How Long To Resolution?
“Justice delayed is justice denied.”
William Gladstone, British politician (1809 – 1898)
It is often said that the introduction of revised disputes processes in 1996, their amendment in 2005, and improvements in IRD practices over the past decade have all contributed to the comparatively timely resolution of tax disputes. This article questions whether that is, in fact, the case.
The 1994 Organisational Review of the IRD reported in Appendix E that at that time “the disputes resolution process, including hearings before the Courts and the Taxation Review Authority, can take an unacceptable length of time for both the Commissioner and the taxpayer”.
This observation is amply borne out by consideration, for example, of the dispute which led to the last Privy Council tax decision, Peterson v CIR  3 NZLR 433. The dispute concerned tax deductions from the 1982, 1983 and 1985 tax years and the two cases were not filed in the TRA until 1998, with Privy Council judgment being given in 2005 – 20 to 23 years later.
On the other hand, large disputes which originated around the time of that review were clearly capable of resolution in a much shorter timeframe – see for example CIR v Auckland Harbour Board  3 NZLR 289 PC (13 years), Auckland Gas Co Ltd v CIR  3 NZLR 6 PC (eight to twelve years), CIR v NZ Forest Research Institute Ltd  3 NZLR 1 PC (seven years), CIR v Wattie  1 NZLR 529 PC (seven years) and Golden Bay Cement Co Ltd v CIR  1 NZLR 385 PC (eleven years). A tax advisor in the mid 1990s could realistically suggest to a client that a Privy Council hearing could be had within ten years of the income year in which the dispute arose.
The recommendations from the 1994 review led, among many other outcomes, to the enactment of Part IVA of the Tax Administration Act 1994 (“TAA”) which has among its purposes to establish procedure which will “promote the prompt and efficient resolution of any dispute”.
Part IVA Process
Despite the cynicism with which those mired in the trenches of either side of a Part IVA dispute may regard that stated purpose, it must be acknowledged that the usual result of the process will be that a dispute, if not earlier resolved, will get to Court or the TRA within four years and two months after the end of the tax year in which the relevant tax return was filed. This usual situation may be extended in three situations: where a taxpayer waives the time bar, where the Commissioner invokes the exception to the time bar in section 108 TAA, and where the dispute relates to the imposition of penalties.
Inland Revenue Practices
The IRD statement IR-SPS-INV-170 “Timeliness in resolving tax disputes” published in TIB Volume Fourteen, No 2, February 2002 “establishes administrative practices and timelines that will assist in the timely progression of cases in dispute”. It has to be said that this statement bears little resemblance to reality, or even to IRD’s other publications. For example, the timelines allowed include one month for the Adjudication phase, whereas the IRD article published in TIB Volume Nineteen, No 10 November 2007 explains that the Adjudication Unit’s performance standards require that 80% of adjudication reports be delivered within 8, 14 or 20 weeks of allocation, according to whether the dispute is of a low, medium or high complexity and the average achieved delivery times were 6, 13 and 25 weeks respectively. Especially allowing for the optimism displayed in the statement, its statement that “in a simple case in dispute can take up to 16 months from the time a Notice of Proposed Adjustment (NOPA) is issued until the assessment is issued” gives some flavour of the delays inherent in the process. A large dispute, in practice, will take all of the four year period for resolution, and will require a time bar waiver if the dispute is to go to Adjudication before an assessment is raised.
An apparent anomaly is that there is no effective time period within which the CIR must raise an assessment of a shortfall penalty. Section 94A TAA provides that the CIR may raise penalty assessments and must do so separately from the tax to which the penalty relates. This makes it clear that the penalty assessment is a different assessment from the substantive income tax assessment. The consequence is that the time bar provision, section 108 TAA, only applies once the CIR has made an initial penalty assessment and provides for a further four years within which the CIR can increase that assessment. But there is no time limit within which the CIR must make his first penalty assessment.
A Long Stop?
There may well be a case for a long-stop time bar. Given the CIR’s extensive powers to require information, the seven-year requirement for taxpayers to retain records and the fact that the onus remains on taxpayers, there seems an element of unfairness in the IRD proposing adjustments for, say, tax periods 15 years old, even when this is done in the context of an allegation of sham. Yet this situation is not unknown. The unfairness is exacerbated if the prospect of penalties can hang indefinitely over the hapless taxpayer, like the sword of Damocles.
How long in Court?
Hard cases make bad law, so it is said, but even easy issues like the (in)effectiveness of the JG Russell tax avoidance template have the potential to make tardy law. Indeed the Russell template cases still persist, a quarter of a century or so after the events with which they are concerned. In some cases the blame for this has been attributed by the Courts to the taxpayers and lawyers concerned; in a recent Russell template case Wire Supplies Ltd v Commissioner of Inland Revenue  3 NZLR 458 CA at 465 Priestly J delivering the Court’s judgment said that the situation “calls for proper restraint on the part of the taxpayers and their advisers and acceptance that repetition of failed arguments, sometimes with hair-splitting variations to the arguments as originally made, does nothing to make them more convincing”.
Leaving aside the Dickensian progress of the Russell cases, a case these days can take anything from a year from filing to decision in the TRA, to say six years from filing in the High Court to final decision from the Supreme Court. As an example of the latter, the Supreme Court has very recently heard the appeal from Glenharrow Holdings Ltd v CIR (2007) 23 NZTC 21,564 CA which in turn was an appeal from High Court proceedings first filed in 2002.
A feature of tax litigation in recent years has been the increasing use of the test case and stay procedure in sections 138R and 138Q of the TAA. Chambers J noted in Bage Investments Ltd v CIR (1999) 19 NZTC 15,531 HC that “parking cases inevitably means delay in resolution”. With due respect to His Honour, this presupposes that one party or the other will not accept that the judgment in the test case effectively determines the outcome of the stayed case. If the CIR’s delegate makes appropriate test case and stay decisions and the Courts deliver fully reasoned judgments, then one would expect that the CIR and reasonable taxpayers properly advised would usually accept the futility of attempting to litigate in the face of a test case decision, whichever way that decision went. It is notable that the recent test cases which have resulted in final judgments (Birkdale Service Station Ltd v CIR; CIR v Birkdale Service Station Ltd  1 NZLR 293 CA, Poverty Bay Electric Power Board v CIR (1999) 19 NZTC 15,001 CA and Erris Promotions Ltd v CIR  1 NZLR 811 HC) appear to have been accepted as decisive by those involved in the stayed cases: there have been no reported decisions which appear to be challenges by stayed case taxpayers.
However, it is clearly correct that a taxpayer (or CIR) who feels on reflection that the test case decision is distinguishable from his or her own case, or who obstinately wants his or her day in court, will inevitably face delay from the test case procedure. As an example of the sorts of delay which the process may give rise to, the test cases in the Trinity scheme were determined in late 2003 by a High Court decision which reviewed the CIR’s decision, B v CIR; CIR v Multiple taxpayers  2 NZLR 86 HC. The test cases were heard by the Supreme Court in June of this year; quite possibly a judgment may be delivered five years after the test case determination meaning that the taxpayers involved in the stayed cases, should they (or the CIR, depending on what the Supreme Court decides) still wish to proceed, will have been delayed for an additional five years over and above the delays in the Part IVA process and ordinary High Court litigation.
Summary of potential time
Adding up the various time periods listed or mentioned above, one could get the following projection for typical resolution of a reasonably complex dispute concerning the taxpayer’s 31 March 2008 income year:
|7 July 2008||Taxpayer’s return filed|
|to July 2011||Investigation|
|12 August 2011||IRD NOPA|
|10 October 2011||Taxpayer NOR|
|16 November 2011||Conference|
|to August 2012||Further investigation|
|13 September 2012||IRD SOP|
|10 November 2012||Taxpayer SOP|
|7 January 2013||IRD SOP reply|
|31 March 2013||Time Bar (taxpayer signs waiver)|
|28 August 2013||Adjudication report|
|20 October 2013||Statement of Claim filed in High Court|
|1 June 2016||High Court judgment|
|13 December 2017||Court of Appeal judgment|
|5 May 2019||Supreme Court judgment|
While this time line represents a case which goes the full distance to final appeal, it does not include all conceivable steps – for example multiple successive NOPAs may be issued or a dispute may commence litigation in the TRA adding a further level of Court determination. The delay of 11 years between the end of the tax year to which the dispute relates and final Supreme Court decision seems realistic if the periods in dispute in the only two substantive tax cases to be argued to date in the Supreme Court – the 1997 and 1998 income tax years in the Trinity litigation and GST period ending 31 March 1998 in Glenharrow – are representative of the norm.
This duration is about the same as, or slightly worse than, the average of the Privy Council decisions from the late 1990s to the early part of this century mentioned above. It seems as if no time has been saved.
While under the pre-Part IVA TAA process long timeframes were clearly possible (witness the timeframe in Peterson or the Russell Template cases mentioned above), it is worth recalling that the current timeline can be extended indefinitely if one adds into the equation an assessment outside the time bar (on the basis of alleged fraud or omission of income), a stay of litigation because of a test case or the imposition of penalties after the CIR succeeds in the substantive tax dispute.
Fiscal consequences of delay
Up until the early 1980s the consequences of delay in determining taxpayer liability were very taxpayer-friendly. Under section 398 ITA 1976 as originally enacted and in force until 1 April 1984, a once-only 10% charge was imposed on unpaid tax after one month: no requirement to pay the tax in dispute, no further additional tax and no interest. Especially in a period of double-digit inflation, a canny taxpayer would be ahead by continuing a dispute, win or lose.
Those halcyon days are well and truly over. Interest is payable on the tax in dispute, and the tax in dispute may be collected by the CIR in the course of the dispute if there is considered to be a risk as to the taxpayer’s ultimate ability or willingness to pay the sum.
The time taken to resolve a tax dispute has not reduced appreciably from what appears to be the standard of 10-15 years ago. What appears to have happened is that the relatively costly exercise of preparing NOPAs, NORs and SOPs has replaced what used to be a period of inaction (at least from the taxpayer’s perspective) in disputes.
While there are advantages in having an objective look at the case by the Adjudication Unit, and the CIR no longer has the ability in most cases to delay indefinitely (although, as noted above, there are several exceptions, some of them new), one has to ask whether taxpayers are any better off overall under the current system. The new system costs more and seems to take the same time or more in most cases as the old system. Is it a failed experiment?
 Section 89A(1)(d)
 Sections 16 to 19 TAA
 Section 24 TAA
 Section 149A TAA
 Cicero: Tusculan Disputations V, XXI – XXII
 What is now a trite phrase appears to have been reported first in Winterbottom v Wright (1842) 10 M&W 109 at p 116
 PLM Software Ltd v CIR (2001) 20 NZTC 17,336 HC at 17,342, para 20
 Although it may be questionable whether a decision which had originated in the TRA would easily receive leave for appeal to the Supreme Court
 Section 138I TAA