Some observations about the Jenkins Report and SR&ED…

It seems that the Harper government is about to introduce tax-regime changes based on last year’s Jenkins report (see It’s important for governments to question government largesse and the effectiveness of its expenditures. But as stated previously in this blog, SR&ED may not be the most efficient way of stimulating R&D, but it does help create jobs in Canada and it does keep technology companies in Canada.  In fact, I’ve seen it reverse the tide of outsourcing for some companies.. It all reminds me of Winston Churchill’s statement about Democracy:  “Democrasy is the worst form of government, except for all those other forms that have been tried from time to time”.  It ain’t perfect, but without SR&ED it is almost certain that a large chunk of our technology industry would disappear.

The problem I see with SR&ED has largely been the enforcement (or lack thereof) and the complex, oblique and ambiguous rules surrounding the program. It just lends itself to exploitation the way it is being administered at present.

The changes proposed to SR&ED by the Jenkins report don’t seem all that bad: Base it more on labour costs; increase the rates (I like that part); reduce the amount of refundable credits (will surely encourage profitability); and make the pre-review process more of a pre-approval process. This all seems quite reasonable.

They should also create some more “bright lines” in the rules. For instance: If they’re going to make the credit primarily labour-hours-based, they should prescribe that time records be kept, plain and simple. No exceptions. Currently, you are supposed to have good supporting evidence, preferably in the form of time records, but they don’t prescribe the nature of the data to be kept in those time records (and believe me there is a lot of latitude in this area). And if you don’t have good time records, they will often let you hobble together other information to support the time claimed – but that often ends up being an “allocation” based on questionable methodologies. It seems to me that it would be virtually costless to always insist on seeing the time records, and to check that they at least meet some basic characteristics before approving a claim. That would save probably $3,000,000 of the approximately $3,500,000 the government spends on SR&ED each year. And who could reasonably argue against such a rule? The beneficiaries? Hardly. It is other peoples’ money after all.