D expenditures are eligible for either a basic research credit or an incremental credit. (28) The Tax Reform Act of 1986 reduced the statutory rate of credit from 25 percent to 20 percent and limited the type of expenses that qualify for the credit. (29) Under section 41 of the U.S. Internal Revenue Code, the credit is based on a fixed-based percentage formula. The credit is computed by applying the statutory rate of 20 percent to the excess of the current year's R&D over a computed base period amount.

The base period amount of R&D is computed by taking the ratio of aggregate R&D expenditures to aggregate sales during the statutory base period (1984-1988) and multiplying the result by the average of gross receipts (sales) for the 4 most recent taxable years prior to the taxable year of the credit. (30) The section 174 deduction for R&D expenditures is decreased by 100 percent of the credit taken. The loss of the section 174 deduction and the limitation under section 41(c)(2) with respect to the minimum base amount reduce the tax benefit of the credit well below its statutory rate of 20 percent. For a firm in the 35-percent tax bracket, the maximum tax savings associated with the 20-percent R&D credit is only 6.5 percent of qualifying R&D expenditures. (31) Indeed, the 6.5 percent effective rate of credit is the best-case scenario. Considering the narrow definition of R&D costs that qualify for the credit, and the reduced statutory rate of the incremental credit, the credit may not meaningfully influence R&D intensity.32

An alternative incremental credit is allowed for taxpayers that cannot qualify for the credit under the incremental credit formula. To qualify, the taxpayer must have incurred qualified research expenses in excess of 1 percent of gross receipts for the prior 4 years. Expenditures in excess of 1 percent of gross receipts are eligible for rates of credit between 2.65 to 3.75 percent. (33) Once elected, taxpayers must continue using the alternative incremental credit formula unless the Treasury allows requesting taxpayers to return to the incremental credit formula.

Japanese R&D Tax Incentives

Effective April 1, 2003, Japanese tax incentives for R&D are revised to allow a special tax credit of 8 percent of the total amount of R&D expenses (plus an additional 2 percentage points for R&D investments between 2003 and 2005). In any year, the credit can offset up to 20 percent of total corporate tax liability and any excess is carried forward to the following fiscal year. For companies with R&D-to-sales ratios (average R&D-to-sales ratio over the last 4 years, including the current year) below 10 percent, the rate of credit would be reduced to 8 percent plus 20 percent of the aforementioned ratio (plus an additional 2 percentage points for R&D investments between 2003 and 2005). Where the R&D-to-sales ratio over the 3 prior years exceeds 10 percent, the credit rate will be increased to 12 percent (8 percent plus 2 percentage points plus an additional 2 percentage points for R&D investments between 2003 and 2005). A credit rate of 12 percent will apply to joint R&D, university R&D contracts, R&D contracts with public research institutes and to R&D expenditures for the enhancement of technology of small and medium-size companies (plus an additional 3 percentage points for R&D investments between 2003 and 2005). In addition, taxpayers are allowed to take a special 50 percent depreciation allowance for R&D investment between 2003 to and 2005. The credit is available on R&D conducted overseas.

Canadian R&D Tax Incentives

Canadian tax laws allow an investment tax credit and an immediate tax deduction for both machinery and equipment and total scientific research expenditures incurred during the current year. (34) The rate of credit is 20 percent but the rate is increased to 35 percent for small businesses/private companies. (35) Moreover, Canadian provinces also allow additional incentives to stimulate innovation within provinces in which research is conducted. The credit is also available on contract research.

Tax credits for research expenses can be used to reduce tax liability in the current year and any unused credits can be carried back to the 3 prior tax years or forward over the next 10 tax years. Under Canadian tax laws, the amount of credit claimed reduces the tax deduction for R&D expenditures in the ensuing tax year. The credit is refundable for small businesses/private companies. Unlike the U.S. credit, the Canadian credit applies to a much broader category of research-related expenses and the definition for purposes of the credit is the same as that used for obtaining a tax deduction. Examples of qualifying expenditures include salaries and wages, materials, and certain lease payments related to R&D. The amount of deductible and creditable R&D is reduced by provincial R&D-related benefits.

French R&D Tax Incentives

Under French laws, a research tax credit of 50 percent is granted on qualified R&D expenses. The annual credit is limited to Eu6.1 million per company in each year. Moreover, royalties received on French patents are subject to a reduced tax rate of 20 percent rather than the 35percent effective tax rate on corporate income so long as the income attributable to the patent is not distributed to shareholders. In addition, R&D expenditures are deductible when determining taxable income. (36)

The credit is determined by taking 50 percent of the difference between the amount of the research expenses for the current year and the average of R&D expenses incurred in the 2 previous years. A reduction in research expenses does not entail a repayment of any prior credits, but a portion, equal to 50 percent of the negative variation in R&D expenditures below the 2 year average, reduces R&D tax credit in subsequent years. The credit reduces corporate tax liability and any excess can be carried forward for 3 years or can be discounted with a bank.

The French Treasury refunds any unused credits remaining after expiration of the 3-year carry-forward period to the taxpayer. The definition of creditable expenses is quite expansive and includes costs such as: (1) personnel expenditures (including social security contributions), (2) operating expenses of up to 75 percent of personnel expenses, (3) equipment depreciation, (4) expenditures related to patents, and (5) contract research costs.

United Kingdom R&D Tax Incentives

Under the United Kingdom provisions, large corporations are allowed to deduct up to 125-percent of total research expenses in the current year. The benefits of the enhanced tax deduction for research expenses are limited by taxable profits and are not refundable as is the case with the French credit. The effective rate of the U.K. tax-based incentives is a maximum of 7.5 percent (25% additional R&D deduction x 30% corporate tax rate).

Indian R&D Tax Incentives

India allows a tax super-deduction of 125 percent of certain scientific research expenses and for R&D related capital expenditures in the year such expenses are incurred. (37) Generally, taxpayers may claim a 125 percent super-deduction as long as the prescribed authority of the Indian government (Council of Scientific and Industrial Research, Ministry of Science and Technology) approves the R&D project. (38) In addition, capital expenditures incurred in the 3-year period prior to the commencement of business can be expensed in the year business begins. Prior to the 125-percent deduction for scientific research expenses, India provided a tax holiday for firms performing commercial R&D as a business activity. Tax exemptions on custom duties and excise duties were also common.

Singaporean R&D Tax Incentives

Under Singaporean tax laws, taxpayers conducting a trade or business in manufacturing or in providing certain services may claim a double deduction for R&D expenses. (39) Qualified services are those related to the development of computer software, services related to storage technology, services related to processing or distributing information by the use of computers or services related to computer software. R&D is defined broadly to include studies in the field of science or technology wherein the results of the study will be used for the production or the improvement of products, processes, or materials. Any unused deduction can be carried forward to future years.

Companies involved in R&D also can set aside up to 30 percent of taxable income as an R&D reserve which must be spent on qualified R&D projects within 3 years of setting up the reserve. The 30-percent reserve is offset, however, by the double-deduction for R&D expenses. The R&D related incentives also are supplemented by investment tax credits on machinery, equipment, buildings, patents, and know-how used in R&D projects. Companies are allowed to write-off up to 50 percent of capital expenditures for machinery and equipment, buildings and the cost of know-how and/or patents to the extent the expenditures are for new technology.

In addition to the above, foreign multinationals with substantial operations in Singapore (known as OHQ Companies) are allowed reduced tax rates on Singaporean income and on dividend payments. To qualify, foreign corporations operating in Singapore must be engaged in activities such as technical support services, marketing and sales, treasury functions, or R&D services. Qualified companies performing R&D are allowed additional years of reduced tax rates after the expiration of the initial period of reduced tax rates.

Mexican R&D Tax Incentives

Mexico allows a tax credit of 30 percent on total R&D expenses and investments in research and development of technology. However, the annual tax expenditure for the credit is subject to an annual limit that may reduce the credit to a minimum amount or in the perverse extreme eliminate it altogether. Unused R&D credits can be carried forward for 10 years and the current year's credit is applied against annual income tax liability before monthly estimated tax payments.

Australian Tax-Based R&D Incentives

Australia also provides tax-based incentives for R&D based on a super deduction of 125 percent of qualified R&D expenditures. A premium rate of 175 percent may be used to the extent R&D expenditures in any year exceeds average R&D expenditures over the 3 prior years. R&D expenses on a project performed outside Australia will qualify for the super deduction so long as the project is approved by the Australian Industry R&D Board.

Brazilian Tax-Based R&D Incentives

The Brazilian government has historically offered tax-based incentives in the form of excise tax exemptions for R&D to manufacturers investing in computer and information technology areas. Several restrictions are placed on the eligibility rules. For example, recipients must invest at least 5 percent of their gross sales in qualified R&D activities in Brazil. (40) At a minimum, a portion of such sales must be for cooperative research with universities, research centers or in certain research programs that are considered to be in priority areas of technology. Recently, the tax benefits and the eligibility rules were relaxed for manufacturers of personal computers. (41)

Appendix 2

Computational Formulas for Credit Rates

U.S. [section] 41 (a)(1) Formula

The formula for determining the average rate of credit for the U.S. is presented in Equation 1. (42)

(1) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Where, [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

QRE = Qualified research expenditure is the amount of firms' R&D that qualifies for tax incentives.

GR = Gross receipts, sales

[tau] = Tax Rate

The dollar amount of credit for the United States was calculated through the following steps: (1) determine the fixed base percentage by calculating the QRE to gross receipts ratio over the 1984-1988 period (step-1); (2) compute the mean of gross receipts for the 4 years prior to the year of the credit (step-2); (3) determine the base amount by taking the product of the amounts determined in steps 1 and 2 (step-3); (4) subtract the amount determined in step-3 from R&D expenditures in the year of the credit (step-4); (5) take the product of the amount determined in step-4 and the statutory credit rate of 20 percent (step-5); and (6) multiplying the amount determined in step-5 by 65 percent (1-0.35 tax rate)(step-6). The tax benefit attributable to the regular business deduction for R&D expenditures is ignored. A 35-percent corporate income tax rate is assumed.

U.S. Formula under Proposed H.R. 463, IRC [section] 41(c)(5)

The formula for calculating the average rate of credit under section 41(c)(5) of proposed H.R. 463 would be, as follows: (43)

(2) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Japanese Formula under New Law

The formula for computing the Japanese rate of credit under proposed law is presented below:

(3b) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

if [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

OR

AR[C.sub.t] = [(0.08 + [a.sup.*]0.02 + [b.sup.*]0.02).sup.*]QR[E.sub.t]/QR[E.sub.t]

a = 1 for fiscal years 2003 - 2005 and a = 0 otherwise

b = 1 if [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] and QR[E.sub.j]/N[S.sub.j]>0.10 for j = t-1, t-2, t-3. and b = 0 otherwise

Where, QRE = Qualified research expenditures,

NS = Net Sales

Effective April 1, 2003, Japanese tax incentives for R&D were revised to allow a special tax credit of 8 percent of the total amount of R&D expenses (plus an additional 2 percentage points for R&D investments between 2003 and 2005). In any year, the credit can offset up to 20 percent of total corporate tax liability and any excess is carried forward to the following fiscal year. For companies with R&D-to-sales ratios below 10 percent, the rate of credit would be reduced to 8 percent plus 20 percent of the aforementioned ratio (plus an additional 2 percentage points for R&D investments between 2003 and 2005). Where the R&D-to-sales ratio over the 3 prior years exceeds 10 percent, the credit rate will be increased to 12 percent (8 percent plus 2 percentage points plus an additional 2 percentage points for R&D investments between 2003 and 2005). A credit rate of 12 percent will apply to joint R&D, university R&D contracts, R&D contracts with public research institutes and to R&D expenditures for the enhancement of technology of small and medium-size companies (plus an additional 3 percentage points for R&D investments between 2003 and 2005). In addition, taxpayers are allowed to take a special 50 percent depreciation allowance for R&D investment between 2003 to and 2005. The credit is available on R&D conducted overseas.

Canadian Formula

The formula for calculating the average rate of credit for Canada is presented below.

(4) AR[C.sub.t] = max[0, 0.20(QR[E.sub.t] + C[E.sub.t] - [[tau].sup.*]QR[E.sub.t-1] - [tau]C[E.sub.t-1])]/QR[E.sub.t]

where, CE = capital expenditures

Canadian tax laws allow an investment tax credit and an immediate tax deduction for both capital expenditures and total scientific research expenditures incurred during the current year. The rate of credit is generally 20 percent for large/public companies and 35 percent for private companies. The amount of credit claimed is used to offset the tax deduction for R&D expenditures in the ensuing tax year and it is refundable for small companies only. Provincial R&D related benefits are used to offset federal R&D deductions and investment tax credits in the current year. In computing the average rate of credit, any investment tax credit on machinery and equipment is ignored along with the regular business deduction for R&D expenses. A 28-percent corporate income tax rate is assumed. The credit rates are adjusted for the credit-induced loss of the regular tax deduction based on the credit taken in the prior year.

French Formula

The formula for calculating the average credit rate for France is, as follows:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

CR=R&D credit taken at time j+1.

a = coefficient that assumes values of zero if there was a negative variation at time j and as a one otherwise. A

negative variation occurs when [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Under French laws, a research credit of up to Eu6.1 million can be credited per company in each year. Moreover, royalties received on French patents are subject to a tax rate of 20 percent rather than the 35-percent effective tax rate on corporate income so long as the income attributable to the patent is not distributed to shareholders. In addition, R&D expenditures are deductible when determining taxable income. Unlike the credit programs in a number of other countries, the credit taken does not reduce the tax deduction for R&D. The credit is computed by taking 50 percent of the difference between the amount of the research expenses for the current year and the average of R&D expenses incurred in the 2 previous years. For purposes of computing the average rate of credit for France, the regular business deduction is ignored and the rate is based on the usage of the available credit rather than on the credit earned in each year. Equipment depreciation is also ignored in the calculation. The formula does not reflect any possible reduction in future credits because current spending is less than the base amount. The amount of R&D credit carried forward (out of the Eu6.1 million) is assumed to be refunded by the French Treasury at the expiration of the 3-year carry-forward period.

United Kingdom Formula

The formula for calculating the average rate of credit for the U.K. is presented below.

(6) AR[C.sub.t] = 0.25QR[E.sub.t][tau]/QR[E.sub.t]

Under the United Kingdom provisions, large corporations are allowed to deduct up to 125 percent of total research expenses in the current year. The benefits of the enhanced tax deduction for research expenses are limited by taxable profits and are not refundable as is the case for the French credit. The effective rate of the U.K. credit is a maximum of 7.5 percent (25% additional R&D deduction x 30% corporate tax rate). In computing the U.K. credit, the regular business deduction for R&D is ignored and it is assumed that the credit can be fully used in the year earned. The credit rate for the United Kingdom is based on current tax laws even though 19982000 data are used in the computation. A corporate income tax rate of 30 percent was assumed.

Indian Formula

The formula for calculating the average rate of credit for India is presented below.

(7) AR[C.sub.t] = 0.25QR[E.sub.t][tau]/QR[E.sub.t]

Taxpayers also are allowed a super-deduction equal to 125 percent of in house R&D expenditures that are "approved" by the prescribed authority of the Indian government. (44) The super-deduction is also allowed for firms conducting R&D through universities, institutes, or other qualified government departments or organizations. In computing the rate of credit, the regular business deduction for R&D expenses is ignored and it is assumed that the credit is fully used in the year it is earned. Tax-based incentives on machinery and equipment are ignored.

Singaporean Formula

The formula for calculating the average rate of credit for Singapore is presented below.

(8) AR[C.sub.t] = QR[E.sub.t][tau] + max(0,(0.30NT[I.sub.t] - 2QR[E.sub.t])[tau])/QR[E.sub.t]

Where NTI is net taxable income.

The following is a summary of tax-related R&D incentives offered in Singapore:

(1) Under Singaporean tax laws, taxpayers conducting a trade or business in manufacturing or in providing certain services are allowed a double deduction for R&D expenses.

(2) Companies involved in R&D also can set aside up to 30 percent of taxable income as an R&D reserve which must be spent on qualified R&D projects within 3 years of setting up the reserve. The 30-percent reserve is offset by the double-deduction for R&D expenses.

(3) Companies are allowed to write off up to 50 percent of machinery, equipment, buildings, the cost of know-how and patents to the extent the expenditures are for new technology or new equipment.

In computing the Singaporean rate of credit, the regular business deduction for R&D is ignored along with any investment tax credit on R&D-related machinery and equipment and it is assumed that the credit can be fully used in the year it is earned. In addition, the rate of credit does not include amounts set aside as an R&D reserve. A 24.5-percent corporate income tax rate is assumed in computing the ARC for Singapore.

Mexican Formula

The formula for calculating the average rate of credit for Mexico is presented below.

(9) AR[C.sub.t] = 0.30QR[E.sub.t][tau]/QR[E.sub.t]

The rate of credit under the Mexican program may be somewhat misleading because the tax expenditure for the credit has an annual cap that may reduce the credit to a minimal amount or in the perverse extreme eliminate it altogether. The annual cap for 2002 was $50 million. In computing the average rate of R&D credit for Mexico, it was assumed that the amount of the credit does not offset the regular business deduction for R&D expenses. The rates are not shown in the tables because of the annual cap on the credit.

Australian Formula

The formula for calculating the average rate of credit for Australia is presented below:

(10) AR[C.sub.t] = max[[0.25.sup.*]QR[E.sub.t]/QR[E.sub.t]*[tau], [a.sub.*]*[0.75.sub.*]QR[E.sub.t]/QR[E.sub.t]*[tau]

Australia also provides tax-based incentives for R&D based on a super deduction of 125 percent of qualified R&D expenditures. A premium rate of 175 is allowed to the extent R&D expenditures in any years exceed average R&D expenditures over the 3 prior years. R&D expenses on a project performed abroad will qualify for the super deduction so long as the project is approved by the Australian Industry R&D Board.
Table 1
Export Share of Technology-Intensive Products
(1980-1998) *

Year France Singapore Mexico German Japan

1980 6.94711 1.99871 0.11724 11.07297 12.56059
1981 7.21826 1.60139 0.10385 13.67119 13.88427
1982 9.38512 1.56637 0.07927 14.88049 14.41960
1983 7.15084 1.96837 0.62592 7.15084 16.53761
1984 7.06311 2.18783 0.66795 12.60372 18.74301
1985 6.59547 2.26003 0.64061 12.28559 18.27502
1986 5.73968 2.79424 0.05186 10.83135 17.69768
1987 6.00299 3.64947 0.13273 10.17318 16.64691
1988 5.96502 4.42927 0.19001 9.93412 16.24424
1989 6.18363 4.65998 0.33924 10.20274 15.95776
1990 5.66382 2.94631 0.20951 9.01299 15.75854
1991 6.37221 4.88397 0.24301 9.16928 15.20050
1992 6.07101 5.23876 1.14269 8.21117 14.70772
1993 6.19815 5.97524 1.18089 7.50861 13.64526
1994 5.63970 7.15314 1.53165 7.37702 12.50317
1995 5.50978 7.76518 1.76384 6.45741 12.01317
1996 4.89689 7.76195 1.96701 6.02216 11.37608
1997 5.26113 7.36356 2.17199 6.32067 10.85482
1998 5.35609 6.43020 2.94493 6.49985 9.67646

Year U.K. Canada Italy U.S.

1980 11.23961 2.34974 9.34391 25.65294
1981 10.18472 2.45187 7.85061 23.29033
1982 10.63266 2.54299 7.54228 19.59255
1983 10.11025 2.66741 7.40043 21.06998
1984 9.982971 2.45437 6.55204 18.96947
1985 19.97728 2.5134 6.25236 19.97728
1986 20.73581 2.41403 5.99695 20.73581
1987 9.01386 2.38231 5.80506 20.71789
1988 10.09927 2.16891 4.80585 20.86128
1989 8.94965 2.0623 4.53494 21.20646
1990 8.81769 2.26469 4.23640 22.44114
1991 8.16687 2.41996 3.92109 22.18753
1992 7.28975 2.28541 3.52401 22.09314
1993 7.93535 2.05910 4.24423 19.94097
1994 7.51843 2.17198 3.47523 19.36881
1995 7.72883 2.2772 3.43758 18.72220
1996 7.52035 2.23710 3.49585 19.65529
1997 6.82411 2.10140 3.37223 20.03748
1998 6.43181 2.20412 3.46879 19.77982

Source: World Industry Service, WEFA Group Database (2000).

* The figures are expressed as percentages.

Note: A total of 68 countries are included in the data base.
Technology-intensive industries include aerospace, computers
and machinery, communications equipment, and pharmaceuticals.

Table 2
Total R&D Expenditures as a Percentage of GDP (1981-99)

Country/year United Japan Germany France
 States

1981 2.31 2.13 2.47 1.93
1982 2.48 2.22 2.56 2.02
1983 2.55 2.35 2.56 2.06
1984 2.60 2.43 2.56 2.16
1985 2.72 2.58 2.75 2.22
1986 2.70 2.55 2.77 2.21
1987 2.66 2.62 2.87 2.24
1988 2.62 2.66 2.86 2.24
1989 2.59 2.77 2.86 2.29
1990 2.62 2.85 2.75 2.37
1991 2.69 2.82 2.53 2.37
1992 2.62 2.76 2.41 2.38
1993 2.49 2.68 2.35 2.40
1994 2.40 2.63 2.26 2.34
1995 2.48 2.77 2.26 2.31
1996 2.53 2.80 2.26 2.30
1997 2.55 2.88 2.29 2.22
1998 2.58 3.01 2.31 2.18
1999 2.63 3.01 2.38 2.17

Country/year United Italy Canada
 Kingdom

1981 2.38 0.88 1.24
1982 NA 0.90 1.39
1983 2.20 0.95 1.36
1984 NA 1.01 1.40
1985 2.24 1.12 1.44
1986 2.26 1.13 1.48
1987 2.20 1.19 1.43
1988 2.15 1.22 1.41
1989 2.16 1.24 1.40
1990 2.16 1.29 1.47
1991 2.08 1.23 1.53
1992 2.09 1.18 1.58
1993 2.12 1.13 1.63
1994 2.07 1.05 1.67
1995 1.98 1.00 1.64
1996 1.91 1.01 1.60
1997 1.83 0.99 1.61
1998 1.83 1.02 1.62
1999 1.87 1.04 1.58

Sources: Science and Engineering Indicators (2002); Organisation for
Economic Co-operation and Development (OECD), Main Science and
Technology Indicators database (November 2000).

Table 3
Simulated Rates of Credit for Selected Companies and Countries
(3-year Average, 1998-2000)

Country/ Singapore Canada India U.K.
Company

GM 0.245 0.140 0.088 0.075
Dell 0.245 0.162 0.088 0.075
Hewlett
 Packard 0.245 0.140 0.088 0.075
IBM 0.245 0.144 0.088 0.075
Johnson &
 Johnson 0.245 0.150 0.088 0.075
3M 0.245 0.146 0.088 0.075
Merck 0.245 0.147 0.088 0.075
Microsoft 0.245 0.155 0.088 0.075
Pfizer 0.245 0.157 0.088 0.075
Abbott 0.245 0.145 0.088 0.075
Boeing 0.245 0.142 0.088 0.075
Caterpillar 0.245 0.147 0.088 0.075
Dupont 0.245 0.128 0.088 0.075
Dow
 Chemical 0.245 0.144 0.088 0.075
Ford 0.245 0.145 0.088 0.075
Intel 0.245 0.153 0.088 0.075
Lockheed 0.245 0.145 0.088 0.075
Micron 0.245 0.156 0.088 0.075
Oracle 0.245 0.148 0.088 0.075
Proctor &
 Gamble 0.245 0.151 0.088 0.075
Country
 averages 0.245 0.147 0.088 0.075

Country/ Japan U.S. France
Company

GM 0.063 0.020 0.000
Dell 0.058 0.065 0.034
Hewlett
 Packard 0.066 0.020 0.001
IBM 0.064 0.019 0.001
Johnson &
 Johnson 0.069 0.052 0.003
3M 0.065 0.020 0.006
Merck 0.068 0.025 0.001
Microsoft 0.084 0.065 0.003
Pfizer 0.084 0.065 0.003
Abbott 0.075 0.037 0.002
Boeing 0.061 0.027 0.003
Caterpillar 0.060 0.040 0.011
Dupont 0.066 0.055 0.002
Dow
 Chemical 0.062 0.019 0.003
Ford 0.062 0.030 0.001
Intel 0.070 0.019 0.002
Lockheed 0.060 0.015 0.008
Micron 0.065 0.042 0.025
Oracle 0.075 0.050 0.007
Proctor &
 Gamble 0.061 0.046 0.004
Country
 averages 0.067 0.037 0.004

Table 4
Annual Increases in R&D Proportion by Year and Country

Panel A: Foreign R&D as a Proportion of U.S. R&D

Year/Country 1991 1992 1993 1994 1995

U. K.:
 RDP 0.0232 0.0189 0.0179 0.0282 0.0236
 % [DELTA] in RDP -0.0633 0.0590 -0.0452 0.3213 -0.0168

Germany:
 RDP 0.0352 0.0343 0.0323 0.0339 0.0249
 % [DELTA] in RDP 0.1841 0.0600 -0.0046 0.0795 0.0322

France:
 RDP 0.0087 0.0089 0.0086 0.0130 0.0133
 % [DELTA] in RDP 0.1956 0.0035 -0.0248 -0.0479 0.0947

Italy:
 RDP 0.0081 0.0044 0.0030 0.0051 0.0039
 % [DELTA] in RDP -0.1902 0.0881 0.2007 0.0284 0.0382

Japan:
 RDP 0.0056 0.0080 0.0080 0.0086 0.0114
 % [DELTA] in RDP 0.2270 0.1948 0.2183 0.2347 0.2544

Canada:
 RDP 0.0131 0.0136 0.0121 0.0105 0.0132
 % [DELTA] in RDP 0.0093 0.0170 -0.1177 -0.0812 0.2564

Mexico:
 RDP 0.0027 0.0030 0.0035 0.0074 0.0031
 % [DELTA] in RDP 0.4115 0.1220 0.1622 0.4288 -0.3836

Netherlands:
 RDP 0.0040 0.0053 0.0039 0.0062 0.0059
 % [DELTA] in RDP 0.0187 0.0623 -0.2028 0.3599 -0.0436

Ireland:
 RDP 0.0133 0.0153 0.0333 0.0052 MD
 % [DELTA] in RDP -0.0011 0.4369 0.3576 1.6435

Brazil:
 RDP 0.0021 0.0024 0.0021 0.0014 MD
 % [DELTA] in RDP 0.1656 0.0652 -0.1167 -0.5436

Australia:
 RDP 0.0011 0.0010 0.0042 0.0046 0.0049
 % [DELTA] in RDP -0.0167 -0.0928 0.4528 -0.2329 0.6408

Year/Country 1996 1997 1998 1999 2000

U. K.:
 RDP 0.0244 0.0255 0.0465 0.0531 0.0199
 % [DELTA] in RDP -0.0238 0.0202 0.2890 -0.0491 -0.0386

Germany:
 RDP 0.0241 0.0262 0.0284 0.0292 0.0268
 % [DELTA] in RDP -0.1658 -0.1627 -0.1065 0.1959 -0.0752

France:
 RDP 0.0118 0.0111 0.0142 0.0189 0.0149
 % [DELTA] in RDP -0.0614 0.0653 0.1513 0.0463 -0.0842

Italy:
 RDP 0.0031 0.0057 0.0071 0.0053 0.0058
 % [DELTA] in RDP -0.1260 -0.1740 0.1124 -0.0154 -0.1338

Japan:
 RDP 0.0135 0.0104 0.0088 0.0144 0.0126
 % [DELTA] in RDP -0.0791 -0.0620 -0.0185 0.1748 0.4112

Canada:
 RDP 0.0150 0.0115 0.0178 0.0171 0.0183
 % [DELTA] in RDP 0.1543 -0.0689 -0.0288 -0.0390 0.2039

Mexico:
 RDP 0.0013 0.0016 0.0024 0.0037 0.0037
 % [DELTA] in RDP -0.4137 0.1192 0.7387 0.3923 0.0327

Netherlands:
 RDP 0.0040 0.0044 0.0008 0.0005 0.0064
 % [DELTA] in RDP -0.2041 -0.1692 0.5292 -0.5580 -0.0438

Ireland:
 RDP 0.0030 0.0023 0.0039 0.0046 0.0035
 % [DELTA] in RDP 0.0730 -0.1647 0.2019 0.3369 0.1673

Brazil:
 RDP 0.0027 0.0042 0.0047 0.0046 0.0053
 % [DELTA] in RDP 0.1069 0.1157 0.0818 0.4427 0.0676

Australia:
 RDP 0.0055 0.0040 0.0052 0.0149 0.0074
 % [DELTA] in RDP -0.0810 -0.1583 0.1532 0.3230 -0.1795

Panel B: Foreign R&D as a Proportion of Worldwide R&D

Year/Country 1991 1992 1993 1994 1995

United States
 RDP 0.8827 0.8849 0.8803 0.8661 0.8748
 % [DELTA] in RDP 0.0144 -0.0036 0.0026 -0.0161 0.0042

U.K.:
 RDP 0.0216 0.0175 0.0165 0.0259 0.0213
 % [DELTA] in RDP -0.0607 0.0556 -0.0551 0.3121 -0.1223

Germany:
 RDP 0.0312 0.0330 0.0307 0.0319 0.0241
 % [DELTA] in RDP 0.0634 0.0659 -0.0179 0.0749 -0.0411

France:
 RDP 0.0080 0.0082 0.0079 0.0117 0.0121
 % [DELTA] in RDP 0.1922 0.0119 -0.0373 -0.0571 0.1084

Italy:
 RDP 0.0072 0.0039 0.0027 0.0045 0.0036
 % [DELTA] in RDP -0.1819 0.0915 0.1867 0.0099 0.0540

Japan:
 RDP 0.0052 0.0074 0.0074 0.0078 0.0106
 % [DELTA] in RDP 0.2244 0.1967 0.1995 0.2171 0.2944

Canada:
 RDP 0.0120 0.0125 0.0110 0.0095 0.0119
 % [DELTA] in RDP 0.0114 0.0198 -0.1194 -0.0901 0.2729

Mexico:
 RDP 0.0024 0.0027 0.0031 0.0065 0.0028
 % [DELTA] in RDP 0.4360 0.1375 0.1439 0.4295 -0.3676

Netherlands:
 RDP 0.0036 0.0048 0.0035 0.0055 0.0052
 % [DELTA] in RDP 0.0209 0.0778 -0.2152 0.3570 -0.0432

Ireland:
 RDP 0.0127 0.0147 0.0313 0.0045 0.0019
 % [DELTA] in RDP 0.0157 0.4381 0.3492 1.5994 -0.5800

Brazil:
 RDP 0.0019 0.0022 0.0018 0.0012 MD
 % [DELTA] in RDP 0.1931 0.0721 -0.1181 -0.5452

Australia:
 RDP 0.0010 0.0009 0.0037 0.0041 0.0044
 % [DELTA] in RDP -0.0150 -0.0911 0.4501 -0.2451 0.6601

Year/Country 1996 1997 1998 1999 2000

United States
 RDP 0.8663 0.8583 0.8609 0.8367 0.8881
 % [DELTA] in RDP -0.0097 -0.0210 0.0036 -0.0238 0.0289

U.K.
 RDP 0.0225 0.0233 0.0435 0.0477 0.0189
 % [DELTA] in RDP -0.0226 0.0254 0.3036 -0.0653 -0.0347

Germany:
 RDP 0.0234 0.0248 0.0267 0.0275 0.0253
 % [DELTA] in RDP -0.1646 -0.1658 -0.1100 0.1860 -0.0663

France:
 RDP 0.0106 0.098 0.0126 0.0164 0.0138
 % [DELTA] in RDP -0.0602 0.0457 0.1474 0.0274 -0.0559

Italy:
 RDP 0.0029 0.0051 0.0062 0.0046 0.0053
 % [DELTA] in RDP -0.1255 -0.1775 0.1104 -0.0317 -0.0724

Japan:
 RDP 0.0123 0.0092 0.0083 0.0126 0.0117
 % [DELTA] in RDP -0.0779 -0.0781 0.2050 0.1812 0.4219

Canada:
 RDP 0.0135 0.0103 0.0160 0.0150 0.0169
 % [DELTA] in RDP 0.1471 -0.0716 -0.0344 -0.0626 0.2452

Mexico:
 RDP 0.0011 0.0014 0.0021 0.0030 0.0024
 % [DELTA] in RDP -0.4123 0.1126 0.6573 0.3702 -0.1009

Netherlands:
 RDP 0.0035 0.0038 0.0007 0.0005 0.0059
 % [DELTA] in RDP -0.2057 -0.1741 0.4291 -0.5438 0.0193

Ireland:
 RDP 0.0029 0.0022 0.0035 0.0040 0.0032
 % [DELTA] in RDP 0.0781 -0.1528 0.1553 0.3048 0.1860

Brazil:
 RDP 0.0024 0.0037 0.0040 0.0038 0.0048
 % [DELTA] in RDP 0.1059 0.1086 0.0715 0.4359 0.1384

Australia:
 RDP 0.0048 0.0034 0.0044 0.0121 0.0068
 % [DELTA] in RDP -0.0899 -0.1668 0.1494 0.2695 -0.1311

Legend

MD=missing data

RDP=R&D spending by foreign affiliates of U.S. firms in foreign
countries relative to total R&D spending in the companion industry in
the 12 countries examined.

Table 5
Analysis-of-Variance Test for Effect of Tax-Based R&D Incentives

Panel A: Model Statistics for effect of tax incentive based on
growth rate in R&D relative to the U.S. (24)

Class Degrees of Sum of Mean F-Value P-Value
Variables Freedom Squares Square (Prob. > f)

Model 1 0.5589 0.5589 3.8600 0.0501
Error 426 61.6890 0.1448

Panel B: Least-square means effect of tax incentive based on growth
rate in R&D relative to the U.S.

Class Variables Least Square Means Standard Hypothesis
 For Annual Change Error (Prob. > t)

Country Years with 0.0961 0.0238 <0.0001
 Incentives
Country Years without 0.0224 0.0289 0.4386
 Incentives

Table 6
Analysis-of-Variance Test for Country
Differences Compared with the U.S.

Panel A: Model Statistics for growth rate
in R&D relative to the U.S. (25)

Class Variables Degrees of Sum of Mean F-Value P-value
 Freedom Squares Square (Prob. > f)

Model 10 2.1418 0.2142 1.4900 0.1418
Error 417 60.1062 0.1441

Panel B: Least-square means for differences among countries relative to
the U.S.

Class Variables Least Square Means Standard Hypothesis
 For Annual Change Error (Prob. > t)

Countries
United Kingdom 0.0551 0.0572 0.3362
Germany 0.0054 0.0517 0.9165
France 0.0347 0.0512 0.4980
Italy -0.0153 0.0633 0.8092
Japan 0.1545 0.0507 0.0025
Canada 0.0257 0.0572 0.6536
Mexico 0.1577 0.0792 0.0470
Netherlands -0.0122 0.0705 0.8630
Ireland 0.2507 0.0828 0.0026
Brazil 0.1044 0.0705 0.1396
Australia 0.0798 0.0624 0.2020
United States NA NA NA

Table 7
Analysis-of-Variance Test for Effect of Tax-Based R&D Incentives

Panel A: Model Statistics Effect of Tax Incentives for growth
rate in R&D relative to the U.S. (26)

Class Variables Degrees of Sum of Mean F-Value P-Value
 Freedom Squares Square (Prob. > f)

Model 1 0.5745 0.5745 4.1000 0.0436
Error 456 63.9666 0.1403

Panel B: Least-square means effect of credit based on growth rate in
R&D relative to the U.S.

Class Variables Least Standard Degrees Expected
 Square Error of Under Null
 Means Freedom Hypothesis
 For Annual (Prob. > t)
 Change

Country Years with 0.0902 0.0223 283 <0.0001
Incentives
Country Years without 0.0173 0.0283 175 0.5416
Incentives

Table 8
Analysis-of-Variance Test for Overall Country Differences

Panel A: Model Statistics for growth rate in
R&D relative to other countries (27)

Class Variables Degrees of Type 1 Means F-Value P-Value
 Freedom Sum of Square (Prob. > f)
 Scores

Model 11 2.5928 0.2357 1.7000 0.0713
Error 446 61.9483 0.1389

Panel B: Least-square means for differences among countries as compared
to other countries

Class Variables Least Square Means Standard Hypothesis
 For Annual Change Error (Prob. > t)

Countries
United Kingdom 0.0545 0.0562 0.3320
Germany -0.0148 0.0498 0.7667
France 0.0334 0.0503 0.5063
Italy -0.0118 0.0621 0.8490
Japan 0.1804 0.0494 0.0003
Canada 0.0272 0.0562 0.6288
Mexico 0.1254 0.0761 0.0999
Netherlands -0.0119 0.0692 0.8632
Ireland 0.2449 0.0813 0.0027
Brazil 0.1155 0.0692 0.0957
Australia 0.0816 0.0613 0.1838
United States -0.0035 0.0731 0.9615


(1) The term "R&D "is used to indicate research and development expenses for financial reporting purposes even though only certain research expenses are eligible for the credit. The Internal Revenue Code limits the credit to expenditures that are (1) deductible under section 174 of the Internal Revenue Code, (2) substantially all the development activities constitute a process of experimentation, (3) the development activities are meant to discover information that is technological in nature, and (4) the development activities relate to a new or improved function, performance, reliability, or quality. Therefore, R&D for financial reporting purposes is broader than the definition for purposes of the credit. Technology intensity is assumed present where the cost of R&D exceeds 2.5 percent of value-added cost.

(2) Organisation for Economic and Cooperative Development (1995).

(3) While foreign firms also significantly increased R&D expenditures in the United States, a large proportion of such increase may be attributable to acquisitions of U.S. pharmaceutical and other R&D intensive firms by foreign firms. D.H. Dalton & M.G. Serapio, Jr., U.S. Research Facilities of Foreign Companies, NTIS PB93-134328 (1993).

(4) More often, products are considered to be technology-intensive when more than 2.5 percent of their valued-added cost consists of R&D-related costs.

(5) A. Auquier & R. Caves, "Monopolistic Export Industries, Trade Taxes, and Optimal Competition Policy, 89 Economic Journal 559-81 (September 1979); J. Brander & B. Spencer, "Strategic Commitment With R&D: The Symmetric Case," 14 Bell Journal of Economics 225-35 (Spring 1983).

(6) European Community members took action against the United States in the WTO Resolution Council regarding the legality of tax incentives for exports offered under the Domestic International Sales Corporation (DISC). The U.S. later modified the DISC provisions (and, subsequently, the Foreign Sales Corporation (FSC) provisions) in an effort to bring them in conformity with the WTO agreement, though the WTO ruled that the U.S. Extraterritorial Income (ETI) regime also constituted an illegal export subsidy.

(7) The [P.sub.r]/[P.sub.Y] proportion is essentially the input-to-output ratio of R&D costs.

(8) A. Huchzermeier & C.H. Loch, "Project management under risk: Using the real options approach to evaluate flexibility in R&D," 47 Management Science 85-101 (2001).

(9) The rates are based on the 3-year average over the 1998-2000 period. The U.S. rate reflects the higher of the section 41(a)(1) and 41(c)(4) rates. See Appendix 1 for a discussion of the tax-based provisions. The rates are based on the formulation in Appendix 2.

(10) See Bureau of Economic Analysis, FDIUS: Operations of U.S. Affiliates of Foreign Companies (see http://bea.gov/bea/uguide.htm).

(11) The super-deduction for R&D in the United Kingdom was enacted after the year 2000.

(12) Both industries contained most of the items that qualify for the Brazilian R&D incentives. Some items in the electronic equipment industry, however, do not qualify for the incentives. The remaining industries were coded as not being eligible for tax-based incentives during the 1990-2000 period.

(13) Data for 1990 were lost because of taking the annual change in R&D investments.

(14) As a result, the mean values may be misleading in some situations because some countries may not have R&D investment in all 7 industries from foreign affiliates of U.S. companies. Moreover, one-time increases or contractions in the ratios may skew the observed rates of change. Because the primary interest was determining whether the observed changes were a function of tax-based R&D incentives and the extent to which countries differed from each other, systematic patterns in the mean values among countries were never expected to meaningfully affect the results.

(15) The year 1990 was lost in the computation of the annual rate of change because there were no data for 1989.

(16) For purposes of the study, a p-value of less than 10 percent is used to indicate the cutoff for the acceptable level of statistical significance.

(17) Considering the p-value was 0.1418, the differences among the countries were only marginally non-significant.

(18) For purposes of the study, a p-value of less than 10 percent is used to indicate the cutoff for the acceptable level of statistical significance.

(19) Even though the p-value for Brazil was not statistically significant when U.S. R&D was used to scale annual R&D investment for the companion industry in each country, it was only marginally non-significant.

(20) The descriptive statistics showed that some of the annual rates of change in R&D were negative. Therefore, the number 2 was added to all rates of change to avoid taking the log of negative numbers.

(21) See A. Auquier & and R. Caves, "Monopolistic Export Industries, Trade Taxes, and Optimal Competition Policy. 89 Economic Journal 559-81 (September 1979): J. Brander & B. Spencer, "Strategic Commitment With R&D: The Symmetric Case," 14 Bell Journal of Economics 225-35 (Spring 1983).

(22) See Public Law No. 97-34 (Economic Recovery Tax Act of 1981); R.A. Eisner & M. Sullivan, "The New Incremental Tax Credit for R&D: Incentive or Dis-incentive," 37 National Tax Journal 171-83 (June 1984); Joint Committee on Taxation, Factors Affecting the International Competitiveness of the United States, JCS 6 (May 31, 1991).

(23) R.D. Wheeler, "Is the current U.S. R&D credit globally competitive? Should the scope of the credit be restructured or expanded?" 53 Tax Executive 129-31 (2001).

(24) The Wilcoxon Rank-Sum One-sided test failed to show statistical significance at the 0.10 level between the groups. The test, however, was marginally significant (p=0.1955). The Wilcoxon test is less subject to the distribution of the data than is the Analysis-of-Variance test.

(25) The Kruskal-Wallis Test failed to show significant differences among the countries at the 0.10 level (p=0.3381). The Wilcoxon test is less subject to the distribution of the data than is the Analysis-of-Variance test.

(26) The Wilcoxon Rank-Sum One-sided test also showed statistical significance at the 0.10 level between the groups (p=0.0824). The Wilcoxon test is less subject to the distribution of the data than is the Analysis-of-Variance test.

(27) The Kruskal-Wallis Test failed to show significant differences among the countries at the 0.10 level (p=0.4001). The Wilcoxon test is less subject to the distribution of the data than is the Analysis-of-Variance test.

(27A) See B.A. Billings & R. Paschke, "Would H.R. 463 Improve the Competitiveness of U.S. R&D Tax Incentives," 99 Tax Notes 10 (June 10, 2003).

(28) The basic research credit is available for cash grants to universities or similar organizations. For practical purposes, the incremental credit is the primary R&D tax incentive because it has fewer restrictions on the method of funding used and on the status of the organization performing the research.

(29) The credit is subject to periodic renewal by Congress. It was recently extended to June 30, 2004.

(30) Taxpayers may instead elect to use an alternative incremental research credit formula.

(31) Take, for example, Corporation A, which spent $10 million on qualified R&D expenses in 1990. The base amount for purposes of the incremental credit cannot be less than $5 million (50% x $10 million). The maximum credit is $1 million (20% x $5 million). The otherwise deductible amount under section 174 is reduced by the $1 million credit (100% x $1 million). Assuming no general business credit limitations apply, the federal income tax liability will increase by $350,000 (0.35 x $1 million) because of the credit induced loss of the section 174 deduction. Thus, the effective rate of credit (net tax benefit/R&D expenditures) is 6.5 percent [($650,000/$10 million) x 100].

(32) Under section 41, R&D in the experimental or laboratory sense qualifies for a 20-percent tax credit. Qualified expenses must be for in-house research, contract research, ar basic research. In addition, the project must be for discovering information that is technological in nature, the information discovered must be potentially useful in a new or improved business component, and substantially all the activities of the research must be elements of experimentation for a functional process.

(33) Under the alternative incremental research credit election, a credit rate of 2.65 percent is applied to qualified research expenses exceeding a base amount computed using a fixed-based percentage of 1 percent of average gross receipts for the 4 prior years but not to exceed a fixed-base percentage of 1.5 percent of average gross receipts for the 4 prior years. A credit rate of 3.2 percent applies to qualified research expenses exceeding a base amount using a fixed-base percentage of 1.5 percent of average gross receipts for the 4 prior years but not to exceed 2 percent of the 4 prior years' gross receipts. A rate of 3.75 percent is applied to qualified expenses in excess of the 2 percent base amount. Assuming a marginal tax rate of 35 percent the maximum credit rate under the AIRC is 2.44 percent.

(34) Generally, the credit is available on machinery and equipment only, not on land and building, incurred during the taxable year if such machinery and equipment are used in Canada. However, the credit is available on special purpose buildings. The credit ranges from 30 percent, in the Atlantic and Gaspe provinces, to 20 percent in the rest of Canada.

(35) The 35-percent rate for small businesses/private companies is reduced under certain conditions.

(36) Taxpayers can fully deduct research expenses or can elect to capitalize and depreciate such expenses if the research has a reasonable chance of profitability during the period.

(37) The rate is increased to 150 percent for research related to biotechnology.

(38) See section 35 of the indian Income Tax Act of 1961.

(39) See also M. Sulivan, "Low-Tax Gets Quatity, Quality from U.S. Multinationals," 28 Tax Notes International 787 (November 25, 2002).

(40) Export sales are not included in determining the 5-percent threshold. Moreover, the exemption varies among regions of the country.

(41) See Soares da Silva & David Roberto, "PC Manufacturers win Tax Breaks," 29 Tax Notes International 1151 (February 10, 2003).

(42) By election, taxpayers may chose to use an alternative incremental research credit formula, which allows firms to compute the credit as a sum of (i) 2.65 percent of the excess of qualified research expenses for the taxable income over a fixed base percentage of 1 percent of the average of gross receipts (sales) for the 4 most recent taxable years preceding the taxable year (but not exceed 1.5 percent of such average); (ii) 3.2 percent of so much of such expenses as exceeds 1.5 percent of such average (but not exceed 2 percent of such average); and (iii) 3.75 percent of so much of such expenses as exceeds 2 percent of such average.

(43) The rate would be 6 percent of total qualified R&D for start-up taxpayers. A 35-percent corporate income tax rate is assumed.

(44) The rate is increased from 125 percent to 150 percent for R&D related to biotechnology.

B. ANTHONY BILLINGS is a professor of accounting at Wayne State University in Detroit, Michigan. He has published widely on issues related to domestic and international tax policy and international competitiveness, including a paper entitled "The Role of Taxes in Corporate R&D Spending," which appeared in the October 2001 issue of R&D Management (Blackwell Publishers). Another article on state-level research tax credits appeared in the June 2002 issue of CPA Journal.
COPYRIGHT 2003 Tax Executives Institute, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:research and development
Author:Billings, B. Anthony
Publication:Tax Executive
Date:Jul 1, 2003
Words:13255
Previous Article:TEI's 2003-2004 goals and objectives.
Next Article:Elimination of the federal capital tax: June 19, 2003.
Topics:


Related Articles
U.S. Taxation of International Income: Blueprint for Reform.
Will Clinton give industry a green edge?
Research credit redux: Federal Circuit brings the research credit back to life - Fairchild v. United States.
State economic development initiatives.
Mayor proposes tax incentives. (Sault Ste. Marie).
Problems with current U.S. policy.
The "Geek Gap": the erosion of America's manufacturing base is undermining our research and development capacity--and, unless reversed, will lead to...
Tax reform: challenges and opportunities.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters