Economists talk with industry people all the time. We’re obviously talking with different people than you so that particular point won’t get resolved here.
The limitations of the tests that have been done involve the poor quality of the available data, not the use of math. Statistics tests are a useful way of finding out if people do what they say, like spend more on R&D. Qualitative research with retrospective interviews aren’t worth much because it’s too easy for industry to say that this or that project had gone forward because of the credit after the fact.
This is a fascinating discussion. I must admit, this is the first time I have ever read/heard any credible analysis arguing against the R&D tax credit. It's practically gospel in tech that the R&D tax credit is vital. Perhaps we should re-evaluate. But at the same time, I don't discount the fine points made by those on this forum arguing in its favor. If there is one thing we cannot afford in the U.S., it is to fall behind in R&D. We are already, in my opinion, feeling the effects of reduced spending on basic sciences and physics research (Bell Labs, etc.).
So G-Linden agree the credit has been ineffectual?
The credit has been in existence for 31 years. At its enactment in 1981, it was in place for 5 years before expiring. It has been renewed ever since. Companies are fairly confident that it would be renewed because everyone is in favor of it.
So there has been plenty of time for this credit to do its job, but it just hasn't
I agree that the credit does not encourage companies to manufacture overseas, but our system does. And if our system also rewards those that do so with a research credit then this is just icing on the top of the cake we have baked for these multinationals.
We can help bring US manufacturing back by tying the credit to US manufacturing. If a company does not manufacture its product here, it should not get a credit.
In its testimony before the Senate Finance Committee on September 20, 2011, the representative of the OECD stated that any additional research that is done by a company because of a research incentive tends to the kind that produces the lowest returns to the company.
The reason for this is that companies are going to do a certain amount of research in any event and the R&D that they do because of a tax credit is from the bottom of their barrel of research projects.
Another witness said that since we have a limit on the number of researchers in this country, the R&D credit does not increase R&D but increases R&D costs.
If there were a multiplier effect for the economy it wouldn't be a very strong one.
Yes, there are studies. Based on my own interaction with those in the industry I know those studies are completely incorrect, but I am not the only one.
"As currently structured, the U.S. R&E credit probably has had at most a minor and transitory effect on industry R&D spending.’’ Dr. Gregory Tassey, National Institute of Standards and Technology
‘‘In reality, it is impossible for policymakers to know how much research spending taxpayers would have done without the credit.’’ Government Accountability Office,
The R&D credit is project based. By statute, each project must separately qualify and the IRS requires documentation showing that each project satisfies the requirements of the credit.
The fact that industry has not announced even one project that it has done because of the credit is pretty telling,
The large companies do not choose a level of spending with the credit in mind. This an academic theory and academia has been misled here. In reality the R&D credit is not involved in the decision making over the level of R&D spending. Academia cannot refute this because except in one case 25 years ago they have not actually surveyed the decision-makers but have relied on mathematical formulas in determining the effect of the credit.
In several branches of the industry, the real innovation happens in start-ups, who will have very little turnover and thus benefit very marginally from any tax credit. Generally the innovations of start-ups then become the products of the bigger firms. By focussing the use of the $10bn to aid start-ups by some means, whether it be through grants or by a government backed VC/angel funding program, there would be a lot more effective multiplication of the value of the tax revenue used.
Giving tax breaks to large profitable corporations is a very inefficient way to get any new innovation.
The R&D tax credit IS valuable. Don't throw the baby out with the bath water. The firm I founded and own, Tax Point Advisors, Inc., works with CPAs & their clients across the U.S. by helping companies claim this credit. Except for a handful of large companies, the rest - nearly all - are under $100m in sales. And the great majority are $5m - $50m in sales. The credits - often cash refunds - they receive from the R&D tax credit program are an enormous benefit to such small companies. I have seen time and again clients use cash from the credits to hire new technical staff, to buy new equipment, etc. That is not to negate Mr. Rashkin's point that a majority for the $8b - $10b claimed each year in Federal R&D credits goes to major corporations, which, generally, do not need tax "breaks." So, limit the credit for the small to mid-size companies which DO need the help and financial incentives, and which - studies have shown - use this benefit to create new jobs? About 70-75% of the U.S. workers for companies of 100 or fewer employees, and such small business have always been the engine of economic recovery after recessions. So, why not help them, while taking away subsidies for Apple and the like? Simply 1) make the R&D tax credit a permanent part of the U.S. Tax Code, and 2) make it limited to small to mid-size companies (I would suggest up to $100m in annual sales).
Jeffrey Feingold,Founder and Managing Partner
Tax Point Advisors, Inc.
Offices across the U.S., including CA, MA, NY, OH, and TX
www.taxpointadvisors.com; (800) 260-4138
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...