Archived information is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.
David N.H. Horler, Ph.D. Submission in Response to Joint Finance Canada - Canada Revenue Agency Consultation Improving the Scientific Research and Experimental Development Tax Incentives:
Submission to the Department of Finance Canada and the Canada Revenue Agency on the SR&ED Tax Incentives
David N.H. Horler, Ph.D.
Horler Information Inc.
9th Floor, 275 Slater Street
Ottawa, Ontario K1P 5H9
This submission primarily addresses question (d) of the Department of Finance Consultation Paper as well as questions (a) and (b) to the extent that they are affected by the Canada Revenue Agency's (CRA's) administration of the Scientific Research and Experimental Development (SR&ED) tax incentive program.
It is important in a public consultation process to know who the respondents are, what they bring to the table and what their interests are. My interests and experience related to SR&ED are summarized in the Appendix. I have had continuous experience with the SR&ED program since its inception in 1985, as a claimant for ten years, as a consultant to CRA for five years, and as a consultant to industry for seven years. I am a scientist by training and I work primarily on the technological side of the SR&ED program. I also have a good working knowledge of the financial rules. My SR&ED experience is mostly in software and Information Technology (IT) claims.
The views expressed in this paper are my own. While I am grateful to clients and professional colleagues for comments received on drafts of this document, the views herein do not express those of any client or organization with which I may be associated.
I have read various recent reports and representations on the SR&ED program, most notably the Canadian Advanced Technology Alliance (CATA) 2007 consultation report, and I have first-hand experience with most of the problems they describe. I see files from coast to coast, and my experience indicates that the problems are widespread, both geographically and in terms of the number of files affected. The problems cannot be dismissed as isolated anomalies. My own experience is consistent with CATA's statistics that 30-50% of large claimants are having problems with the SR&ED program. I do not want to reiterate the problems in detail but I have summarized some key problems below for discussion:
- Inconsistent interpretation of SR&ED;
- CRA ignoring its own policies;
- The lack of any effective mechanism to take a second look at the scientific facts of a case (as opposed to reviewing the process in an administrative sense) at any stage up to and including the Notice of Objection (NOO) stage;
- The removal of checks and balances that had been put in place to address the above problems;
- The lack of corporate memory and therefore repeating the same mistakes as in the past and not having the expertise to resolve problems.
3. Basic Framework
In this section, I want to set out a framework for the subsequent discussion and analysis.
The tension between the views of the SR&ED program as an audit program or an incentive program has always existed and, depending on the dominant ideology within CRA at any given time, is the root cause of the pendulum swings that are observed in the delivery of the SR&ED program every ten years or so. Some of the characteristics that the SR&ED program tends to have under these two ideologies are shown in Figure 1.
Table 1 – SR&ED as an "Audit" versus "Incentive" Program
|Characteristics||"Audit Program"||"Incentive Program"|
|Objective||Tax recovery||Delivering an incentive|
|Interpretation of Income Tax Act (ITA)||Narrow||Broad|
|Tone of audit||More confrontational||Collaborative|
|Audit process||Time-consuming, arduous||Reasonable|
|Risk to CRA||None apparently perceived by CRA||Tax credits perceived to be allowed that should not be|
|Tax policy risk||SR&ED loses its incentive nature||Tax expenditure may become excessive|
|Technological interpretation issues:|
|Concept of SR&ED Project||Small||Small or large|
|Concept of "technology"||Lowest level only||Multi-level|
|Interpretation of Experimental Development (e.g., in software development)||For the purpose of scientific advancement only (e.g., computer science)||For the purpose of technological advancement (e.g., IT)|
|Interpretation of supporting work (ITA subsection 248(1) SR&ED paragraph (d))||Mostly seen as routine work which is ineligible||Eligible|
|Notion of "business" or "commercial" objective||Universally applied||Secondary to SR&ED purpose|
|Documentation requirements||Onerous: special SR&ED documentation may be required||Reasonable|
|Research and Technology Advisor's (RTA) review methodology||Reconstruction of entire claim from historical documentation||Use of T661 project description and interviews to understand the project|
A second framework element for the subsequent discussion is shown in Table 2, which is a compliance matrix of the possible claim decisions made by claimants (the columns headed "Claimant's SR&ED Filing") and a matrix of the possible outcomes of CRA's review of the Investment Tax Credit (ITC) claimed (the two columns at the right). Optionally, Table 2 can be skipped for now because it does not provide any commentary in itself, but the reader may wish to refer back to it during the subsequent analysis.
Table 2 – SR&ED Claim and Outcome Matrix
|Claimant's SR&ED Filing||CRA's Determination on ITC Claimed in Boxes A or C|
|"Truth"||ITC is Claimed||ITC is Not Claimed||Claim is Accepted||Claim is Rejected|
Work and Expenditures are SR&ED
A. Correct determination by claimant.
B. Claimant under-claims or does not claim at all.
Error of omission by taxpayer. (Compliance problem.)
|E. Correct determination by CRA.||
F. Incorrect determination by CRA.
Error of omission by CRA.
|Work and Expenditures are not SR&ED||
C. Claimant over-claims.
Error of commission by claimant. (Compliance problem.)
|D. Correct determination by taxpayer.||
G. Incorrect determination by CRA.
Error of commission by CRA.
|H.Correct determination by CRA.|
In a perfect world, claimants would only claim their "true" entitlements (i.e., only SR&ED claims that fall into box A would be claimed) and CRA would accept all claims (i.e., CRA's determinations would all be in box E). My experience, as well as surveys by CRA and CATA, indicates that the majority of claims are covered by this situation and we should recognize that the system is working in most cases. However, I want to address the issues where:
- CRA rejects valid claims because:
- CRA makes an error in disallowing an expenditure (i.e., the outcome is in box F when it should be in box E);
- CRA moves the goal posts ("Truth"), so that a claim to which the claimant believes it is entitled (top row) is considered by CRA not to be SR&ED (bottom row). CRA views this situation as being where the outcome falls into box H, while the claimant views this as being the same as above (i.e., the outcome falls into box F). An example might be that work that, in truth, is a supporting activity for an SR&ED project is evaluated by CRA in an excessively narrow way as ineligible work because it is "standard practice";
- The claimant does not claim its entitlement (box B);
- The claimant over-claims (box C);
- CRA focuses on minimizing tax expenditures in box G.
4. Analysis and Causes
Arising from Table 2, some problematical situations were identified, which I want to discuss below.
CRA rejects valid claims (Table 2, box F)
Table 2 used the word "truth", which might seem strange in this context. I am using the word to mean an absolute reference. Unfortunately, there is no completely factual or objective way of defining SR&ED. There is a subjective element to the determination of SR&ED under subsection 248(1) of the Income Tax Act (ITA). This is unavoidable but it is possible to define more and more precisely what SR&ED is or is not, which has been achieved with the experience built up during more than 20 years of the SR&ED program.
The application of the SR&ED definition requires consensus and consistency. Consensus requires consultation. If one side or the other (industry or government) tries to unilaterally define the line between SR&ED and "non-SR&ED" (i.e., the line between the top and bottom rows of Table 2), the result is contention. Even worse, if one side or the other unilaterally tries to push the line one way or the other, the result is also contention. At the moment, there is a general sense in the industry that CRA is pushing towards a more restrictive definition of SR&ED with no consensus or consistency.
Table 1 further introduces some concepts regarding the technological interpretation of SR&ED. These are key issues that are causing contention right now. The definition of SR&ED is extremely sensitive to these issues. This means that a small shift in policy will have a large effect on the size of the SR&ED program in terms of tax expenditures and number of claims. The technological interpretation issues listed in Table 1 urgently need to be clarified with the aim of determining administrative policy that works for both CRA and industry.
Let us go back to our perfect world for a moment: suppose that all SR&ED claims submitted were fully compliant (box A). Under an "incentive" philosophy, CRA would review claims, accept them and thank claimants for making CRA's job easier. Unfortunately, the perfect world cannot exist under the "audit" philosophy because auditors have tax recovery targets. Even if a claim is fully compliant, CRA may try to disallow part of it (to put it into box F). The auditor may stay on the case for as long as it takes to find a way of meeting the recovery target. So, the next time around, some taxpayers will have learned to build something into the claim for the auditor to find, and everyone is happy. With respect to the goal of providing claimants their true entitlements, the audit mode of delivering the SR&ED program encourages aggressive claimants and penalizes honest claimants. I do not want to play the audit game in SR&ED.
SR&ED reviewers are increasingly coming in with the attitude that only they understand SR&ED and that the claimant has to be "educated". They tend to talk down to claimants and to create an aura of mystery around the whole process so that only they can make the "true" determination of what qualifies. This attitude usually sets the stage for downward adjustments to claims.
So, what is the true path to SR&ED enlightenment? Clearly, it is to deliver the SR&ED program as an incentive, not as an audit program.
Given that there is no completely objective test to determine if the definition of SR&ED is met, consistency of interpretation is paramount. From the public perspective, the application of the SR&ED provisions of the ITA must be fair from one taxpayer to another and must be consistent over time. This goes to the heart of the fairness and predictability of the tax system. From the CRA administrative perspective, consistency is multifaceted and includes consistency between Coordinating Tax Services Offices (CTSOs), between Research and Technology Advisors (RTAs) and Financial Reviewers (FAs), between claimants, and over time.
The delivery of the SR&ED program depends above all on the technical determinations of work that meets the definition of SR&ED in subsection 248(1) of the ITA. This determination is the responsibility of approximately 200 RTAs assigned to 10 CTSOs. These individuals are the primary points of contact between CRA and claimants at which the program is actually delivered.
The success of the SR&ED program thus depends on the consistent delivery of the program by the RTAs. In the present decentralized delivery model, there are approximately 200 separate delivery points with almost no alignment between them. There appears to be slight coordination of science eligibility decisions within CTSOs and virtually none between CTSOs. Given that CRA is free to organize the delivery of the SR&ED program any way it wishes, the present organization is surprising. An organizational structure that has several hundred delivery points operating in relative isolation from one another is hardly one that comes to mind as a model likely to achieve consistency. There are few checks and balances in place to mitigate the enormous financial power of an individual RTA.
Thus, it is not surprising that we are in fact seeing multiple different SR&ED programs actually being administered in the field. RTAs appear to consider that they have free rein to interpret the criteria as they wish, to create their own written guidelines and even in some cases their own versions of "prescribed" information requirements. The diversity of approaches to reviews and the diversity of interpretations are staggering. The problems appear to be getting worse.
I do not want to give the impression that the RTAs are deliberately trying to undermine the program. The vast majority of RTAs are attempting, in their own minds, to evaluate claims fairly and conscientiously. To their credit, the majority are also, I believe, still trying to deliver an incentive program despite working within an organization that has an audit culture. The problems can arise, in different specific situations, from such factors as lack of experience, lack of training and lack of management. Under the present delivery model, similar problems are bound to exist, and do exist, on the financial review side, but my experience indicates that they are more acute on the scientific side.
We know from experience that the consistency and predictability of SR&ED scientific determinations is very sensitive to the exact interpretation of the SR&ED criteria. As noted in Table 1, questions that can cause a huge variance in the ITCs allowed include:
- the interpretation of what is a "technology";
- what is considered a technological objective versus a business objective;
- the appropriate level of aggregation of an SR&ED project;
- what work is recognized to be subject to technological uncertainty versus what is routine engineering or standard practice;
- what is appropriate to claim as supporting work under paragraph (d) of the definition of SR&ED;
- the documentation required to substantiate a claim.
To my knowledge, CRA does not try to measure the consistency of the SR&ED program. There appears to be no headquarters oversight, no role played by the National Technology Sector Specialists (NTSS), no specific SR&ED quality assurance, no CTSO management involvement, no technical second review process, and even the Notice of Objection (NOO) process reduced only to a procedural review. This means that up to and including the NOO stage, only one RTA will have evaluated the technical facts and there is no mechanism for anyone else to take another look at the technical issues. It is becoming the case that there is no way to have CRA reconsider an initial science opinion other than to appeal the assessment to the Tax Court of Canada.
Thus, CRA seems to be deliberately and systematically removing all mechanisms that aid consistency. In addition to those mentioned above, joint training and consultation with industry are important such mechanisms.
One reason for the present situation that is probably not well understood outside CRA may relate to the dynamics between headquarters and the field. Headquarters "interference" in files can cause internal friction. The concern regarding interference can be well founded if senior management at headquarters or the Minister is being lobbied on a particular file or group of files: direct political pressure in a file clearly jeopardizes the integrity of the tax system. To my understanding, headquarters has completely withdrawn from any involvement in files unless requested. The problem is that headquarters still needs to play an oversight role to ensure the correct and consistent application of policy. Headquarters appears to have given up all the tools available to it to achieve this administrative policy objective.
In a program like SR&ED that requires judgment, consistency over time depends largely on the stability and experience of personnel. The turnover of staff causes some concern but is to some extent inevitable. What is more troubling is when expertise available to the Agency is deliberately pushed out for ideological reasons. This kind of action tends to exacerbate the pendulum swings of the SR&ED program. We are now going through a period similar to one we experienced ten years ago, but almost all the players in CRA are different. We should not have to relive the past just so that each generation of managers can learn anew the mistakes of the past. In my experience, CRA has an unfortunate habit of burying corporate memory rather than using it.
To give one specific example of inconsistency over time, which I believe is linked to the shift towards an audit-driven mindset, CRA has, within the last couple of years, started to take highly aggressive positions to deny SR&ED contract labour. To my knowledge, there is no associated change in the ITA or in published administrative policy. Furthermore, I am not aware of any tax policy objective served by these restrictive positions. Nevertheless, CRA is coming up with new interpretations to attempt to deny expenditures that would not have been challenged a few years ago.
The claimant does not claim its entitlement (Table 2, box B)
By definition, what is not claimed is not measured by CRA. However, this box has great tax policy importance. I am seeing situations, as CATA and others have reported, where claimants are deciding that claiming SR&ED is not worth the hassle. In many cases, CRA's audit process is so contentious, time-consuming and protracted that there is no possible incentive value to the claimant in making a claim. It can now often be the case that the lost opportunity cost of defending an SR&ED claim outweighs the ITC. There is also considerable uncertainty of the outcome because of CRA's inconsistency and the perception that CRA is moving the line on eligibility.
There appears to be somewhat of a dichotomy in this respect between CRA's treatment of large and small companies. Generally, CRA is still operating in "incentive" mode with small companies but frequently in "audit" mode with large corporations. To my knowledge, there is no basis in tax policy for this distinction in how reviews are conducted.
It can be argued that under-claiming of SR&ED (box B) is a compliance problem, just as is over-claiming (box C). The taxpayer that does not claim its SR&ED entitlement is failing to comply with the provisions of the Income Tax Act. Obviously, CRA is not going to put enforcement effort into hounding taxpayers that do not claim SR&ED but, on the other hand, CRA should not drive away claimants by making SR&ED entitlements difficult to access.
The claimant over-claims (Table 2, box C)
We need to acknowledge the problem of aggressive claims. What is happening within CRA is undoubtedly in part a reaction to what CRA sees as the opening up of the program to abuse. As more interpretation papers have been published in recent years, employing more inclusive language, although no doubt with the best of intentions, there must inevitably have been some taxpayers and consultants seeking to push the limits. For example, I am hearing that "experimental production" is an area that needs to be clarified.
Aggressive claims do not equate to large claims. Claims by corporations with multibillion dollar revenues may be large in absolute terms but still small in relation to the corporation's activities and may be prepared in a conservative way. At the other end of the spectrum, small claims can be highly aggressive but can slip through CRA's risk assessment procedures because the number of small claims is vastly greater than the number of large claims. It may be the case that CRA's acceptance of most small claims is actually making its delivery statistics look good, and without too much financial impact to the program, but is hiding real compliance problems.
Industry and SR&ED consultants need to understand what kinds of claims CRA considers to be aggressive, and why. I do not speak for CRA, but it may be that CRA considers that industry had too much control over the program in recent years and that the "partnership" model of program delivery created "loopholes" that need to be closed. That view is probably not entirely without merit. Industry cannot expect to run the program, but that does not mean that all consultation should cease. The program cannot function without consensus. This problem needs to be addressed in further steps that should follow these consultations and is a key issue if we are to achieve a stable long-term model for the delivery of the SR&ED program.
CRA focuses on minimizing tax expenditures (Table 2, box G)
The tax expenditures by the government on the SR&ED program are given by the sum of ITCs in boxes E and G (Table 2). There is a strong perception in industry that CRA is putting a heavy emphasis on eliminating tax loss to the federal treasury by ensuring that there are no claims that fall into box G. However, the way that CRA is doing this is resulting in more valid claims being rejected that should in "truth" fall into box E.
CRA is moving towards increasingly restrictive interpretations of SR&ED, i.e., the "Audit" column of Table 1. Thus, CRA's perception of the claimant's "true" entitlement is shrinking. Also, as reported in the CATA survey, CRA does not always follow its own policies or apply them consistently. CRA's move towards an audit-driven ideology means that CRA tends to lose more and more of the understanding of why and how R&D is performed in industry. As CRA breaks work down into smaller and smaller "nuggets" (a word that is once again in common use within CRA), the overall purpose and challenges cannot be seen and everything starts to look routine. Therefore, by focusing on eliminating "box G" combined with the "audit" view of the program, there is nothing to stop the SR&ED program from being shrunk to the point where it is irrelevant to industry. Unfortunately, there is insufficient senior scientific expertise with CRA to prevent this from happening.
The reason most frequently cited by CRA for the focus on "box G" is potential scrutiny by the Auditor General. However, it is necessary to accept that there is risk in administering a program that has somewhat subjective criteria. The risk on one side is that credits are allowed for ineligible work, and the Auditor General has previously embarrassed CRA (then Revenue Canada) by listing some apparently ridiculous examples of projects that were accepted. However, the risk on the other side is that extreme auditing will cause the program to fail to achieve its tax policy objective as an incentive. It is not clear to me why the Auditor General should focus only on one side of the risk equation, but perhaps because it is easier to see and to measure, and can be portrayed more dramatically.
CRA management is apparently placing emphasis on achieving the delivery targets for processing SR&ED claims, and should be commended for that. However, my experience indicates that there are delivery problems that are not revealed in the published statistics. I suggest that the following questions be examined by the Department of Finance and CRA:
- What is the inventory of claimed, unassessed ITCs?
- What is the breakdown of the inventory by size of claimant (e.g., Canadian Controlled Private Corporations (CCPC) versus non-CCPC)?
- What does the aged inventory look like (ITC versus days in the system)?
- Is CRA using delay status properly and reasonably?
- What proportion of files is assessed without concurrence?
- How many files are continuing on to the Notice of Objection (NOO) stage?
- What is the aged inventory for NOO files?
The answers might reveal a fuller picture of program delivery with respect to timeliness.
Certainty and predictability
As the program is presently being administered, it is ceasing to meet these objectives. When filing the claim, the claimant does not know which published policies CRA considers still to be in effect and which are not, what approach the auditors will take (and they are really auditors these days although they are still called reviewers), and what methods they will use to endeavour to deny work or expenditures. These days, the auditors frequently also maintain an atmosphere of total uncertainty regarding what expenditures will be allowed until the very end of the process, which in a large file can easily be two years or more after the claim is filed.
The net result is that, from the corporation's tax planning perspective, SR&ED often has more the characteristics of a lottery than a stable, predictable incentive. Whatever credits are received tend to be treated as a windfall rather than as something that can be counted on for R&D planning purposes.
Some caveats need to be stated here because CRA's role to ensure fiscal responsibility means that CRA will always need to challenge aggressive filing positions. However, my point is that CRA itself is becoming much more aggressive and restrictive in challenging reasonable claims that in my opinion comply with taxpayers' true entitlements.
We also need to note that services such as the Account Executive service, Preclaim Project Review (PCPR) and Process Review are intended to increase certainty and predictability. Claimants often speak well of these services. However, my concern is that the guidance given to claimants under these services is subject to the overall mindset of delivering either an "audit" or an "incentive" program and that the balance has swung to "audit", requiring the claimant to satisfy a very high standard of proof that work is SR&ED. The claimant runs the risk of under-claiming if CRA attempts to shrink claims so that it is absolutely certain that no ineligible expenditures (box G) are claimed.
The SR&ED program has become overburdened with complex rules and mountains of interpretation. It is now virtually impossible for most taxpayers to obtain their full entitlement to SR&ED ITCs without engaging specialized consulting expertise. A determined effort needs to be put into simplification. This will not be an easy process but, in the end, industry will buy in if the process is open and transparent, and the results reasonable. The concern right now is that changes are being drawn up behind closed doors.
CRA has not yet reached the end state envisaged by the current management team for the SR&ED program. For example, there are rumours of the wholesale cancellation and re-issuance of policies. Another example may be that, having removed any NTSS role in promoting consistency, the next step may be to abolish them.
Unfortunately, draconian and unilateral steps are likely to work against CRA in the medium to long term. For example, the judiciary is less likely to give weight to interpretations issued by CRA without consultation than to interpretations agreed upon with industry. Therefore, CRA may be setting itself up for some unfavorable court decisions, which is a pattern that has occurred in the past. If we are going to rely on the courts to resolve contentious issues, it will take many years and great cost to build a broad body of case law and we may end up with a mish-mash of policy.
This consultation will undoubtedly serve more to identify problems than detailed solutions. There need to be follow-up steps. A Ministers' Conference, jointly hosted by the Ministers of Finance and National Revenue, would be a good start. The eventual resolution of the problems requires that they be elevated to that level.
Although the problems are serious and have important consequences, I do not see that they are unduly difficult to fix. The problems are not new. Solutions have already been devised in the past. If the mechanisms that were previously put in place to fix problems are removed, it is not surprising that the same problems resurface. The only problem that has never been adequately resolved is consistency, but I do not believe that it is intractable if effort is put into addressing it.
The SR&ED program cannot be treated as just another CRA audit program, despite the inclinations of the current CRA management. The program's stakeholders include the Department of Finance, other government departments and Canadian industry. The program is a vital instrument of this country's technological leadership and economic success.
The paramount need is to ensure the stability and smooth operation of the program so that it meets the tax policy and administrative policy objectives of timeliness, cost-effectiveness, simplicity, consistency and predictability. All these objectives are presently at risk.
SR&ED Experience and Interests
David N.H. Horler, Ph.D.
This summary is provided in a spirit of full disclosure of my background and interests relevant to this consultation process.
I am a scientist by training and I work primarily on the scientific and technological side of the SR&ED program. I also have a good understanding of the financial rules. I have experience in SR&ED program administration and experience with a similar US program.
I claimed SR&ED tax credits for software development performed by the R&D team of my own company continuously for ten years (1985-94). We survived three Revenue Canada audits. All our claims were accepted in full or with minimal adjustments. We also had NRC IRAP grants.
I was a consultant to the SR&ED group at Revenue Canada headquarters (now the CRA SR&ED Directorate) for five years (1995-2000) during which time I had several roles:
- I reviewed many Notice of Objection files and other claims referred to headquarters;
I advised the Department of Justice on appeals to the Tax Court of Canada and was an expert witness for the Crown;
I participated in the review of SR&ED in the IT sector and was the principal author of Information Circular (IC) 97-1, which arose out of an 18 month consultation process with industry, CRA and other government departments;
I planned and played a leading role in delivering approximately 30 cross-Canada joint CRA-industry seminars and training sessions on SR&ED in software development just prior to the publication of IC 97-1;
I took part in several Quality Assurance reviews of CTSOs;
I participated in formulating the new initiatives announced by the Minister of National Revenue in 1997 intended to improve the delivery of the SR&ED program, including the Account Executive service, the Pre-claim Project Review (PCPR) service, the National Technology Sector Specialist (NTSS) initiative, and the creation of three new CTSOs for SR&ED.
From the year 2000 to the present, I have been an SR&ED consultant to industry. I work with SR&ED claimants throughout Canada, mostly in the software and IT sector. I deal in some way with up to approximately 50 claims per year from large and small companies.
I have also consulted to the US Internal Revenue Service (IRS) in three capacities: (a) to advise on policy for the US research and experimentation tax credit, (b) to train experts engaged by the IRS to review IT claims and (3) as an expert witness on several large cases.
I am clearly an interested party in this consultation. However, I have no vested interest other than wishing to make the SR&ED program work well for the benefit of claimants. I am not seeking to drive the SR&ED program in any particular direction, other than to have it achieve the tax policy objectives of the Department of Finance and the published administrative policy objectives of the CRA. I have no conflict of interest in any respect regarding my own financial or professional interest. I do not expect to be short of work whatever happens, but I would prefer to spend my time helping claimants rather than fighting CRA. I do not represent any group or organization for the purpose of this consultation.