Canadian Tax Journal Revue Fiscale Canadienne

2020 ■ VOLUME 68, No 4

CANADIAN TAX JOURNAL REVUE FISCALE CANADIENNE

PEER-REVIEWED ARTICLES Interest Relief on Income Tax Debts: Canada Versus the United States Michael H. Lubetsky Tax Literacy: A Canadian Perspective Anthony Pham, Antoine Genest-Grégoire, Luc Godbout, and Jean-Herman Guay

POLICY FORUM Crisis, Cleanup, and the Prospect of Long-Term Fiscal Change Shirley Tillotson Editors’ Introduction—The GST/HST Responsibilities of Non-Resident E-Commerce Firms Alan Macnaughton and Daniel Sandler The GST/HST Obligations of Non-Resident E-Commerce Firms— Jurisprudence and Policy Nicholas Shatalow Carrying On About Carrying On Business: A Response to “The GST/HST Obligations of Non-Resident E-Commerce Firms” Zvi Halpern-Shavim Much Ado About Doing Not Much: Some Reflections on the Jurisdiction To Tax Business Transactions Malcolm Gammie

FEATURES Finances of the Nation: Provincial Debt Sustainability in Canada: Demographics, Federal Transfers, and COVID-19 Trevor Tombe Current Cases: (FCA) Iberville Developments Limited v. Canada; (FCA) Landbouwbedrijf Backx BV v. Canada; (BCCA) Collins Family Trust v. Canada (Attorney General) Kirsten Kjellander, Ryan L. Morris, John Sorensen, Ehsan Wahidie, and Anita Yuk Selected US Tax Developments: New Proposed Regulations Under the Section 1061 Carried Interest Rules Peter A. Glicklich and Gregg M. Benson Current Tax Reading Robin Boadway and Kim Brooks ■ CANADIAN TAX JOURNAL EDITORIAL BOARD/ COMITÉ DE RÉDACTION DE LA REVUE FISCALE CANADIENNE

■ Editors/Rédacteurs en chef Alan Macnaughton Daniel Sandler University of Waterloo EY Law llp Kevin Milligan (on leave) Frances Woolley University of British Columbia Carleton University

■ Practitioners and Academics/Fiscalistes et Universitaires Richard M. Bird University of Toronto Allison Christians McGill University David G. Duff University of British Columbia Luc Godbout Université de Sherbrooke Kenneth J. Klassen University of Waterloo Wojciech Kopczuk Columbia University Rick Krever The University of Western Australia Angelo Nikolakakis EY Law LLP, Montreal Shawn D. Porter Deloitte, Toronto Diane Ring Boston College Lindsay Tedds University of Calgary www.ctf.ca/www.fcf-ctf.ca Call for Book Proposals

The Canadian Tax Foundation, an independent, not-for-profit research and educational organization, is seeking proposals for books in the areas of taxation and public finance. Since its inception in 1945, the Foundation has published many books and articles on a wide range of subjects within its areas of interest. The Foundation seeks proposals for research projects that will

n result in a book on a single topic of interest in the area of taxation or public finance; n be undertaken by an experienced researcher who has expertise in an area of taxa - tion or public finance; and n be carried out within a time frame that is reasonable, given the nature of the project.

Projects selected by the Foundation may qualify for its full or partial financial support of the research and for its underwriting of the publication costs. The Foundation retains the absolute right at its sole discretion to choose whether to support a given proposal or to publish a project. Interested parties should send a brief written outline of a proposal, for initial consider - ation by the Foundation, to:

Heather Evans Executive Director and Chief Executive Officer Canadian Tax Foundation/Fondation canadienne de fiscalité 145 Wellington Street West, Suite 1400 Toronto, Ontario M5J 1H8 [email protected]

For further information, please contact the director, as indicated above, or the co-chairs of the Canadian Tax Foundation Research Committee:

Hugh Woolley c/o Canadian Tax Foundation/Fondation canadienne de fiscalité

Kim Brooks c/o Canadian Tax Foundation/Fondation canadienne de fiscalité

■ i ■ Appel de propositions de livres

La Fondation canadienne de fiscalité (FCF ) / Canadian Tax Foundation, un organisme sans but lucratif indépendant de recherche et à caractère éducatif, souhaite recevoir des propo - sitions de livres dans les domaines de la fiscalité et des finances publiques. Depuis sa fondation en 1945, la FCF a publié de nombreux livres et articles sur divers sujets dans ses champs d’intérêt. La FCF souhaite obtenir des propositions de projets de recherche qui :

n mèneront à la rédaction d’un livresur un sujet unique d’intérêt en fiscalité ou en finances publiques; n seront dirigés par un chercheur chevronné ayant une expertise dans un domaine de la fiscalité ou des finances publiques; n seront effectués dans un délai raisonnable, compte tenu de la nature du projet.

Les projets qui seront sélectionnés par la FCF pourront être partiellement ou totalement admissibles à une aide financière pour la recherche et les frais de publication. La FCF se réserve le droit absolu, et à sa seule discrétion, d’appuyer une proposition particulière ou de publier un projet. Toute personne intéressée doit faire parvenir un bref sommaire de la proposition pour examen initial par la FCF à :

Heather Evans Directrice exécutive et chef de la direction Canadian Tax Foundation/Fondation canadienne de fiscalité 145 Wellington Street West, Suite 1400 Toronto, Ontario M5J 1H8 [email protected]

Pour plus d’information, veuillez communiquer avec le directeur, tel qu’il est mentionné plus haut, ou avec les co-présidentes du comité de recherche de la Fondation canadienne de fiscalité :

Hugh Woolley a/s Canadian Tax Foundation/Fondation canadienne de fiscalité

Kim Brooks a/s Canadian Tax Foundation/Fondation canadienne de fiscalité

■ ii ■ ■ Recent and Upcoming Events * ■ Activités récentes et à venir *

YOUNG PRACTITIONER WEBINAR SERIES Various Topics: January 13 and 28, February 17 and 24, 2021

TRANSFER PRICING CONFERENCE AND LIVE WEBCAST February 3-4, 2021

ESTATES & TRUST SEMINAR AND LIVE WEBCAST April 14-15, 2021

JOURNÉE D’ÉTUDES FISCALES ET WEBÉMISSION le 6-7 juillet 2021

THE BELOW DATES ARE TENTATIVE: PRAIRIE PROVINCES TAX CONFERENCE AND LIVE WEBCAST May 31-June 1, 2021, Winnipeg

BRITISH COLUMBIA TAX CONFERENCE AND LIVE WEBCAST September 20-21, 2021, Vancouver

ONTARIO TAX CONFERENCE AND LIVE WEBCAST October 25-26, 2021, Toronto

ATLANTIC PROVINCES TAX CONFERENCE November 5-6, 2021, Halifax

73RD ANNUAL TAX CONFERENCE November 28-30, 2021, Toronto

* For further details on upcoming events, please visit the Canadian Tax Foundation website at www.ctf.ca. / Pour plus de renseignements, veuillez consulter le site Web de la Fondation à www.fcf-ctf.ca.

Disclaimer. The material contained in this publication is not intended to be advice on any particular matter. No subscriber or other reader should act on the basis of any matter contained in this publication without considering appropriate professional advice. The publisher, and the authors and editors, expressly disclaim all and any liability to any person, whether a purchaser of this publication or not, in respect of anything and of the consequences of anything done or omitted to be done by any such person in reliance upon the contents of this publication. Opinions expressed by individual writers are not necessarily endorsed by the Canadian Tax Foundation and its members. Photocopying and reprinting. Permission to photocopy or reprint any part of this publication for distribu - tion must be applied for in writing; e-mail: [email protected]. Advertising. Inquiries relating to advertisements should be directed to Christine Escalante, e-mail: [email protected]. Exonération de responsabilité. Le contenu de cette publication ne doit être interprété d’aucune façon comme un avis ou une opinion. L’abonné ou le lecteur ne devrait pas fonder ses décisions sur le contenu de cette publication sans envisager une consultation professionnelle appropriée. L’éditeur et les auteurs réfutent toute responsabilité envers toute personne, qu’elle soit abonnée ou non, rela - tivement à toute conséquence résultant d’actes ou omissions faits en fonction du contenu de la présente publication. Les opinions exprimées par les auteurs particuliers ne sont pas nécessairement appuyées par la Fondation canadienne de fiscalité et ses membres. Photocopie et réimpression. L’autorisation de photocopier ou de réimprimer toute portion de cette publication à des fins de distribution devra être obtenue en adressant une demande écrite à permissions @ctf.ca. Annonces publicitaires. Toutes demandes concernant les annonces publicitaires devront être adres - sées à Christine Escalante, courriel : [email protected].

Canadian Tax Journal Revue fiscale canadienne Published four times per year Publiée quatre fois l’an Price: $75 per copy (plus applicable taxes) Prix : 75 $ l’exemplaire (taxes en sus) Subscription rate: $343.75 per year (plus Abonnement : 343,75 $ par an (taxes en sus) applicable taxes) (Numéro d’enregistrement de TVH : (HST registration no. R-106867260) R‑106867260)

Canadian Tax Foundation La Fondation canadienne de fiscalité 145 Wellington Street West 1250, boul. René-Lévesque ouest Suite 1400 Bureau 2935 Toronto, Canada M5J 1H8 Montréal (Québec) H3B 4W8 Telephone: 416-599-0283 Téléphone : 514-939-6323 Toll Free: 1-877-733-0283 Télécopieur : 514-939-7353 Facsimile: 416-599-9283 Internet : www.fcf-ctf.ca Internet: www.ctf.ca

2020, vol. 68, no. 4 2020, vol. 68, n o 4 (Issued January 2021) (publication : janvier 2021)

ISSN 0008-5111 ISSN 0008-5111

Printed in Canada Imprimée au Canada 5,600 01-21 5,600 01-21

■ iv ■ ■ 2020 VOLUME 68, No 4

Canadian Tax Journal Revue fiscale canadienne

PEER-REVIEWED ARTICLES 931 Interest Relief on Income Tax Debts: Canada Versus the United States MICHAEL H. LUBETSKY 987 Tax Literacy: A Canadian Perspective ANTHONY PHAM, ANTOINE GENEST-GRÉGOIRE, LUC GODBOUT, AND JEAN-HERMAN GUAY

POLICY FORUM 1009 Crisis, Cleanup, and the Prospect of Long-Term Fiscal Change SHIRLEY TILLOTSON 1031 Editors’ Introduction—The GST/HST Responsibilities of Non-Resident E-Commerce Firms ALAN MACNAUGHTON AND DANIEL SANDLER 1035 The GST/HST Obligations of Non-Resident E-Commerce Firms—Jurisprudence and Policy NICHOLAS SHATALOW 1053 Carrying On About Carrying On Business: A Response to “The GST/HST Obligations of Non-Resident E-Commerce Firms” ZVI HALPERN-SHAVIM 1069 Much Ado About Doing Not Much: Some Reflections on the Jurisdiction To Tax Business Transactions MALCOLM GAMMIE

FEATURES 1083 Finances of the Nation: Provincial Debt Sustainability in Canada: Demographics, Federal Transfers, and COVID-19 TREVOR TOMBE 1123 Current Cases: (FCA) Iberville Developments Limited v. Canada; (FCA) Landbouwbedrijf Backx BV v. Canada; (BCCA) Collins Family Trust v. Canada (Attorney General) KIRSTEN KJELLANDER, RYAN L. MORRIS, JOHN SORENSEN, EHSAN WAHIDIE, AND ANITA YUK 1159 Selected US Tax Developments: New Proposed Regulations Under the Section 1061 Carried Interest Rules PETER A. GLICKLICH AND GREGG M. BENSON 1173 Current Tax Reading ROBIN BOADWAY AND KIM BROOKS

■ v ■ ■ Canadian Tax Journal

The Canadian Tax Journal publishes research in, and informed comment on, taxation and public finance, with particular relevance to Canada. To this end, the journal invites interested parties to submit manuscripts for possible publication as peer-reviewed articles, and it especially welcomes work that contributes to the analysis, design, and implementation of tax policies. Articles may be written in English or French and should present an original analysis of the topic. Submitted work, or any substantial part or version thereof, must not have been previously published, either in print or online, and it must not be submitted or scheduled for publication elsewhere. The journal welcomes shorter submissions (from 4,000 to 8,000 words) focused on specific topics as well as longer submissions (to a maximum of 20,000 words) that analyze issues in depth. Submitted articles are subject to a double-blind peer review; authors’ identities are not known to reviewers, and reviewers’ identities are not known to authors. (Non-peer-reviewed contributions may appear elsewhere in the journal.) Final decisions on publication of articles are made by the editors, Alan Macnaughton, Daniel Sandler, and Kevin Milligan, on the advice of reviewers. Many reviewers are drawn from the editorial board (listed on the inside front cover of this journal), although ad hoc reviewers are also consulted. Submissions may be (1) accepted outright; (2) accepted if recommended revisions are made; (3) revised by the authors, as requested by the editors on the advice of reviewers, and resubmitted for further review; or (4) rejected with reasons. The time from submission to the first edi - torial decision is usually two months or less. Prospective contributors should submit a copy of the manuscript to the journal’s editorial department. The preferred method of submission is by e-mail with an attached Word document. E-mail inquiries are welcome: write to [email protected]. Contributors are responsible for providing complete and accurate citations to sources, a detailed abstract (200 to 400 words), and up to six keywords for indexing purposes. The full text of many articles that have appeared in the Canadian Tax Journal since 1991 can be found on the Canadian Tax Foundation’s website: www.ctf.ca. Additionally, the journal in its entirety appears in the Canadian Tax Foundation’s TaxFind, which is updated regularly. The Canadian Tax Journal is indexed in EconLit, ABI Inform, LegalTrac, Index to Canadian Legal Literature, CCH Canadian’s Canadian Income Tax Research Index, Carswell’s Income Tax References, Accounting and Law Index, Current Law Index, Canadian Index, Canadian Periodicals Index, Index to Canadian Legal Periodical Literature, Index to Legal Periodicals and Books, and PAIS International in Print.

■ vi ■ ■ Revue fiscale canadienne

La Revue fiscale canadienne publie des recherches et des commentaires éclairés sur la fiscalité et les finances publiques, particulièrement pertinents pour le Canada. À cette fin, la revue invite les personnes intéressées à soumettre des articles en vue d’une éventuelle publication en tant qu’articles revus par des pairs, et elle accueille tout particulièrement les travaux qui contribuent à l’analyse, à la conception et à la mise en oeuvre des politiques fiscales. Les articles peuvent être rédigés en anglais ou en français et doivent présenter une analyse originale du sujet. Les articles soumis, ou toute partie substantielle ou version des articles, ne doivent pas avoir été publiés antérieurement en format papier ou électronique, et ne doivent pas être soumis ou prévus pour publication ailleurs. Vous pouvez soumettre pour publication, dans la revue fiscale, des articles plus courts (4 000 à 8 000 mots) sur des sujets particuliers ainsi que des articles plus longs (maximum de 20 000 mots) analysant des sujets en profondeur. Les articles soumis sont sujets à une double revue à l’aveugle par des pairs; l’identité des auteurs n’est pas connue des réviseurs et celle des réviseurs n’est pas connue des auteurs (certains articles non soumis à cette révision par des pairs peuvent paraître ailleurs dans la revue.) La décision finale de publier ou non un article est celle des rédacteurs en chef Alan Macnaughton, Daniel Sandler et Kevin Milligan, à la recommandation des réviseurs. Bien que certains réviseurs ad hoc soient aussi consultés, la majorité des réviseurs sont choisis parmi les membres du Comité de rédaction (énumérés à l’endos de la page couver - ture de la revue). Les articles soumis peuvent être 1) acceptés d’emblée; 2) acceptés après modifications; 3) modifiés par les auteurs tel que demandé par les rédacteurs en chef sur l’avis des réviseurs, et resoumis à une nouvelle révision; ou 4) rejetés avec raisons. Le temps écoulé entre la soumission d’un article et la première décision éditoriale est habitu - ellement de deux mois ou moins. Les aspirants contributeurs doivent soumettre un exemplaire de l’article proposé au service éditorial. Il est préférable que la soumission se fasse par courriel, avec une pièce jointe en Word. Les demandes de renseignements par courriel sont les bienvenues. Elles doivent être adressées à [email protected]. Les contributeurs doivent soumettre l’ensemble de leurs sources, un précis détaillé de leurs articles (entre 200 et 400 mots), et jusqu’à six mots clés aux fins d’indexation. On peut trouver le texte intégral de nombreux articles publiés dans la Revue fiscale canadienne depuis 1991 sur le site Internet de le Fondation : www.fcf-ctf.ca. De plus, la revue dans son entier se trouve dans TaxFind, qui est mis à jour régulièrement. La Revue fiscale canadienne est indexée sous EconLit, ABI Inform, LegalTrac, Index to Canadian Legal Literature, Canadian Income Tax Research Index de CCH Canadian, Income Tax References de Carswell, Accounting and Law Index, Current Law Index, Canadian Index, Canadian Periodicals Index, Index to Canadian Legal Periodical Literature, Index to Legal Periodicals and Books, et PAIS International in Print.

■ vii ■ ■ Canadian Tax Foundation

The Canadian Tax Foundation is Canada’s leading source of insight on tax issues. The Foundation promotes understanding of the Canadian tax system through analysis, research, and debate, and provides perspective and impartial recommendations concerning its equity, efficiency, and application. The Canadian Tax Foundation is an independent tax research organization and a regis - tered charity with over 12,000 individual and corporate members in Canada and abroad. For more than 70 years, it has fostered a better understanding of the Canadian tax system and assisted in the development of that system through its research projects, conferences, publications, and representations to government. Members find the Foundation to be a valuable resource both for the scope and depth of the tax information it provides and for its services, which support their everyday work in the taxation field. Government policy makers and administrators have long respected the Foundation for its objectivity, its focus on current tax issues, its concern for improvement of the Canadian tax system, and its significant contribution to tax and fiscal policy.

MEMBERSHIP Membership in the Foundation is open to all who are interested in its work. Membership fees are $399.00 a year, except that special member rates apply as follows: (a) $199.00 for members of the accounting and legal professions in the first three years following date of qualification to practise; (b) $199.00 for persons on full-time teaching staff of colleges, universities, or other educational institutions; (c) $40.00 for students in full-time attend - ance at a recognized educational institution; and (d) $171.00 for persons who have reached the age of 65 and are no longer actively working in tax. Memberships are for a period of 12 months dating from the receipt of application with the appropriate payment. Applications for membership are available from the membership administrator for the Canadian Tax Foundation: facsimile: 416-599-9283; Internet: www.ctf.ca; e‑mail: [email protected].

■ viii ■ ■ Fondation canadienne de fiscalité

La Fondation canadienne de fiscalité est un organisme indépendant de recherche sur la fiscalité inscrit sous le régime des œuvres de charité. Elle compte environ 12 000 membres au Canada et à l’étranger. Depuis plus de 70 ans, la FCF favorise une meilleure compréhension du système fiscal canadien et aide au développement de ce système par le biais de ses projets de recherche, conférences, publications et représentations auprès des gouvernements. Les membres considèrent l’étendue et le détail de l’information offerte par la FCF comme une importante ressource. Ils apprécient également les autres services de la FCF qui facilitent leur travail quotidien dans le domaine de la fiscalité. Les décideurs et administrateurs gouvernementaux respectent depuis longtemps l’objectivité de la FCF, son attention aux questions fiscales de l’heure, sa préoccupation envers l’amélioration du système fiscal canadien et son importante contribution au développement des politiques fiscales.

ADHÉSION Toute personne intéressée aux travaux de la FCF peut en devenir membre. Les droits d’adhésion sont de 399,00 $ par année, à l’exception des tarifs spéciaux suivants : a) 199,00 $ pour les personnes faisant carrière en comptabilité ou en droit pendant les trois premières années suivant leur admission à la profession; b) 199,00 $ pour le personnel enseignant à temps plein dans un collège, une université ou une autre maison d’enseignement; c) 40,00 $ pour les étudiants fréquentant à temps plein une maison d’enseignement reconnue; et d) 171,00 $ pour les personnes qui ont 65 ans et plus et qui ne travaillent plus activement en fiscalité. La période d’adhésion est de 12 mois, à compter de la réception de la demande accompagnée du paiement approprié. Il est possible de se procurer les demandes d’adhésion auprès de l’administratice responsible de l’adhésion à la FCF : télécopieur : 514-939-7353; Internet : www.fcf-ctf.ca; courriel : [email protected].

■ ix ■ ■ BOARD OF GOVERNORS/CONSEIL DES GOUVERNEURS Elected November 27, 2020/Élu le 27 novembre 2020

Albert Anelli, QC1* Ted Gallivan, ON3 John Oakey, NS1 Cheryl Bailey, ON1 Rachel Gervais, ON1* Shamim Panchbhaya, ON3 Jeffery Blucher, NS2 Siobhan Goguen, AB2 Ryan Rabinovitch, QC2 Cathie Brayley, BC2* Paul Grower, MB2 Mitchell Sherman, ON2 Alycia Calvert, ON1* Eric Hamelin, QC3 Michael Smith, AB1 Marlene Cepparo, ON1* Lucy Iacovelli, ON1 Martin Sorensen, ON2* Grace Chow, ON1 Soraya Jamal, BC2 Sean Sprackett, QC1 Michael Coburn, BC2 Dean Landry, ON1* Linda Tang, ON1 Giancarlo Di Maio, ON1 Rick McLean, ON1 John Tobin, ON2 Nancy Diep, AB2 Stefanie Morand, ON2 Hugh Woolley, BC1 Marie-Claire Dy, BC2 Michael Munoz, AB3 Barb Worndl, ON2* Olivier Fournier, QC2

* Executive Committee of the Board of Governors Comité de direction du conseil des gouverneurs 1 Nominee of the Chartered Professional Accountants of Canada 2 Nominee of the Canadian Bar Association 3 Non-sponsor

■ OFFICERS/MEMBRES DE LA DIRECTION Chair/Présidente du conseil Barb Worndl Vice-Chair and Chair of the Executive Marlene Cepparo Committee/Vice-présidente du conseil et présidente du comité de direction Second Vice-Chair/Deuxième vice-présidente Cathie Brayley Past Chair/Présidente sortant du conseil Alycia Calvert Executive Director and Chief Executive Heather Evans Officer/Directrice exécutive et chef de la direction Director, Membership Development and Wayne Adams Community Relations/Directeur, Développement du programme de l’adhésion et relations avec la communauté Regional Director, Quebec/ Lucie Bélanger Directrice régionale du Bureau du Québec Director of Finance and Treasurer/ Shelly Ali Directrice financière et trésorière

■ STAFF/PERSONNEL Events and Web Manager/Directrice Roda Ibrahim des événements et du site Web Librarian/Bibliothécaire Judy Singh Managing Editor/Directeur de Michael Gaughan la rédaction

■ x ■ ■ Canadian Tax Foundation Publications

The Foundation’s publications comprise a range of forms and delivery formats. A number of the regularly issued publications are distributed without charge to Foundation members: the Canadian Tax Journal (4 issues), Perspectives on Tax Law & Policy (4 issues, delivered electronically), Tax for the Owner-Manager (4 issues, delivered electronically), Canadian Tax Focus (4 issues, delivered electronically), and the annual conference report. Monographs and books may be purchased on the Foundation’s website at www.ctf.ca.

Canadian Tax Journal — issued quarterly to members via www.ctf.ca (Non-Members $75 per copy, $343.75 per year). Newsletters Perspectives on Tax Law & Policy — issued quarterly to members via www.ctf.ca. Tax for the Owner-Manager — issued quarterly to members via www.ctf.ca. Canadian Tax Focus — issued quarterly; available to members and non-members via www.ctf.ca. Canadian Tax Highlights — archives (2000-2019) available to members via www.ctf.ca. Conference Reports — Reports of the proceedings of annual tax conferences (Members $40; Non‑Members $95). Latest issue: 2019 (Members $40; Non-Members $350). — Tax Dispute Resolution, Compliance, and Administration in Canada: Proceedings of the June 2012 Conference (Members $30; Non-Members $195) — Collections of papers delivered at regional and special tax conferences (British Columbia, Prairie Provinces, Ontario, and Atlantic Provinces) are available in USB format (Members $445; Non-Members $495). Finances of the Nation — Review of expenditures and revenues and some budgets of the federal, provincial, and local governments of Canada. PDFs for the years 2002-2012 are available on the CTF website at no cost. In 2014, “Finances of the Nation” began to appear as a feature in issues of the Canadian Tax Journal. Monographs 2020. Taxation of Private Corporations and Their Shareholders, 5th edition, Rachel Gervais, John Sorensen, David Stevens, and Dave Walsh, eds. (Members and Non-Members $170; Students $70) 2019. Funding the Canadian City, Enid Slack, Lisa Philipps, Lindsay M. Tedds, and Heather L. Evans, eds. ($40 each) 2018. Tax Treaties After the BEPS Project: A Tribute to Jacques Sasseville, Brian J. Arnold, ed. (Members $60; Non-Members $90) 2018. Reforming the Corporate Tax in a Changing World, School of Public Policy of the University of Calgary (Members $30; Non-Members $50) 2017. Income Tax at 100 Years: Essays and Reflections on the Income War Tax Act, Jinyan Li, J. Scott Wilkie, and Larry F. Chapman, eds. (Members $60; Non-Members $90) 2016. Reform of the Personal Income Tax in Canada, School of Public Policy of the University of Calgary (Members $35; Non-Members $50) 2016. Canadian Taxation of Trusts, Elie S. Roth, Tim Youdan, Chris Anderson, and Kim Brown (Members $150; Non-Members $200; Students $50) 2016. User Fees in Canada: A Municipal Design and Implementation Guide, Catherine Althaus and Lindsay M. Tedds ($40 each) 2015. Timing and Income Taxation, 2d edition, Brian J. Arnold, Colin Campbell, Michael Hiltz, Richard Marcovitz, Shawn D. Porter, and James R. Wilson (Members $25; Non-Members $125; Students $25) 2015. Effective Writing for Tax Professionals, Kate Hawkins and Thomas E. McDonnell, QC (Members $35; Non-Members $40) 2014. After Twenty Years: The Future of the Goods and Services Tax, School of Public Policy of the University of Calgary (Members $25; Non-Members $35)

■ xi ■ 2013. Essays on Tax Treaties: A Tribute to David A. Ward, Guglielmo Maisto, Angelo Nikolakakis, and John M. Ulmer, eds. ($100 each) 2012. Tax Policy in Canada, Heather Kerr, Ken McKenzie, and Jack Mintz, eds. (Members $75; Non-Members $100; Students $50) 2011. Canadian Tax Foundation Style Guide, 5th edition (Members $35; Non-Members $40) 2011. International Financial Reporting Standards: Their Adoption in Canada, Jason Doucet, Andrée Lavigne, Caroline Nadeau, Jocelyn Patenaude, and Dave Santerre (Members $30; Non‑Members $40) 2011. Tax Expenditures: State of the Art — Selected Proceedings of the Osgoode 2009 Conference, Lisa Philipps, Neil Brooks, and Jinyan Li, eds. (Members $45; Non-Members $55; Students $30) 2010. Taxation of Private Corporations and Their Shareholders, 4th edition (Members $75; Non-Members $100; Students $25) Tax Professional Series (Please specify title and author when ordering.) 2003. The Taxation of Business Profits Under Tax Treaties, Brian J. Arnold, Jacques Sasseville, and Eric M. Zolt, eds. (softcover edition, $75) 2003. International Taxation in the Age of Electronic Commerce: A Comparative Study, Jinyan Li. Co-published with International Fiscal Association (Canadian Branch) (Members $95; Non-Members $145; Students $45) 1999. Countering Tax Treaty Abuses: A Canadian Perspective on an International Issue, Nathalie Goyette ($75 each) Canadian Tax Paper Series (Please specify publication number when ordering.) No. 112: 2009. Effective Responses to Aggressive Tax Planning: What Canada Can Learn from Other Jurisdictions, Gilles N. Larin and Robert Duong, with a contribution from Marie Jacques No. 111: 2009. Reforming Canada’s International Tax System: Toward Coherence and Simplicity, Brian J. Arnold (Members $100; Non-Members $125) No. 110: 2006. Financing Education and Training in Canada, 2d edition, Douglas Auld and Harry Kitchen No. 109: 2004. The Canadian Federal-Provincial Equalization Regime: An Assessment, Alex S. MacNevin No. 108: 2004. Venture Capital and Tax Incentives: A Comparative Study of Canada and the United States, Daniel Sandler No. 107: 2002. Municipal Revenue and Expenditure Issues in Canada, Harry M. Kitchen (Members $20; Non‑Members $40) No. 106: 2002. Taxes and the Canadian Underground Economy, David E.A. Giles and Lindsay M. Tedds No. 105: 2000. The Income Tax Treatment of Financial Instruments: Theory and Practice, Tim Edgar No. 104: 1999. Rationality in Public Policy: Retrospect and Prospect, A Tribute to Douglas G. Hartle, Richard M. Bird, Michael J. Trebilcock, and Thomas A. Wilson, eds. No. 103: 1999. Canadian Tax Policy, 3d edition, Robin W. Boadway and Harry M. Kitchen No. 102: 1997. Financing the Canadian Federation, 1867 to 1995: Setting the Stage, David B. Perry No. 101: 1997. General Payroll Taxes: Economics, Politics, and Design, Jonathan R. Kesselman No. 100: 1995. Growth of Government Spending in Alberta, Paul Boothe No. 99: 1995. Financing Education and Training in Canada, Harry Kitchen and Douglas Auld Special Studies in Taxation and Public Finance (Please specify publication number when ordering.) No. 2: 2000. Gambling and Governments in Canada, 1969-1998: How Much? Who Plays? What Payoff? François Vaillancourt and Alexandre Roy No. 1: 1998. Federal-Provincial Tax Sharing and Centralized Tax Collection in Canada, Ernest H. Smith

■ xii ■ ■ Les publications de la Fondation canadienne de fiscalité

Les publications de la Fondation existent sous différentes formes et elles sont disponibles de diverses façons. Certaines de ces publications régulières sont distribuées gratuitement aux membres de la Fondation : la Revue fiscale canadienne (4 numéros), Perspectives en fiscalité et en politique fiscale (4 numéros, offerts électroniquement), Actualités fiscales pour les propriétaires exploitants (4 numéros, offerts électroniquement), Canadian Tax Focus (4 numéros, offerts électroniquement) et le Rapport de la conférence annuelle. Les livres et monographies peuvent être achetés sur le site Web de la Fondation www.fcf-ctf.ca.

Revue Fiscale Canadienne — parution trimestrielle aux membres sur www.fcf-ctf.ca (Non-membres 75 $ par numéro, 343,75 $ par année). Bulletins Perspectives en fiscalité et en politique fiscal — parution trimestrielle disponible aux membres sur www.fcf-ctf.ca. Actualités fiscales pour les propriétaires exploitants — parution trimestrielle disponible aux membres sur www.fcf-ctf.ca. Canadian Tax Focus — parution trimestrielle disponible aux membres et non-membres sur www.fcf-ctf.ca. Faits saillants en fiscalité canadienne — archives (2000-2019) accessibles aux membres sur www.fcf-ctf.ca. Rapports des Conférences — comptes rendus des conférences annuelles sur la fiscalité (Membres 40 $; Non-membres 95 $). Dernière édition : 2019 (Membres 40 $; Non-membres 350 $). — Tax Dispute Resolution, Compliance, and Administration in Canada: Proceedings of the June 2012 Conference (Membres 30 $; Non-membres 195 $) — Collections contenant les travaux présentés aux conférences régionales sur la fiscalité, soit British Columbia, Prairie Provinces, Ontario et Atlantic Provinces, sont disponibles en format USB (Membres 445 $; Non-membres 495 $). Finances of the Nation — Analyse des recettes et dépenses, et quelques budgets, des gouvernements fédéral, provinciaux et locaux au Canada. Les copies PDF pour les années 2002-2012 sont disponibles sur le site Web de la FCF pour téléchargement gratuit. Dans le numéro 62:3 (2014), « Finances of the Nation » est apparu dans les éditions de la Revue fiscale canadienne à titre de nouvelle rubrique. Monographies 2020. Taxation of Private Corporations and Their Shareholders, 5 e édition, Rachel Gervais, John Sorensen, David Stevens et Dave Walsh, éds. (Membres et Non-membres 170 $; Étudiants 70 $) 2019. Funding the Canadian City, Enid Slack, Lisa Philipps, Lindsay M. Tedds et Heather L. Evans, éds. ($40 chacun) 2018. Tax Treaties After the BEPS Project: A Tribute to Jacques Sasseville, Brian J. Arnold, éd. (Membres 60 $; Non-membres 90 $) 2018. Reforming the Corporate Tax in a Changing World, l’École de politique publique de l’Université de Calgary (Membres 30 $; Non-membres 50 $) 2017. Income Tax at 100 Years: Essays and Reflections on the Income War Tax Act, Jinyan Li, J. Scott Wilkie et Larry F. Chapman, éds. (Membres 60 $; Non-membres 90 $) 2016. Reform of the Personal Income Tax in Canada, l’École de politique publique de l’Université de Calgary (Membres 35 $; Non-membres 50 $) 2016. Canadian Taxation of Trusts, Elie S. Roth, Tim Youdan, Chris Anderson et Kim Brown (Membres 150 $; Non-membres 200 $; Étudiants 50 $) 2016. User Fees in Canada: A Municipal Design and Implementation Guide, Catherine Althaus et Lindsay M. Tedds (40 $ chacun)

■ xiii ■ 2015. Timing and Income Taxation, 2 ième édition, Brian J. Arnold, Colin Campbell, Michael Hiltz, Richard Marcovitz, Shawn D. Porter et James R. Wilson (Membres 25 $; Non-membres 125 $; Étudiants 25 $) 2015. Effective Writing for Tax Professionals, Kate Hawkins et Thomas E. McDonnell, QC (Membres 35 $; Non-membres 40 $) 2014. After Twenty Years: The Future of the Goods and Services Tax, l’École de politique publique de l’Université de Calgary (Membres 25 $; Non-membres 35 $) 2013. Essays on Tax Treaties: A Tribute to David A. Ward, Guglielmo Maisto, Angelo Nikolakakis et John M. Ulmer, éds. (100 $ chacun) 2012. Tax Policy in Canada, Heather Kerr, Ken McKenzie et Jack Mintz, éds. (Membres 75 $; Non-membres 100 $; Étudiants 50 $) 2011. Canadian Tax Foundation Style Guide, 5 ième édition (Membres 35 $; Non-membres 40 $) 2011. International Financial Reporting Standards: Their Adoption in Canada, Jason Doucet, Andrée Lavigne, Caroline Nadeau, Jocelyn Patenaude et Dave Santerre (Membres 30 $; Non-membres 40 $) 2011. Tax Expenditures: State of the Art — Selected Proceedings of the Osgoode 2009 Conference, Lisa Philipps, Neil Brooks et Jinyan Li, éds. (Membres 45 $; Non-membres 55 $; Étudiants 30 $) 2010. Taxation of Private Corporations and Their Shareholders, 4 ième édition (Membres 75 $; Non-membres 100 $; Étudiants 25 $) Collection Tax Professional (Prière d’indiquer le titre et le nom de l’auteur sur votre commande.) 2003. The Taxation of Business Profits Under Tax Treaties, Brian J. Arnold, Jacques Sasseville et Eric M. Zolt, éds. (édition brochée, 75 $) 2003. International Taxation in the Age of Electronic Commerce: A Comparative Study, Jinyan Li. Publié en collaboration avec l’Association fiscale internationale (chapitre canadien). (Membres 95 $; Non-membres 145 $; Étudiants 45 $) 1999. Contrer l’abus des conventions fiscales : Point de vue canadien sur une question internationale, Nathalie Goyette (75 $ chacun) Canadian Tax Paper Series (Prière d’indiquer le numéro de la publication.) No 112 : 2009. Des résponses efficaces aux planifications fiscales agressives : leçons à retenir des autres juridictions, Gilles N. Larin et Robert Duong, avec la contribution de Marie Jacques No 111 : 2009. Reforming Canada’s International Tax System: Toward Coherence and Simplicity, Brian J. Arnold (Membres 100 $; Non-membres 125 $) No 110 : 2006. Financing Education and Training in Canada, 2 ième édition, Douglas Auld et Harry Kitchen No 109 : 2004. The Canadian Federal-Provincial Equalization Regime: An Assessment, Alex S. MacNevin No 108 : 2004. Venture Capital and Tax Incentives: A Comparative Study of Canada and the United States, Daniel Sandler No 107 : 2002. Municipal Revenue and Expenditure Issues in Canada, Harry M. Kitchen (Membres 20 $; Non-membres 40 $) No 106 : 2002. Taxes and the Canadian Underground Economy, David E.A. Giles et Lindsay M. Tedds No 105 : 2000. The Income Tax Treatment of Financial Instruments: Theory and Practice, Tim Edgar No 104 : 1999. Rationality in Public Policy: Retrospect and Prospect, A Tribute to Douglas G. Hartle, Richard M. Bird, Michael J. Trebilcock et Thomas A. Wilson, éds. No 103 : 1999. Canadian Tax Policy, 3 : édition, Robin W. Boadway et Harry M. Kitchen No 101 : 1997. General Payroll Taxes: Economics, Politics, and Design, Jonathan R. Kesselman) No 102 : 1997. Financing the Canadian Federation, 1867 to 1995: Setting the Stage, David B. Perry No 100 : 1995. Growth of Government Spending in Alberta, Paul Boothe No 99 : 1995. Financing Education and Training in Canada, Harry Kitchen et Douglas Auld Special Studies in Taxation and Public Finance Series (Prière d’indiquer le numéro de la publication.) No 2 : 2000. Gambling and Governments in Canada, 1969-1998: How Much? Who Plays? What Payoff? François Vaillancourt et Alexandre Roy No 1 : 1998. Federal-Provincial Tax Sharing and Centralized Tax Collection in Canada, Ernest H. Smith ■ xiv ■ canadian tax journal / revue fiscale canadienne (2020) 68:4, 931 - 86 https://doi.org/10.32721/ctj.2020.68.4.lubetsky

Interest Relief on Income Tax Debts: Canada Versus the United States

Michael H. Lubetsky*

PRÉCIS Le paragraphe 220(3.1) de la Loi de l’impôt sur le revenu autorise le ministre du Revenu national à renoncer aux intérêts sur les dettes fiscales ou à les annuler. Ce pouvoir est généralement exercé dans quatre circonstances : lorsque les intérêts se sont accumulés en raison de circonstances indépendantes de la volonté du contribuable; lorsque les intérêts se sont accumulés en raison d’une erreur ou d’un retard de l’Agence du revenu du Canada; lorsque les intérêts accumulés causent un préjudice; ou dans le cadre d’une divulgation volontaire. Au sud de la frontière, l’article 6404 du Internal Revenue Code autorise le secrétaire du Trésor à « annuler » (abate) les intérêts sur les dettes fiscales. En pratique, l’allègement discrétionnaire des intérêts prévu par l’article 6404 n’est possible que dans des circonstances très limitées. L’approche restrictive de l’allègement discrétionnaire des intérêts est toutefois compensée par un plus grand nombre de dispositions d’allègement des intérêts, ainsi que par le pouvoir de règlement amiable (compromise) du secrétaire relativement aux obligations fiscales pour divers motifs, dont certains recoupent les motifs d’allègement des intérêts reconnus au Canada. Cet article compare les régimes canadien et américain d’allègement des intérêts afin de cerner les aspects du régime américain auxquels le Canada devrait prêter plus d’attention. Les différences dans l’approche américaine qui présentent un intérêt particulier sont les suivantes :

n un éventail plus large, et sans doute plus cohérent, de dispositions d’allègement applicables aux intérêts, notamment en ce qui concerne la compensation et les reports rétrospectifs des intérêts;

* Of Davies Ward Phillips & Vineberg LLP, Toronto (e-mail: [email protected]). I wish to thank Marilyn A. Banks, Nat Boidman, Connor Campbell, John Gamino, Nathalie Goyette, Matthias Heilke, Heath Martin, Ashley Perley, Christine Purden, Bobby J. Sood, R. Michael Teper, Geoffrey Turner, Anne-Sophie Villeneuve, and two anonymous reviewers for this journal for their insight and comments during the preparation of this article. I disclose that I have acted and/or am acting for clients in various disputes with the Canada Revenue Agency and the Agence du revenu du Québec that have raised some of the issues discussed in this article. However, the views expressed herein are mine alone.

931 932 n canadian tax journal / revue fiscale canadienne (2020) 68:4

n la compétence de la Cour fiscale des États-Unis à examiner les refus d’annuler les intérêts ou d’accepter une offre de règlement amiable; n le traitement des situations de préjudice et des circonstances extraordinaires sous l’égide du régime de l’offre de règlement amiable (offer-in-compromise), qui permet de prendre en considération l’obligation fiscale sous-jacente en plus des intérêts, et qui permet également de subordonner l’allègement au respect futur par le contribuable de ses obligations de déclaration et de paiement; n dans certains cas plus anciens, la volonté d’utiliser l’allègement des intérêts pour régler des litiges fiscaux complexes et de longue date; n l’absence de délais prescrits relativement au pouvoir du secrétaire d’annuler les intérêts ou d’accepter un règlement amiable à leur égard.

L’étude comparative révèle également que le Canada et les États-Unis accordent un poids différent aux justifications politiques qui sous-tendent l’allègement des intérêts. Le Canada s’attache principalement à faire en sorte que les conséquences de l’inobservation pour les contribuables qui sont des particuliers soient justes et équitables. Les États-Unis, en revanche, se concentrent davantage sur la réhabilitation à long terme des contribuables contrevenants, ainsi que sur la garantie que les intérêts reflètent une juste compensation pour l’utilisation de l’argent du Trésor public par ces contribuables — deux points qui pourraient faire l’objet d’une plus grande attention de ce côté-ci de la frontière.

ABSTRACT Subsection 220(3.1) of the Income Tax Act authorizes the minister of national revenue to waive or cancel interest on income tax debts. This power is typically exercised in four circumstances: where interest has accumulated owing to circumstances beyond a taxpayer’s control; where the interest has accumulated owing to error or delay by the Canada Revenue Agency; where the accumulated interest causes hardship; or in the context of a voluntary disclosure. South of the border, section 6404 of the Internal Revenue Code authorizes the secretary of the Treasury to “abate” interest on tax debts. As a practical matter, discretionary interest relief under section 6404 is available only in very limited circumstances. The restrictive approach to discretionary interest relief is, however, offset by a greater array of interest-relieving provisions, as well as by the power of the secretary to “compromise” tax liabilities on various grounds, some of which overlap with grounds for interest relief recognized in Canada. This article compares the Canadian and US interest relief regimes, with a view to identifying aspects of the US regime that may merit further consideration in Canada. The differences in the US approach that are of particular interest include

n a wider, and arguably more coherent, range of relieving provisions applicable to interest, particularly with regard to interest netting and carrybacks; n the jurisdiction of the United States Tax Court to review refusals to abate interest and/or to accept an offer in compromise; n dealing with situations of hardship and extraordinary circumstances under the aegis of the offer-in-compromise regime, which allows for consideration of the underlying tax liability in addition to the interest, and which also allows for relief to interest relief on income tax debts: canada versus the united states n 933

be made conditional on the taxpayer’s future compliance with filing and payment obligations; n in certain older cases, a willingness to use interest relief to settle longstanding and complex tax disputes; and n the absence of statutory time limits on the power of the secretary to abate or compromise interest.

The comparative study also reveals how Canada and the United States place different weight on policy rationales that underlie interest relief. Canada focuses mainly on ensuring that the consequences of non-compliance for individual taxpayers are fair and equitable. The United States, on the other hand, focuses more on rehabilitating non-compliant taxpayers in the long term, as well as ensuring that interest reflects fair compensation for such taxpayers’ use of the public treasury’s money—both of which could be given greater attention on this side of the border. KEYWORDS: INTEREST n FAIRNESS n RELIEF n UNITED STATES n COMPARATIVE ANALYSIS n TAX ADMINISTRATION

CONTENTS Introduction 934 The ITA Interest Relief Regime 939 Subsection 220(3.1) 939 Other Relieving Provisions 942 Issues Relating to the Application of Subsection 220(3.1) 946 The US Regime: Grounds for Interest Relief 948 Overview of IRC Section 6404 948 IRC Section 6404(a): Interest on Excessive, Unlawful, or Erroneous Taxes 948 IRC Section 6404(c): Small Tax Balances 950 IRC Section 6404(d): Certain Mathematical Errors by the IRS 951 IRC Section 6404(e)(1): Unreasonable Errors and Delays in Performing Ministerial or Managerial Acts 952 IRC Section 6404(e)(2): Erroneous Refunds 957 IRC Section 6404(f ): Erroneous Written Advice by the IRS 959 IRC Section 6404(g): Audit Delays 959 IRC Section 6404(i): Federally Declared Disaster or Terroristic or Military Action 961 The OIC Regime 962 Other Relieving Provisions 970 Judicial Recourses Against a Refusal To Suspend or Abate Interest 973 Key Differences in Income Tax Administration 973 Mandatory Interest Relief 975 IRC Section 6404(h) 976 IRC Sections 6320 and 6330 (CDP Hearings) 980 Observations for Canada 982

934 n canadian tax journal / revue fiscale canadienne (2020) 68:4

INTRODUCTION Bad things happen if you fail to pay federal income taxes when due. Chief Justice John Roberts in Hinck v. US, 550 US 501 (2007), at 502 I doubt that there is anyone on this panel who has not heard more than one heart-breaking story from constituents who find themselves facing crushing back taxes and penalties and interest because they simply were unable to comply with a Tax Code they have no hope of understanding. Congressman Amo Houghton, Hearing Before the Subcommittee on Oversight of the Committee on Ways and Means, 106th Cong., 2d sess. ( January 27, 2000), at 7 Section 161 of the Income Tax Act1 imposes interest on delinquent income tax debts at prescribed rates, generally without regard either for the circumstances that have resulted in the delinquency or for the impact of the added liability on the particular taxpayer.2 Interest can accumulate quickly for reasons beyond the taxpayer’s control and can cause considerable hardship, especially for taxpayers who have fallen behind in paying their taxes precisely because of difficult financial circumstances. This hardship can be magnified by the various penalties that the ITA imposes on non- compliant taxpayers, and those penalties also can bear interest retroactive to the taxation year at issue.3 To alleviate the potentially inequitable consequences of non- compliance, subsection 220(3.1) of the ITA grants broad discretion to the minister of national revenue (“the minister”) to waive or cancel interest and penalties on income tax debts.4 Subsection 220(3.1) was enacted in 1991 to assist taxpayers who fall into non-compliance “because of extraordinary circumstances beyond their control,” including natural disasters, civil disturbances, illness, or erroneous information from revenue officials.5 Consistent with the overall objective of ensuring that the

1 Income Tax Act, RSC 1985, c. 1 (5th Supp.), as amended (herein referred to as “the ITA”). 2 The limited number of relieving provisions in the ITA that attenuate the application of section 161 are outlined below. The methodology for determining the interest rate appears in income tax regulation 4301. 3 ITA subsection 161(11). 4 This article focuses on federal income tax. However, provisions analogous to subsection 220(3.1) of the ITA appear in other federal fiscal statutes, including the Excise Tax Act, RSC 1985, c. E-15, subsections 88(1) and 281.1(1); the Air Travellers Security Charge Act, SC 2002, c. 9, s. 5, section 30(1); the Excise Act, 2001, SC 2002, c. 22, sections 173 and 255.1; and the Softwood Lumber Products Export Charge Act, 2006, SC 2006, c. 13, section 37(1). For a discussion of interest relief in provincial income tax regimes, see Michael H. Lubetsky, “Interest Relief Under the Federal and Provincial Regimes” (2015) 20:1 Tax Litigation 1182-92. 5 Canada, Department of Finance, Explanatory Notes to Legislation Relating to Income Tax (Ottawa: Department of Finance, May 1991), at 354-55. See also Canada Revenue Agency, “The Application of the Fairness Provisions to Penalty and Interest,” memorandum, March 1996. For a thorough discussion of subsection 220(3.1), including its history, its interactions with other provisions of the ITA, leading case law, and various issues concerning its administration, interest relief on income tax debts: canada versus the united states n 935 consequences of non-compliance are fair and equitable, subsection 220(3.1) addresses only the direct consequences of non-compliance (that is, interest and penalties) rather than the tax liability itself. In the United States, section 6601 of the Internal Revenue Code likewise imposes interest on tax debts at prescribed rates.6 And similarly to the ITA, IRC section 6404 authorizes the secretary of the Treasury (“the secretary”) to “abate” tax, interest, and/or penalties in various circumstances. As it pertains to interest, IRC section 6404 seems at first impression to be quite similar to subsection 220(3.1) of theITA , with two immediately obvious points of difference: (1) section 6404 is more directive as to the circumstances in which interest relief is warranted; and (2) section 6404 confers jurisdiction, in certain situations, on the United States Tax Court to hear appeals of decisions to refuse interest abatement. However, a closer examination—including a review of Treasury regulations, published guidance, and case law—reveals that the US interest relief regime differs significantly from Canada’s in both conceptualization and administration. TheIRC ’s penalty and interest abatement provisions are largely distinct, and they aim to respect the underlying policy distinction between penalties and interest—namely, that penalties serve to deter taxpayer non-compliance while interest serves to compensate the public treasury for the use of its money by the taxpayer.7 One consequence of this distinction is that while most penalties can be abated on a showing of “reasonable cause,” interest relief is available only in certain specific, legislatively prescribed circumstances.8 The specified grounds on which the secretary may abate interest are narrow, such that US practitioners routinely advise clients facing large assessments of interest that they have no practical option but to pay it.9 On the other hand,

see John A. Sorensen, “A Comprehensive Review of Penalty and Interest Relief Under the Income Tax Act,” in Report of Proceedings of the Sixty-Seventh Tax Conference, 2015 Conference Report (Toronto: Canadian Tax Foundation, 2016), 41:1-49. An extensive and current bibliography of sources on subsection 220(3.1), including CRA guidance, case law, and academic and practitioner commentary can be found in David M. Sherman, ed., Practitioner’s Income Tax Act, 58th ed. (Toronto: Thomson Reuters, 2020), notes on subsection 220(3.1). 6 Internal Revenue Code of 1986, as amended (herein referred to as “the IRC”). The methodology for determining the interest rate is set out in IRC section 6621. The IRC encompasses estate, gift, payroll, and excise taxes, as well as income taxes, and many of the provisions discussed in this article also apply to some or all of those other taxes. 7 United States, Department of the Treasury, Office of Tax Policy,Report to the Congress on Penalty and Interest Provisions of the Internal Revenue Code (Washington, DC: Department of Treasury, Office of Tax Policy, October 1999) (herein referred to as “the Treasury penalty and interest report”), at 2, 7-8, 16-18, and 32-35 (https://home.treasury.gov/system/files/131/ Report-Penalty-Interest-Provisions-1999.pdf ). 8 Ibid., at 44, 128, and 133-35. 9 Dave DuVal, “The Intriguing World of IRS Interest Abatement,” CPA Practice Advisor, August 23, 2018 (www.cpapracticeadvisor.com/tax-compliance/article/12422740/the- intriguing-world-of-irs-interest-abatement). 936 n canadian tax journal / revue fiscale canadienne (2020) 68:4 the restrictive approach to discretionary interest relief is offset to some extent by a greater array of interest-relieving provisions, as well as by a statutory framework that allows the secretary to “compromise” tax liabilities (including tax, interest, and penalties) on the basis of “doubt as to liability,” “doubt as to collectability,” “economic hardship,” or “public policy or equity grounds” (“the OIC regime,” discussed below). This article compares the circumstances in which interest relief can be obtained and recourses for the taxpayer where relief is refused in the United States and Can- ada, with a view to identifying aspects of the US regime that may merit further consideration in this country. The differences in the US approach that are of particular interest include

n a wider, and arguably more coherent, range of relieving provisions applicable to interest, particularly with regard to interest netting and carrybacks; n the jurisdiction of the US Tax Court to review refusals to abate interest and/or to accept an “offer in compromise” OIC( ) under the OIC regime; n dealing with situations of hardship and extraordinary circumstances under the aegis of the OIC regime, which allows for consideration not only of interest but also of the underlying tax liability, and which also allows for interest relief to be made conditional on the taxpayer’s future compliance with filing and payment obligations; n in certain older cases, a willingness to use interest relief to settle longstanding and complex tax disputes; and n the absence of statutory time limits on the power of the secretary to abate or compromise interest.

This article begins with a brief summary of the ITA’s interest relief regime, including subsection 220(3.1) and other relieving provisions, and then reviews the various grounds for interest relief set out in IRC section 6404, the OIC regime, and other interest-relieving provisions of the IRC. The article then discusses the judicial recourses available to US taxpayers when a request for interest relief is denied. Finally, the article summarizes key points of difference between the Canadian and US regimes, and identifies possible avenues for reform on this side of the border. To assist the reader in navigating the many provisions cited in this article, table 1 provides a rough concordance between grounds of interest relief recognized in Canada, the statutory basis for such relief, and the approximate counterpart provisions in the United States. interest relief on income tax debts: canada versus the united states n 937

TABLE 1 Relief from Interest on Income Tax Debts, Legislative Regimes and Administrative Practices in Canada and the United States

Ground of interest relief Canada United States

Delay caused by revenue ITA subsection 220(3.1); IRC section 6404(d) officials ...... action of the CRA, Information (mathematical errors by IRS Circular IC 07-1R1, “Taxpayer officials when preparing a Relief Provisions” taxpayer’s return) IRC section 6404(e)(1) (errors or delays in performing managerial or ministerial acts) IRC section 6404(g) (audit delays) IRC section 6151(b)(1) (lower-income taxpayers who elect to have the IRS calculate their taxes due) IRC section 6601(c) (delay in assessing amounts subject to waiver) Interest on refunds ITA subsection 220(3.1); IRC section 6404(e)(2) issued in error . . . . . action of the CRA, Information (erroneous refunds) Circular IC 07-1R1, “Taxpayer Relief Provisions” ITA paragraph 160.1(1)(b) (interest does not run on a GST credit or Canada child benefit issued in error) Delay caused by natural ITA subsection 220(3.1); IRC sections 6404(i) and 7508A disaster, terrorist attacks, circumstances beyond a (federally declared disaster or etc...... taxpayer’s control, Information terroristic or military action) Circular IC 07-1R1, “Taxpayer Relief Provisions” Financial hardship/ ITA subsection 220(3.1); IRC section 7122 inability to pay . . . . hardship, Information Circular (the OIC regime, “doubt as to IC 07-1R1, “Taxpayer Relief collectability” and “economic Provisions” hardship”)

(Table 1 is continued on the next page.) 938 n canadian tax journal / revue fiscale canadienne (2020) 68:4

TABLE 1 Continued

Ground of interest relief Canada United States

Delay caused by other ITA subsection 220(3.1); IRC section 7122 personal circumstances circumstances beyond a (the OIC regime, “public policy outside the taxpayer’s taxpayer’s control, Information and equity”) control (bereavement, Circular IC 07-1R1, “Taxpayer illness, etc.) ...... Relief Provisions” ITA subsection 161(6) (delays caused by foreign- currency controls) ITA subsection 161(6.2) (flowthrough share adjustments)

Voluntary disclosures . . ITA subsection 220(3.1); na (The IRS’s voluntary Information Circular IC 00-1R6, disclosure regime does not “Voluntary Disclosures provide interest relief.) Program” Retroactive reduction of ITA subsection 161(7) IRC section 6601(d) tax liabilities owing to carrybacks ...... Offsets of underpayments ITA section 161.1 IRC section 6621(d) by overpayments . . . . Grace period for ITA section 161.2 IRC section 6601(e)(3) payments ...... De minimis amounts . . ITA sections 161.3 and 161.4 IRC section 6404(b) Settlement of tax na (The CRA, as a matter of IRC section 6404(g) disputes ...... policy, does not waive interest (for certain tax shelter in such circumstances.) settlements) IRC section 7122 (the OIC regime, including automatic suspension of interest once an OIC is accepted) Instalments/estimated ITA subsections 161(4), (4.1), na (The IRC does not impose taxes ...... (4.2), and (8), as well as a CRA interest on missed estimated administrative position that taxes, but rather a penalty interest is not assessed on under section 6601(h).) missed instalments by trusts

(Table 1 is concluded on the next page.) interest relief on income tax debts: canada versus the united states n 939

TABLE 1 Concluded

Ground of interest relief Canada United States

Other relieving ITA subsection 161(6.1) IRC section 692 provisions ...... (foreign tax credit adjustments) (forgiveness of unpaid tax and interest for individuals dying in active military service, for astronauts dying in the line of duty, and for some victims of terrorism) IRC section 7508 (periods of active military service in combat zones) Servicemembers’ Civil Relief Act, sections 3937, 4000

CRA = Canada Revenue Agency; GST = goods and services tax; IRC = Internal Revenue Code; IRS = Internal Revenue Service; ITA = Income tax Act; OIC = offer in compromise.

THE ITA INTEREST RELIEF REGIME Subsection 220(3.1) The power conferred on the minister in subsection 220(3.1) to waive or cancel interest has been delegated to various officials in the Canada Revenue AgencyCRA ( ).10 The CRA has published guidance about the circumstances in which it will and will not provide interest relief in, inter alia, two information circulars—IC07-1R111 and IC00-1R6.12 Generally speaking, the CRA groups situations in which interest relief may be warranted into four broad categories:

1. circumstances beyond a taxpayer’s control (such as natural or human-made disasters, civil disturbances or disruption in services, serious illness or

10 These delegations are done administratively pursuant to subsection 220(2.01) of the ITA. A list of officials with the authority to grant interest relief is provided in Canada Revenue Agency, “Income Tax Act—Authorization To Exercise Powers or Perform Duties of the Minister of National Revenue” (www.canada.ca/en/revenue-agency/services/tax/technical-information/ delegation-powers-duties-functions/delegation-ministerial-powers-duties-functions/delegation -under-9.html). 11 Information Circular IC 07-1R1, “Taxpayer Relief Provisions,” August 18, 2017 (www.canada.ca/ en/revenue-agency/services/forms-publications/publications/IC07-1/taxpayer-relief-provisions -1r1.html). 12 Information Circular IC 00-1R6, “Voluntary Disclosures Program,” December 15, 2017 (www.canada.ca/en/revenue-agency/services/forms-publications/publications/IC00-1/ IC00-1r6-voluntary-disclosures-program.html). 940 n canadian tax journal / revue fiscale canadienne (2020) 68:4

accident, or serious emotional or mental distress—for example, because of bereavement);13 2. actions of the CRA (such as processing errors and delays, provision of incorrect information to the taxpayer, or undue delays in concluding an audit or resolving an objection or appeal);14 3. inability to pay or financial hardship (for example, where collection has been suspended because the taxpayer is unable to pay, or where payment of interest would cause “a prolonged [!] inability to provide basic necessities . . . such as food, medical care, transportation, or accommodation”);15 and 4. in the context of the CRA’s voluntary disclosures program.16

Under the general category of “inability to pay or financial hardship,” the CRA will sometimes agree to conditionally waive or cancel interest when negotiating payment arrangements with taxpayers who are unable to pay their entire tax liabilities at once. In such cases, the CRA may cancel “all or part of the interest for the period from when payments start until the amounts owing are paid, as long as the agreed payments are made on time and compliance with the act is maintained.”17 Interest relief in such situations thus serves to both assist the taxpayer and motivate compliance. The list of categories is not exhaustive, and—at least in principle—the CRA may cancel interest in other circumstances.18 Other situations in which relief may be warranted are discussed below. While the CRA has repeatedly stated that it will not use interest relief as a negotiation tool to settle disputes with taxpayers over their tax liabilities,19 it will consider requests to waive or cancel interest in accordance with IC07-1R1 in the course of settlement discussions. Thus, as a practical matter, interest relief may sometimes be put on the table for settlement purposes. Decisions

13 IC 07-1R1, supra note 11, at paragraph 25. Sorensen, supra note 5, at 41:29. IC 07-1R1 labels these cases as “extraordinary circumstances,” but the label is misleading insofar as the courts have held that an event does not have to be “extraordinary” to justify interest relief. See 3500772 Canada Inc. v. Canada (National Revenue), 2008 FC 554 (CanLII), at paragraphs 40-41; and Nixon v. Canada (National Revenue), 2008 FC 917, at paragraph 11. 14 IC 07-1R1, supra note 11, at paragraph 26. Sorensen, supra note 5, at 41:30. 15 IC 07-1R1, supra note 11, at paragraphs 27-28.2. Sorensen, supra note 5, at 41:30-31. The word “prolonged” is curious insofar as it suggests that the CRA would not cancel interest that has caused the taxpayer to be unable to procure food, medical care, transportation, or accommodation for only short periods of time. 16 IC 07-1R1, supra note 11, at paragraphs 42-43; and IC 00-1R6, supra note 12, at paragraphs 6, 9, and 15-18. 17 IC 07-1R1, supra note 11, at paragraph 27(b); see also paragraph (d). 18 Ibid., at paragraph 24. Sorensen, supra note 5, at 41:28-29. 19 See, for example, IC 07-1R1, supra note 11, at paragraph 21; Sorensen, supra note 5, at 41:35-37; and Brooke Sittler, “Review of Penalty and Interest Relief Requests Under the Income Tax Act,” in the 2015 Conference Report, supra note 5, 42:1-12, at 42:3. interest relief on income tax debts: canada versus the united states n 941 to grant interest relief in such circumstances can be included in the minutes of settlement executed with the taxpayer.20 Subsection 165(1.2) of the ITA specifies that no objection is possible from an assessment made pursuant to subsection 220(3.1). Since no objection is possible, no appeal to the Tax Court of Canada is possible either.21 If a taxpayer applies for interest relief and is refused, the only remedy is to apply for judicial review by the Federal Court, which has exclusive jurisdiction over judicial review matters “against any federal board, commission or other tribunal.”22 Federal Court jurisprudence on subsection 220(3.1) is voluminous and establishes, inter alia, that the power to waive interest is discretionary and entitled to deference; the court will quash a refusal by the CRA to grant interest relief only if the decision is “unreasonable.”23 The only statutory restriction on the minister’s power to waive or cancel interest is that the minister may waive or cancel only interest that has accrued during the 10-year period prior to the taxpayer’s application for relief (or, if the minister cancels interest proactively, over the 10 prior taxation years).24 This restriction was added in 2004; previously, the minister had been permitted to cancel or waive interest for any taxation year from 1985 onward.25 With the passage of time, however, it apparently became increasingly difficult for revenue officials to verify claims pertaining to taxation years dating back to the 1980s, and the government was prompted to introduce the 10-year limitation.26 Interest relief for periods prior to those permitted by subsection 220(3.1) can be obtained through a remission order from the governor in council (the federal Cabi- net) pursuant to section 23 of the Financial Administration Act27—a process that “is rarely successful and takes years.”28 The CRA remission guide (an internal reference document not published on the CRA’s website) describes remission orders as an “extraordinary measure” to be granted in “rare instances where relief would be

20 Sherman, supra note 5, notes on subsection 220(3.1). 21 ITA subsection 169(1). 22 Federal Courts Act, RSC 1985, c. F-7, section 18(1)(a). 23 Sherman, supra note 5, provides a long list of applicable sources in his notes on subsection 220(3.1). 24 As explained by the Federal Court of Appeal in Bozzer v. Canada, 2011 FCA 186, the 10-year limitation period set out in subsection 220(3.1) does not prevent the CRA from cancelling or waiving interest accruing within the past 10 years on tax debts that originated earlier. 25 For a review of the earlier legislation and the scope of the 1985 restriction, see Montgomery et al. v. MNR, 95 DTC 5032 (FCA). 26 Canada, Department of Finance, 2004 Budget, Budget Plan, March 23, 2004, at 347. 27 RSC 1985, c. F-11. For discussion and sources, see Sherman, supra note 5, notes on subsection 220(3.1), and annotations to the section entitled “Selected Remission Orders.” 28 Sherman, supra note 5, notes on “Selected Remission Orders.” 942 n canadian tax journal / revue fiscale canadienne (2020) 68:4 justified but cannot be granted under the existing tax laws.”29 The remission guide sets out four “guidelines” for determining when remission may be appropriate:

1. “extreme hardship”; 2. “financial setback coupled with extenuating factors” (with the two main extenuating factors being circumstances beyond a person’s control and taxpayer error leading to overpayment of tax that the CRA should have detected and corrected); 3. “incorrect action or advice on the part of CRA officials”; and 4. “unintended results of the legislation.”30

While the Federal Court has jurisdiction to review a decision by the CRA not to recommend a remission order, challenges to such decisions are almost invariably rejected.31

Other Relieving Provisions Several other statutory provisions in the ITA relieve taxpayers of the obligation to pay interest on income tax liabilities, most of which have received little, if any, judicial consideration. Two of these provisions that offer particular points of contrast with the United States are section 161.1, which concerns offsets, and subsection 161(7), which concerns carrybacks. Section 161.1, enacted with the 1999 budget (in the wake of a comparable provision adopted in the previous year in the United States),32 allows corporations to apply in writing to reallocate overpayments to underpayments if interest is accruing on both amounts. Such an application must be made within 90 days of the reassessment creating the overpayment or underpayment (or final disposition of an objection to or appeal from such reassessment), whichever is later. How this time limit applies in situations where an overpayment or underpayment is varied by a subsequent reassessment is unclear and a matter of some debate.33 If a taxpayer applies within the 90-day period, the underpayment is deemed to have been paid on

29 Canada Revenue Agency, “CRA Remission Guide—A Guide for the Remission of Income Tax, GST/HST, Excise Tax, Excise Duties or FST Under the Financial Administration Act” (October 2014), at section I(1) (available on Taxnet Pro). 30 Ibid., at section III. 31 Sherman, supra note 5, notes on “Selected Remission Orders.” While this article was being prepared for publication, the Federal Court issued its decision in Mokrycke v. Canada (Attorney General), 2020 FC 1027, which seems to be the first reported case where a taxpayer has successfully challenged a decision by the CRA to refuse to recommend the issuance of a remission order. 32 Income Tax Amendments Act, 1999, SC 2000, c. 19, section 48(1). The analogous US provision is IRC section 6621(d), discussed below. 33 Robert Madden, “Interest Offset” (2006) 14:6Canadian Tax Highlights 3-4. interest relief on income tax debts: canada versus the united states n 943 a date specified by the taxpayer that must fall on or after the date of both the overpayment and the underpayment. The minister is obligated to issue a reassessment giving effect to the reallocation and recalculating interest accordingly. The 2000 budget proposed to enact an analogous regime applicable to individuals, but the proposal was not implemented and has remained pending ever since.34 Subsection 161(7) concerns the carryback of losses and other amounts arising in subsequent years. It provides that where there is a retroactive reduction of unpaid tax resulting from such carryback, the interest on the offset amount of tax runs until the latest of four dates: (1) the first day of the year following the carryback year; (2) the filing date for the taxpayer’s return for the carryback year; (3) in certain cases, the filing date for an amendment to the taxpayer’s return for the carryback year; and (4) the date of a written request by the taxpayer that the minister apply the carryback. It follows that if the taxpayer late-files its tax return in the carryback year, amends the return in certain cases, or asks to apply the carryback only after filing the return, interest can accrue on tax liabilities during periods in which the tax has been completely offset by the carryback. Subsection 161(7) can be particularly troubling in the case of carrybacks used to offset audit adjustments, since the CRA interprets the provision as requiring that interest be charged until the date on which the taxpayer actually asks to apply the carryback.35 Since a taxpayer can hardly request to apply a carryback until the CRA proposes the audit adjustment, the CRA’s interpretation results in interest being assessed on notional amounts of tax that have been extinguished, often for considerable periods of time.36 Taxpayers that have applied for discretionary interest relief under subsection 220(3.1) in such situations have had mixed results.37 Some other interest-relieving provisions aim to simplify tax administration for both the minister and taxpayers. In particular, section 161.2 provides a grace period for payment when the minister notifies a taxpayer that a specified amount is due,

34 Canada, Department of Finance, 2000 Budget, Notice of Ways and Means Motion and Supplementary Information, February 28, 2000. See Sherman, supra note 5, notes on section 161.1. 35 It bears noting that the CRA does not apply this approach to subsection 161(7) consistently insofar as it does not assess interest on taxpayers that request a carryback of losses to replace tax credits or other categories of losses, provided that the substitution does not change the amount of tax payable. See CRA document nos. 2005-0141251C6, October 7, 2005; and 2017-0736291E5, March 21, 2018. 36 Ironically, the CRA itself has expressed the view that “[t]he intent of s. 161(7) is to cover situations where a taxpayer ignores payment of taxes, with the expectation of incurring losses to be carried back to erase the tax liability. Where there is no such intent, subparagraph 161(7)(b)(ii) [that is, the filing date for the subsequent year’s return]would be applicable instead of subparagraph 161(7)(b)(iv)”: CRA document no. 2009-0313781I7, April 14, 2009 (emphasis added). 37 See Sorensen, supra note 5, at 41:33-35. 944 n canadian tax journal / revue fiscale canadienne (2020) 68:4 while sections 161.3 and 161.4 provide for the disregarding of small amounts. Both of these provisions were enacted as part of the 2003 budget.38 Amendments to paragraph 160.1(1)(b) enacted in 199439 provide that taxpayers do not have to pay interest on repayments of the goods and services tax (GST) credit or the Canada child benefit that have been, for whatever reason, overpaid.40 This provision addresses the circumstances of lower-income taxpayers who might not readily detect an overpayment of benefits and who may well find the imposition of interest on such overpayments onerous.41 Most of the other interest-relieving provisions in the ITA deal with situations where circumstances beyond a taxpayer’s control make the timely payment of tax difficult or impossible. These provisions include the following:

n Subsection 161(6), a provision dating back to 1950,42 applies when a taxpayer has income in a country with a currency-control regime that prevents the taxpayer from transferring the income to Canada. In such cases, the minister may postpone the deadline for payment of tax on the foreign income if the minister is satisfied that payment of the whole of the part I tax “reasonably attributable to income from sources in that country would impose extreme hardship on the taxpayer.”43 When such a postponement is granted, “no interest is payable under [section 161] . . . during the period of postponement.”44

38 Canada, Department of Finance, 2003 Budget, Budget Plan, February 18, 2003, at 347-50 and 372. Also see Sherman, supra note 5, notes on section 161.2. 39 SC 1994, c. 7, schedule II (SC 1991, c. 49), section 132(1); SC 1994, c. 7, schedule VII (SC 1992, c. 48), section 16(1). 40 “Canada Tax Service—McCarthy Tétrault Analysis,” under Repayment of Excessive Refunds (available on Taxnet Pro). 41 In 2014, the Quebec Ombundsperson excoriated Revenu Québec for lax controls in the provincial equivalent of the GST credit, which resulted in systematic overpayments followed by demands for repayment with interest. Most of the people affected were lower-income taxpayers for whom the demands for repayment caused considerable difficulty. See Michael H. Lubetsky, “Quebec Ombudsman Lambastes Revenu Québec” (2014) 22:10 Canadian Tax Highlights 1-2. 42 SC 1950, c. 40, section 18, enacting subsection 50(7) of the 1948 Income Tax Act (SC 1948, c. 52). 43 ITA subsection 161(6). 44 Ibid. The minister provided longstanding guidance on the application of subsection 161(6) in Interpretation Bulletin IT-351 (Cancelled), “Income from a Foreign Source—Blocked Currency,” November 15, 1976. Counterintuitively, the guidance in the bulletin on what constituted “extreme hardship” for the purpose of subsection 161(6) was significantly less onerous than the current guidance in IC 07-1R1 on “hardship” for the purposes of subsection 220(3.1). IC 07-1R1, supra note 11. interest relief on income tax debts: canada versus the united states n 945

n Subsection 161(6.1), originally enacted in 1983,45 applies when the amount of tax due in a particular year is increased retroactively because of post-year-end adjustments in foreign tax credits. Such adjustments may arise as a result of audit action by foreign tax authorities, or if the taxpayer disposes of property after the taxation year-end in situations described in subsection 126(4.2).46 In such cases, interest starts running from the moment that the taxpayer is informed of the adjustment by foreign tax authorities, or the taxpayer triggers the adjustment by disposing of the property in question. n Subsection 161(6.2), enacted in 1997,47 applies when the amount of tax due in a particular year is increased retroactively because of a reduction in flowthrough share renunciations during the one-year “lookback” period, in accordance with subsection 66(12.73) of the ITA.48 In this case, the amount of additional tax due is deemed to have been paid, and an equal amount is due in the following taxation year. Essentially, this allows the taxpayer a one-year grace period to pay the additional tax without interest.49

In addition, although not part of the ITA, interest on income tax debts is suspended when a court issues a stay of proceedings pursuant to insolvency legislation such as the Companies’ Creditors Arrangement Act or the Bankruptcy and Insolvency Act.50 Finally, subsections 161(2.2), (4), (4.01), (4.1), and (8) of the ITA provide various forms of relief for interest that accrues on deficient tax instalments (that is, amounts of tax that must be paid prior to the balance-due date for the year). A review of these provisions falls outside the scope of this article, given that in the United States, interest is not charged on deficient instalments (which the IRC calls “estimated taxes”); instead, a penalty is imposed.51

45 SC 1980-81-82-83, c. 140, section 108; the provision was amended in 1999 to add paragraph 161(6.1)(b) (SC 1999, c. 22, section 64). 46 Subsection 126(4.2), added by the 1998 budget, is an anti-avoidance measure that seeks to prevent parties from using short-term transactions to trade in foreign tax credits. Canada, Department of Finance, 1998 Budget, Budget Plan, February 24, 1998, at 216-20. 47 SC 97, c. 25, section 50(6). 48 The Federal Court of Appeal reviewed this regime in some detail in Tusk Exploration Ltd v. Canada, 2018 FCA 121. 49 Technical Notes, December 1996 (available on Taxnet Pro). 50 Companies’ Creditors Arrangement Act, RSC 1985, c. C-36; and Bankruptcy and Insolvency Act, RSC 1985, c. B-3. For related court decisions, see National Bank of Canada v. Twin Butte Energy Ltd, 2017 ABQB 608, at paragraphs 27-37; and Nortel Networks Corporation (Re), 2015 ONCA 681, passim. 51 IRC section 6601(h); Treas. reg. section 301.6601-1(f)(6). 946 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Issues Relating to the Application of Subsection 220(3.1) Insofar as subsection 220(3.1) pertains to interest relief (as opposed to penalty relief, which falls outside the scope of this article), both the provision itself and the CRA’s approach to its application have been the object of criticism. The following list identifies some perennial points of contention:52

n Gaps. Certain situations arguably deserving of discretionary interest relief fall outside the categories described in IC07-1R1 and IC00-1R6. While those categories are not exclusive, as a practical matter the CRA is generally hesitant to provide interest relief outside the parameters of its published guidance.53 One well-known gap concerns “wash transactions” (that is, where the ITA imposes interest on notional amounts that do not, in fact, represent any actual increased amount of tax). This can occur, for example, when income is shifted from one taxpayer to a related taxpayer or from one taxation year to a prior taxation year,54 or when carrybacks are applied to offset adjustments to income.55 The impact of interest in these situations can be magnified by the interest differential between overpayments and underpayments (that is, taxpayers pay the minister a higher rate of interest than they receive from the minister),56 as well as the fact that interest paid by the minister is taxable while interest due to the minister is non-deductible.57 Indeed, the CRA’s published guidance on interest relief for the purposes of the Excise Tax Act includes a memorandum

52 Not included in the list are long processing times for applications, an issue that goes beyond the scope of this article. 53 The Federal Court has repeatedly reminded the CRA that, when considering interest relief applications, it must not mechanically apply its published guidance but rather make a truly individualized assessment of the taxpayer’s situation. See, for example, Nixon, supra note 13, at paragraph 6. Interestingly, in 2000, the national taxpayer advocate in the United States (presumably referring to Canada) suggested that efforts by “other tax agencies” to ensure “uniformity and consistency” in interest and penalty relief paralyzed decision making and led to “disastrous” results: United States, Subcommittee on Oversight of the Committee on Ways and Means, Hearing Before the Subcommittee on Oversight of the Committee on Ways and Means, 106th Cong., 2d sess. ( January 27, 2000) (herein referred to as “the SOCWM hearing”), at 34. 54 As discussed above, ITA section 161.1 provides limited relief to corporations in the latter situation. 55 See the discussion above on ITA subsection 161(7). A remission order was granted in such a case in the Jerry Mathews Remission Order, PC 2006-446, June 1, 2006. 56 The prescribed interest rates for the ITA are available at Canada Revenue Agency, “Prescribed Interest Rates” (www.cra-arc.gc.ca/interestrates). During the fourth quarter of 2020, interest charged on overdue income tax was 5 percent, compared to 1 percent or 3 percent paid on overpaid income tax to corporations and individuals, respectively. 57 ITA paragraphs 12(1)(c) and 18(1)(t). interest relief on income tax debts: canada versus the united states n 947

on the reduction of penalties and interest in wash transaction situations,58 but there is no equivalent for income tax. n The 10-year rule. The 10-year rule has long been criticized as being unfair and inconsistent with the underlying policy objectives of subsection 220(3.1), as well as limiting the effectiveness of theCRA ’s voluntary disclosures program.59 n Disputes over interest relief. The Tax Court’s lack of jurisdiction over interest relief is a longstanding source of consternation to the many litigants, often self-represented, who ask for interest relief in the context of their appeals and learn the hard way that the court cannot offer it.60 n Settlement tool. The CRA’s official refusal to use interest relief to settle disputes with taxpayers is an impediment to dispute resolution, given that the provision of such relief may be the only practical way for the CRA and the taxpayer to split the difference between them. TheITA does not authorize the minister to compromise amounts of tax due, and thus any settlement of a tax liability must be on a principled basis supportable on the facts and law.61 n Inconsistency. The CRA has long been criticized for inconsistency in how it applies subsection 220(3.1).62 This point was raised, for example, in an internal program review in 201763 and by the auditor general in 2018.64 In fairness to the CRA, however, some of this inconsistency has apparently resulted from granting decision-making authority over the waiver or cancellation of interest to officials in a range of departments and field offices—an approach that enhances access to interest relief for taxpayers.65

58 GST/HST Memorandum 16.3.1, “Reduction of Penalty and Interest in Wash Transaction Situations,” April 2010. 59 See David M. Sherman, ed., Practitioner’s Income Tax Act, 26th ed. (Toronto: Carswell, 2004), notes on subsection 220(3.1); and Sorensen, supra note 5, at 41:4. 60 Numerous examples can be found in Sherman, supra note 5, notes on subsection 220(3.1). 61 This is known as the Galway principle, established in Galway v. MNR, 74 DTC 6247 (FCA), a discussion of which lies beyond the scope of this article. 62 As Sorensen documents, supra note 5, at 41:27, these criticisms date back to 1994, not long after subsection 220(3.1) was first enacted. 63 Canada Revenue Agency, Audit, Evaluation and Risk Branch, “Internal Audit—Application of the Taxpayer Relief Provisions: Final Report,” March 2018 (www.canada.ca/en/revenue-agency/ programs/about-canada-revenue-agency-cra/internal-audit-program-evaluation/internal -audit-program-evaluation-reports-2018/application-taxpayer-relief-provisions.html). 64 Office of the Auditor General of Canada,2018 Fall Reports of the Auditor General of Canada to the Parliament of Canada: Report 7—Compliance Activities—Canada Revenue Agency (Ottawa: Office of the Auditor General of Canada, 2018) (www.oag-bvg.gc.ca/internet/English/ parl_oag_201811_07_e_43205.html). 65 See “Internal Audit,” supra note 63, at paragraphs 1 and 3.1.1. In the United States, the national taxpayer advocate has recognized that some degree of inconsistency is inevitable, observing that “[a]lthough uniformity and consistency are important goals in any tax system, where multiple reviews of employee decisions are required, employees can feel disenfranchised 948 n canadian tax journal / revue fiscale canadienne (2020) 68:4

THE US REGIME: GROUNDS FOR INTEREST RELIEF Overview of IRC Section 6404 IRC section 6404 appears in the “Procedural and Administration” subtitle of the IRC and thus applies to income, estate, gift, employment, and various excise taxes. It opens with a broad “general rule” that authorizes the secretary to abate taxes, interest, and penalties that are “excessive in amount” (section 6404(a)), followed by an almost equally broad restriction that severely limits the application of section 6404(a) in respect of income, estate, and gift taxes (section 6404(b)). IRC sections 6404 (c) through (g) and (i) identify specific situations where abatement is available—some discretionary and some mandatory—of which sections 6404(e), (g), and (i) are most relevant in the context of interest relief. IRC section 6404(h) confers jurisdiction on the US Tax Court to deal with actions brought by taxpayers against the secretary for failure to abate interest. Various Treasury regulations supplement section 6404, including Treas. reg. sections 301.6401-2, 301.6401-4, and 301.7508A-1, which pertain to IRC sections 6406(e), (g), and (i), respectively.66 Similar to what has been done in Canada, the powers vested in the secretary have been delegated to officials in the Internal Revenue ServiceIRS ( ).67 The IRS’s primary guidance on the application of section 6404 (as well as other interest-relieving provisions in the IRC) appears in section 20.2.7 of the Internal Revenue Manuals (IRM).68

IRC Section 6404(a): Interest on Excessive, Unlawful, or Erroneous Taxes IRC section 6404(a) (the general rule) authorizes the secretary

to abate the unpaid portion of the assessment of any tax or any liability in respect thereof, which— (1) is excessive in amount, or (2) is assessed after the expiration of the period of limitation properly applicable thereto, or (3) is erroneously or illegally assessed.

and may, in an attempt to guard against making mistakes, simply avoid making any decisions. . . . In my opinion, the [Internal Revenue] Service should train employees well and then let them do their work.” See the SOCWM hearing, supra note 53, at 34. 66 Treasury regulations issued pursuant to the IRC can be accessed and navigated on Cornell Law School, Legal Information Institute (LII) (www.law.cornell.edu/cfr/text/26/chapter-I). As discussed below, IRC section 6404(i) is a cross-reference to IRC section 7508A. 67 See IRC sections 7701(a)(11)(B) and (12)(A). A list of IRS officials with the authority to grant interest abatement appears in Internal Revenue Service, “Internal Revenue Manuals” (herein referred to as “the IRM”), at section 1.2.2.13 (www.irs.gov/irm). 68 IRM, supra note 67. Other relevant sections of the IRM include sections 5.1.15.16.3 (additional information about abatement under IRC section 6404(e)(1)) and 25.6.1.10 (overview of different categories of abatement). interest relief on income tax debts: canada versus the united states n 949

Provisions such as section 6404(a) have always appeared in the IRC, and analogous provisions have featured in US tax statutes since at least the 19th century.69 While section 6404(a), especially section 6404(a)(1), superficially resembles subsection 220(3.1) of the ITA, it has a different purpose and is much more limited in its scope. First, section 6404(a) was not enacted as a relieving provision. Rather, its purpose was to provide the secretary with the legal authority to reduce assessments that are subsequently determined to be unfounded or issued in error.70 Without such authority, the assessment would be valid and binding, and the only remedy for the taxpayer would be to pay the assessed amount in full and sue the US government for a refund.71 In other words, section 6404(a) simply aimed to provide taxpayers and the secretary alike with an expeditious process to vary or cancel erroneous assessments; apart from this limited purpose, the provision has been described as “administratively inconsequential.”72 Second, the scope of section 6404(a) is circumscribed by section 6404(b), which states that “[n]o claim for abatement shall be filed by a taxpayer in respect of an assessment of any tax imposed under subtitle A or B [that is, income, estate, or gift taxes].” The case law has repeatedly interpreted this prohibition as also barring claims for the abatement of interest on income, estate, or gift taxes under section 6404(a).73 Although the IRS will “informally” consider requests from taxpayers to abate amounts in respect of income taxes under section 6404(a),74 such abatement is possible only by “the unilateral and voluntary action” of the IRS, and the taxpayer has no “control over the situation.”75

69 Myles A. Cane, “Statutory Abatement Under the Federal Tax Laws” (1955) 33:11 Taxes: The Tax Magazine 845-50, at 845; see also Internal Revenue Service, Office of Chief Counsel Memorandum (OCCM) PMTA 2010-10, February 12, 2010, at 4 (www.irs.gov/pub/lanoa/ pmta_2010-10.pdf ). 70 Cane, supra note 69, at 845. 71 Ibid., at 846 and 850. This is why IRC section 6404(a) refers to the “unpaid portion” of an assessment. That said, the IRS chief counsel now takes the position that section 6404(a) “is permissive and that the IRS is not prohibited from abating the paid portion of assessment”: OCCM 201520010, May 15, 2015 (www.irs.gov/pub/irs-wd/201520010.pdf ). 72 Cane, supra note 69, at 850. 73 Bax v. CIR, 13 F 3d 54 (2d Cir. 1993), at 58; Asciutto v. CIR, TC Memo 1992-564, at 7, note 5; aff’d 26 F 3d 108 (9th Cir. 1994);Amlie v. CIR, 1993 US App. LEXIS 33443 (4th Cir. 1993), at 4; Frantz v. CIR, 1993 US App. LEXIS 39167 (5th Cir. 1993), at 4; Melin v. CIR, 54 F 3d 432 (7th Cir. 1995), at 434; Urbano v. CIR, 122 TC 384 (2004), at 395; Goettee v. CIR, 192 F Appx. 212 (4th Cir. 2006), at 216; aff’g TC Memo 2003-43, motion for reconsideration denied TC Memo 2004-9; Greene-Thapedi v. CIR, 126 TC 1 (2006), at note 21; Kersh v. CIR, 2009 TCM 260, at 15; Corson v. CIR, TC Memo 2009-95, at 10-11; and Adams v. CIR, TC Memo 2019-99, at 8-9. 74 OCCM 201550042, November 30, 2015 (www.irs.gov/pub/irs-wd/201550042.pdf ). 75 Cane, supra note 69, at 848. 950 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Third, in King, the Seventh Circuit76 confirmed that the phrase “excessive in amount” in section 6404(a) simply means “in excess of the correct tax liability.”77 Consequently, the provision does not confer any discretion on the secretary to abate amounts that a taxpayer is lawfully required to pay under the IRC.78 This decision shut down a school of thought that had emerged in the US Tax Court (in the employment tax context, which falls outside the scope of section 6404(b)) that held that section 6404(a) authorized the secretary to grant interest relief in any circumstances where it would be just and fair to do so.79 The Seventh Circuit rejected such thinking as “a monkey wrench tossed into the machinery of tax collection.”80 Indeed, the IRS was so determined to extract this “monkey wrench” that it prosecuted the King appeal to the Seventh Circuit even though the amount at issue was only $200, the taxpayer died while the appeal was pending, and his widow declined to appear, leaving the IRS as the only party before the court.81 Consequently, to the extent that section 6404(a) authorizes the discretionary cancellation of interest on income tax debts, it applies only to situations in which taxpayers discover, and can demonstrate to the IRS’s satisfaction, that they have been assessed an improper amount of tax with interest but have no administrative or judicial recourse available.82 The IRS is not required to abate interest in such situations and apparently does not routinely do so. Moreover, if the IRS declines to provide relief for whatever reason, the taxpayer has no judicial remedy.

IRC Section 6404(c): Small Tax Balances IRC section 6404(c), which has the heading “Small Tax Balances,” authorizes the secretary

to abate the unpaid portion of the assessment of any tax, or any liability in respect thereof, if the Secretary determines under uniform rules prescribed by the Secretary that the administration and collection costs involved would not warrant collection of the amount due.

76 The US federal court system consists of 14 “circuits” (13 based on geography, one based on subject matter), each with a court of appeal (“circuit court”) that serves as the most senior court. Appeals from the US Tax Court are heard, generally, by the geographical circuit court in the circuit of the taxpayer’s residence. 77 King v. CIR, 829 F 3d 795 (7th Cir. 2016), at 798-99; rev’g TC Memo 2015-36. 78 Ibid. The Fifth Circuit came to a similar conclusion in Matter of Brugge v. CIR, 9 F 3d 740 (5th Cir. 1996), at 745 (holding that the accidental abatement by the IRS of an amount of tax that was properly owing was ineffective). 79 See H&H Trim & Upholstery Co. v. CIR, TC Memo 2003-9; Law Offices of Michael B.L. Hepps v. CIR, TC Memo 2005-138; and King, supra note 77 (USTC). 80 King, supra note 77 (7th Cir.), at 798. 81 Ibid., at 797-98. 82 For a discussion of such a situation, see OCCM 201520010, supra note 71. interest relief on income tax debts: canada versus the united states n 951

This provision was enacted as part of the wide-reaching revision of federal income tax legislation leading to the Internal Revenue Act of 1954, but reflected prior administrative practice.83 Most of the judicial consideration of section 6404(c) appears as obiter dicta in cases where the IRS abated tax or penalties in error and then sought to reverse the abatement and reinstate a previous assessment after expiration of the normal limitation period. In the context of these cases, the courts have distinguished two different kinds of abatement: one that cancels an assessment irrevocably, and one that merely suspends the effects of an assessment until circumstances change. Abatements under section 6404(c) have been grouped into the latter category.84 It follows that, while section 6404(c) superficially resembles section 161.3 of the ITA, it does not necessarily provide any permanent relief for the taxpayer.

IRC Section 6404(d): Certain Mathematical Errors by the IRS IRC section 6404(d) authorizes discretionary interest relief where an IRS employee makes a “mathematical error”—defined as “an error in addition, subtraction, mul- tiplication, or division shown on any return”85—when preparing a tax return while “acting in his official capacity to provide assistance to taxpayers in the preparation of income tax returns.” Interest relief may be provided up to the time when the secretary notifies the taxpayer of the resultant deficiency. Section 6404(d) was enacted as part of the Tax Reform Act of 1976.86 It was the first provision ever enacted to provide for the cancellation of interest on a discretionary basis.87 The provision is not referenced in any Treasury regulations, it is discussed only cursorily in the IRM,88 and it does not appear to have ever been judi- cially considered.89 One might speculate that situations described by section 6404(d) arise so rarely—especially nowadays, when mathematical operations are typically performed by computer—that the provision has little, if any, practical effect.

83 Cane, supra note 69, at 845-46 and 846, note 15. 84 See In re Becker, 407 F 3d 89 (2d Cir. 2005), at 97 (review of jurisprudence); and US v. Webb, 2020 US Dist. LEXIS 168583 (S Dist. IN 2020), at 9-11. See also Cane, supra note 69, at 847. Thankfully, there does not seem to be any reported case in Canada in which the minister has sought to reverse a decision to waive or cancel interest and revive a prior assessment. 85 IRC section 6213(g)(2)(A). 86 Pub. L. no. 94-455, at section 1212. 87 United States, Congress, Joint Committee on Taxation, Summary of the Tax Reform Act of 1976 (H.R. 10612, 94TH Congress, Public Law 94-455) (Washington, DC: US Government Printing Office, October 4, 1976), at 65 (https://archive.org/details/summaryoftaxrefo00jcs3176/page/ n5/mode/2up). 88 IRM, supra note 67, at section 20.2.7.4.1. 89 There are a handful of cases that refer to IRC section 6404(d) in passing, generally pointing out that it is not relevant to the dispute before the court. See, for example, Asciutto, supra note 73 (USTC), at 7, note 5. 952 n canadian tax journal / revue fiscale canadienne (2020) 68:4

IRC Section 6404(e)(1): Unreasonable Errors and Delays in Performing Ministerial or Managerial Acts IRC section 6404(e), originally enacted in the Tax Reform Act of 198690 and revised in the Taxpayer Bill of Rights 2 in 1996,91 provides for abatement of interest attributable to “unreasonable errors and delays” by the IRS.92 It contains two paragraphs, the first of which—by far the most litigated abatement provision—reads as follows:

(E) ABATEMENT OF INTEREST ATTRIBUTABLE TO UNREASONABLE ERRORS AND DELAYS BY INTERNAL REVENUE SERVICE (1) In general In the case of any assessment of interest on— (A) any deficiency attributable in whole or in part to any unreasonable error or delay by an officer or employee of the Internal Revenue Service (acting in his official capacity) in performing a ministerial or managerial act, or (B) any payment of any tax described in section 6212(a) [that is, income, estate, gift, and miscellaneous excise taxes] to the extent that any unreasonable error or delay in such payment is attributable to such an officer or employee being erroneous or dilatory in performing a ministerial or managerial act, the Secretary may abate the assessment of all or any part of such interest for any period. For purposes of the preceding sentence, an error or delay shall be taken into account only if no significant aspect of such error or delay can be attributed to the taxpayer involved, and after the Internal Revenue Service has contacted the taxpayer in writing with respect to such deficiency or payment.

While section 6404(e)(1) may appear expansive at first impression, it actually applies to a fairly narrow subset of errors and delays caused by the IRS. First, section 6404(e)(1) applies only to errors and delays that occur after the IRS has contacted the taxpayer in writing with respect to a deficiency or payment. This “contact in writing” typically refers to a letter indicating that a taxpayer’s return has been selected for audit.93 Deficiencies or underpayments that result from IRS error or delay that occur prior to initial contact with a taxpayer do not fall under the aegis of section 6404(e)(1).94

90 Tax Reform Act of 1986, Pub. L. no. 99-514, at section 1563. 91 Taxpayer Bill of Rights 2, Pub. L. no. 104-168, at section 1457. 92 For a discussion of the history and a critical overview of IRC section 6404(e)(1), see John Gamino, “ ‘Widely Perceived as Grossly Unfair’: The Unfilled Promise of Interest Abatement” (2013) 11:2 ATA Journal of Legal Tax Research 38-52 (https://doi.org/10.2308/jltr-50553). 93 Allcorn v. CIR, 139 TC 53 (2012), at 56-57: “[A]s contemplated in the legislative history and by examples in the regulations, the period pursuant to section 6404(e)(1) may begin when the IRS commences an audit.” 94 This point was made during the congressional debates that led to the enactment of IRC section 6404(e): HR rep. no. 99-426, 99th Cong., 1st sess. (1985), at 844-45; S rep. no. 99-313, 99th Cong., 2d sess. (1986), at 208; and HR conf. rep. no. 99-841, 99th Cong., 2d sess. (1986), at II-811 (www.bradfordtaxinstitute.com/Endnotes/Confrpt99-8412.pdf ). These passages have been repeatedly cited by the US Tax Court: Krugman v. CIR, 112 TC 230 (1999); Sims v. CIR, interest relief on income tax debts: canada versus the united states n 953

Second, errors or delays are taken into account “only if no significant aspect of such error or delay can be attributed to the taxpayer involved.” The case law has endorsed a strict approach to this limitation against taxpayers, holding that filing errors by a taxpayer that have no bearing on the amount of tax payable, but that result in interest being incurred on notional amounts, nevertheless disqualify interest relief under section 6404(e)(1).95 Third, unreasonable errors or delays must be attributable to officers and -em ployees specifically of the IRS. Errors or delays attributable to employees of other government departments96 or of the courts,97 or to new case law that unexpectedly renders a taxpayer’s previous reporting position incorrect,98 do not fall within the scope of the provision. Fourth—and perhaps most importantly—the unreasonable errors or delays must be attributable to the performance of “ministerial” or “managerial” acts.99 Treas. reg. section 301.6404-2 defines both terms and provides various examples. A ministerial act is

TC Memo 1999-414, at 8-9; Hawksley v. CIR, TC Memo 2000-354, at 18-19; Donovan v. CIR, TC Memo 2000-220, at 6-7; Pettyjohn v. CIR, TC Memo 2001-227, at 16-17 and 19; Downing v. CIR, 118 TC no. 2 (2002), at 15; Wright v. CIR, TC Memo 2004-69, at 11; aff’d 125 F Appx. 547 (5th Cir. 2005); Mekulsia v. CIR, TC Memo 2003-138, at 19, motion to reconsider denied; aff ’d 398 F 3d 601 (6th Cir. 2004);Guerrero v. CIR, TC Memo 2006-2001, at 7-8 and 10; Matthews v. CIR, TC Memo 2008-126, at 17-19; Bucaro v. CIR, TC Memo 2009-247, at 17-18; McGaughy v. CIR, TC Memo 2010-183, at 10-12; and Hull v. CIR, TC Memo 2014-36, at 17-18. 95 Larkin v. CIR, 626 F Appx. 913 (11th Cir. 2015); aff’g TC Memo 2014-195 (“M. Larkin”). This case concerned interest on a notional tax liability in 2006 that was offset by a loss carryover from 2005 that had been improperly allocated on the taxpayers’ returns. (The taxpayers had simply carried the loss forward, while under the IRC carryover rules, they needed to carry it back to 2003, obtain a refund for 2003, and apply the refund to 2006.) The taxpayers filed amended returns in 2008 that corrected the problem. No tax was due, but the IRS assessed interest on a notional liability from 2006 until the refund was applied to offset it in 2008. Abatement was refused both on the basis that the Larkins’ own errors created the situation, and also because of the lack of any error attributable to a “ministerial” or “managerial” act of the IRS (as discussed below). 96 See Nerad v. CIR, TC Memo 1999-376 (the US navy provided an incorrect form W-2 and refused to correct it). 97 OCCM 201526008, April 21, 2015 (www.irs.gov/pub/irs-wd/201526008.pdf ). 98 Sims, supra note 94, at 9: “Petitioners’ complaint is really one against fate—they filed their return just before the Supreme Court provided definitive guidance on the correct tax treatment to be accorded damages like those awarded here. Section 6404(e) simply does not reach this type of complaint.” 99 When originally enacted in 1986, IRC section 6404(e) applied only to “ministerial” acts of IRS officials. In 1996, Congress expanded section 6404(e) to include both “ministerial” and “managerial” acts, effective for taxation years beginning after July 30, 1996. Much of the case law on section 6404(e) pertains to taxation years prior to the cutoff date, and some of those cases might well be decided differently today. See, for example,Jacobs v. CIR, TC Memo 2000-123, at 23 (delay caused when the appeals officer attended training courses); Jean v. CIR, TC Memo 2002-256, at 9-10 (delay caused when the auditor attended training courses); Dadian v. CIR, TC Memo 2004-121, at 13-14 (delay caused when only one employee was 954 n canadian tax journal / revue fiscale canadienne (2020) 68:4

a procedural or mechanical act that does not involve the exercise of judgment or discretion, and that occurs during the processing of a taxpayer’s case after all prerequisites to the act, such as conferences and review by supervisors, have taken place.100

Examples given include (1) delays in transferring an audit file to a different district office after the transfer has been approved; (2) delays in issuing a notice of deficiency once it has been prepared and reviewed; and (3) underpayments caused by providing a taxpayer with out-of-date information with respect to the amount that the taxpayer is required to pay.101 A managerial act is

an administrative act that occurs during the processing of a taxpayer’s case involving the temporary or permanent loss of records or the exercise of judgment or discretion relating to management of personnel.102

Examples given include (1) delays caused when an auditor or revenue agent is temporarily or permanently absent owing to illness, training, or reassignment, and the auditor’s or agent’s files are reassigned to another officer; and (2) delays caused by misplacing a taxpayer’s case file.103 The Treasury regulations exclude from “ministerial” and “managerial” acts “general administrative decisions” and decisions “concerning the proper application of federal tax law (or other federal or state law).”104 These exclusions bar relief in a broad swath of situations in which undue delay may be clearly attributable to the IRS, including delay caused by

assigned to handle a large number of settlements); Bo v. CIR, TC 2005-150, at 19-20 (delay caused when the auditor took maternity leave and attended job-related training); Jaffe v. CIR, TC Memo 2004-122, at 15-16; aff’d 175 F Appx. 853 (9th Cir. 2006) (delay caused when only one employee was assigned to handle a large number of settlements); and Mathia v. CIR, TC Memo 2009-120, at 34-36 and 39 (delay caused by management of the settlement process). 100 Treas. reg. section 301.6404-2(b)(1). See also HR rep. no. 99-426, supra note 94, at 845; and S rep. no. 99-313, supra note 94, at 209: “A ministerial act is a procedural action, not a decision in a substantive area of tax law.” 101 Treas. reg. section 301.6404-2(c), examples 1, 2, and 11. 102 Treas. reg. section 301.6404-2(b)(2). 103 Treas. reg. section 301.6404-2(c), examples 3 to 6. The loss of a taxpayer’s file also constitutes a “ministerial act” that can support interest abatement. See Palihnich v. CIR, TC Memo 2003- 297, at 10-11: “The Commissioner’s loss of a taxpayer’s return is a procedural or mechanical act that does not involve the exercise of discretion or judgment by the Commissioner. . . . Respondent has identified no purpose served by losing a taxpayer’s returns, and we know of none.” See also Bo, supra note 99, at 20. 104 Treas. reg. section 301.6404-2(b)(1). As Gamino (supra note 92, at 47-48 and 51) points out, the lines between general administrative decisions, decisions concerning the proper application of federal tax law, managerial acts, and ministerial acts are fuzzy and not always applied consistently by the US Tax Court. interest relief on income tax debts: canada versus the united states n 955

n holding a file in abeyance pending audit, investigation, or resolution of a related file;105 n prioritizing the processing of certain kinds of files over others;106 n taking erroneous, unwarranted, or untenable positions during audit or collection activities;107 n providing incorrect guidance to taxpayers on filing requirements or the tax consequences of different transactions;108 n taking unnecessarily aggressive collection action that impedes the taxpayer’s ability to pay;109

105 Treas. reg. section 301.6404-2(c), example 8. Much of the case law on IRC section 6404(e)(1) deals with situations where a taxpayer’s audit is held in abeyance pending the resolution of related proceedings, including criminal investigations, prosecutions, and/or related partnership audits and litigation. The US Tax Court has invariably found such delays not to be an appropriate basis for interest relief. See Lee v. CIR, 113 TC 145 (1999), at 149; Taylor v. CIR, 113 TC no. 16 (1999), at 7-13; aff’d on other grounds 9 F Appx. 700 (9th Cir. 2001); Gorgie v. CIR, TC Memo 2000-80, at 10; Hanks v. CIR, TC Memo 2001-319, at 7; Berry v. CIR, TC Memo 2001-323, at 12-14; Mekulsia, supra note 94 (USTC), passim; Beagles v. CIR, TC Memo 2003-67, at 7-12 (including an overview of the Tax Equity and Fiscal Responsibility Act of 1982 [TEFRA], Pub. L. no. 97-248, and the congressional response to the challenges caused by tax shelter audits); Jaffe, supra note 99 (USTC), at 13-15; Kemp v. CIR, TC Memo 2004-139, at 10-12; Dadian, supra note 99, at 12-13; Kimball v. CIR, TC Memo 2008-78, at 18; Matthews, supra note 94, at 24-26; McGaughy, supra note 94, at 14-16; Mathia, supra note 99, at 34-35; Howell v. CIR, TC Memo 2007-204, at 9-10; Corson, supra note 73, at 7, 14-15; Swanson v. CIR, TC Memo 2010-131; Goettee, supra note 73 (4th Cir.), at 219; Larkin v. CIR, TC Memo 2010-73 (“T. Larkin”); and Roudakov v. CIR, TC Memo 2017-121, at 9. 106 Treas. reg. section 301.6404-2(c), example 8; Goettee, supra note 73 (4th Cir.), at 219-20; Foote v. CIR, TC Memo 2015-187, at 26; aff ’d docket no. 15-73728 (9th Cir. 2017); Mekulsia, supra note 94 (USTC), at 24; Hornbacker v. CIR, TC Memo 2016-64, at 18-19; and Jacobs, supra note 99, at 24. 107 Banat v. CIR, 5 F Appx. 36 (2d Cir. 2001), at 37; Kuykendall v. CIR, TC Memo 2008-277; Foote, supra note 106 (USTC), at 19-21; Gorgie, supra note 105, at 9-10; and Krehnbrink v. CIR, no. 17-2072 (6th Cir. June 18, 2018) (unreported); 2018 US App. Lexis 16467 (“Krehnbrink #1”), at 4-5. 108 Sainte-Yves v. CIR, TC Memo 2002-158 (allegedly incorrect IRS advice on the deductibility of losses); Nelson v. CIR, TC Memo 2004-34, at 7 (incorrect advice from the IRS about the tax treatment of amounts inherited from a retirement savings account); Bo, supra note 99, at 19 (delay created when the IRS auditor advised the taxpayer to make an OIC rather than representations about the substantive issues); Guerrero, supra note 94, at 11 (allegedly incorrect IRS advice about the tax treatment of funds withdrawn from pension plans); Geiger v. CIR, TC Memo 2010-133 (“bad advice” from the IRS about the filing of a form and the procedure to amend a return); and Estate of Telesmanich v. CIR, TC Memo 2011-181 (incorrect advice from the IRS about when to file a return and make a payment pertaining to an inheritance in a foreign country). 109 Sandberg v. CIR, TC Memo 2011-72, at 7 and 15 (IRS refusal to subordinate a federal tax lien, which allegedly prevented the taxpayer from obtaining financing to pay a tax debt). 956 n canadian tax journal / revue fiscale canadienne (2020) 68:4

n not reminding taxpayers about outstanding tax liabilities;110 n not attempting to track down a taxpayer to redeliver a notice of deficiency that is marked “return to sender”;111 n deciding not to advise members of a partnership that a “tax matters partner” is under criminal investigation and taking steps to remove him or her;112 n any decisions concerning the conduct of litigation113 or decisions taken in the course of settlement discussions;114 n shutting down a division and transferring a taxpayer’s file to another city as a consequence;115 and n implementing a new computer system.116

As discussed below, the courts have endorsed a restrictive approach to section 6404(e)(1), repeatedly citing congressional reports that indicate that the provision was not intended to “be used routinely to avoid payment of interest,” but rather was to be used only “where failure to abate interest would be widely perceived as grossly unfair.”117 The courts have also repeatedly reminded taxpayers that hardship is not a relevant consideration for relief under section 6404(e)(1).118

110 Smith v. CIR, TC Memo 2002-1, at 9: “Although intrigued by the novelty of a complaint that the IRS has failed to hound the taxpayer enough, we find therein no basis to grant petitioner relief.” See also OCCM PMTA 2011-27, August 26, 2011 (www.irs.gov/pub/lanoa/ pmta_2011-27.pdf ). 111 Brown v. CIR, TC Memo 2000-61, at 9-10. 112 Mekulsia, supra note 94, at 30-32 (USTC) and 604-6 (6th Cir.). 113 Lee, supra note 105, at 150-51; Mathia, supra note 99, at 35; Swanson, supra note 105, at 13; Corson, supra note 73, at 14-15; Goettee, supra note 73 (4th Cir.), at 219; and Foote, supra note 106 (USTC), at 29-30. 114 Corson, supra note 73, at 14-15; Berry, supra note 105, at 12-14; Wright, supra note 94 (5th Cir.), at 549; Foote, supra note 106 (USTC), at 27-28; and Jacobs, supra note 99, at 23-24. 115 Paneque v. CIR, TC Memo 2013-48, at 27-28. 116 Treas. reg. section 301.6404-2(b)(1). 117 HR rep. no. 99-426, supra note 94, at 844; S rep. no. 99-313, supra note 94, at 208, cited in Lee, supra note 105, at 149; Douponce v. CIR, TC Memo 1999-398, at 6; Jacobs, supra note 99, at 14; Pettyjohn, supra note 94, at 17; Brown, supra note 111, at 7; Bucaro, supra note 94, at 17-18; Sims, supra note 94, at 7; Taylor, supra note 105 (USTC), at 8; Gorgie, supra note 105, at 8; Smith, supra note 110, at 7; Wright, supra note 94 (USTC), at 10; Mekulsia, supra note 94 (USTC), at 20-21; Landvogt v. CIR, TC Memo. 2003-217, at 9; Kemp, supra note 105, at 7-8; Krugman, supra note 94, at 14-15; Corson, supra note 73, at 15; Hawksley, supra note 94, at 19; Donovan, supra note 94, at 6; Nelson, supra note 108, at 8; Jean, supra note 99, at 7-8; Guerrero, supra note 94, at 7; Matthews, supra note 94, at 14; Hancock v. CIR, TC Memo 2012-31, at 7, note 8; Allcorn, supra note 93, at 57; Paneque, supra note 115, at 18 and 30; McGaughy, supra note 94, at 10; Mathia, supra note 99, at 35; Howell, supra note 105, at 8; Swanson, supra note 105, at 10; T. Larkin, supra note 105, at 7; Foote, supra note 106 (USTC), at 14; Hull, supra note 94, at note 8; Hornbacker, supra note 106, at 16; Prakash v. CIR, TC Memo 2016-176, at 8; Santana v. CIR, TC Memo 2017-14, at 18; and Roudakov, supra note 105, at 10. 118 See, for example, Cosgriff v. CIR, TC Memo 2000-241, at 6. interest relief on income tax debts: canada versus the united states n 957

IRC Section 6404(e)(2): Erroneous Refunds The second paragraph of IRC section 6404(e) concerns refunds issued in error:

(E) ABATEMENT OF INTEREST ATTRIBUTABLE TO UNREASONABLE ERRORS AND DELAYS BY INTERNAL REVENUE SERVICE . . . (2) INTEREST ABATED WITH RESPECT TO ERRONEOUS REFUND CHECK The Secretary shall abate the assessment of all interest on any erroneous refund under section 6602 until the date demand for repayment is made, unless— (A) the taxpayer (or a related party) has in any way caused such erroneous refund, or (b) such erroneous refund exceeds $50,000.

The IRM contains a lengthy section (section 21.4.5) on erroneous refunds. It groups such refunds into four broad categories based largely on the mechanisms primarily used to recover them. Category A refunds are recovered primarily through deficiency proceedings (discussed below); category B and C refunds are recovered primarily through assessment or a credit entry; and category D refunds can generally be recovered only through civil action (or, in some cases, an offset). Section 21.4.5 also lists many of the causes of erroneous refunds, including some completely outside the control of taxpayers, such as the misapplication of payments by the IRS, the incorrect computation of interest, direct deposit of a refund to the wrong bank account, and the issuance of refunds beyond the statute-barred date.119 To this list can be added refunds issued to wrong taxpayers owing to identify theft, which the IRM treats separately.120 IRC section 6404(e)(2) applies only to refunds “under section 6602”—that is, erroneous refunds of tax, interest, or penalties that are recoverable by a civil action. The US Tax Court has issued inconsistent decisions on whether section 6404(e)(2) applies only where the IRS demands repayment of a refund through a civil suit, or whether the provision also applies when the IRS pursues repayment through deficiency proceedings or assessment.121

119 IRM, supra note 67, at section 21.4.5.3(1). 120 Ibid., at section 25.23. 121 Contrast Baral v. CIR, TC Memo 2009-113, at 13 (holding that IRC section 6404(e)(2) does not apply to refunds that the IRS recoups by assessment rather than a civil suit) with Allcorn, supra note 93, at 59-61 (holding the opposite). Allcorn refers to Baral and purports to “distinguish” it, but it is difficult to reconcile the reasoning of the two decisions. Allcorn is probably the more persuasive authority, given that (1) it is more recent; (2) it analyzes the issue of what kinds of refunds are subject to section 6404(e)(2) in much more detail; and (3) in Baral, the erroneous refund at issue was clearly triggered by errors on the taxpayer’s return, thus providing an alternative basis for the non-applicability of section 6404(e)(2). See also the dicta in Pettyjohn, supra note 94, at 23-24, in which the court seems to follow the view in Allcorn. 958 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Section 6404(e)(2) provides for mandatory abatement of interest on erroneous refunds of up to $50,000,122 unless the taxpayer or a related party has “in any way” caused the erroneous refund. In the case of an erroneous refund over $50,000, or an erroneous refund that a taxpayer or a related party has “in any way” contributed to causing, section 6404(e)(2) has been construed as authorizing the IRS to abate interest on a discretionary basis.123 As explained in the IRM, the IRS will consider cancelling interest on an erroneous refund over $50,000 that it does not consider to have been caused by the taxpayer or a related party.124 In denying the right to abatement if the taxpayer or a related party “has in any way caused” the issuance of the erroneous refund, the wording of section 6404(e)(2) seems very restrictive, particularly in comparison with the analogous provision in section 6404(e)(1) (that is, “no significant aspect of such error or delay can be attributed to the taxpayer involved”). However, the case law on section 6404(e)(2) has endorsed a “flexible” approach, in which a taxpayer does not lose entitlement to abatement simply by “adding confusion” to a clear IRS error.125 The case law has also held that the IRS bears the burden of proving that a taxpayer “caused” an erroneous refund, and has used the presumption to save the entitlement of taxpayers to abatement.126 The $50,000 threshold has been criticized as arbitrary and leading to unfair results when taxpayers, through no fault of their own, unexpectedly find themselves saddled with large interest-bearing obligations to the IRS.127 The case law has also lamented how the “poor draftsmanship and the lack of regulations relating to section 6404(e)(2) invites speculation and confusion” about its proper application.128

122 A $1 million threshold was proposed initially but was scaled down over the course of congressional proceedings. See HR conf. rep. no. 99-841, supra note 94, at II-811; and Paul S. Caselton, “Abatement of Interest Under Section 6404(e)” (1987) 65:10 Taxes: The Tax Magazine 647-51, at 648 and 650. 123 See Caselton, supra note 122, at 650-51; and Allcorn, supra note 93, at 63-66. 124 IRM, supra note 67, at section 20.2.7.7(2). 125 Allcorn, supra note 93, at 62-63, citing Converse v. US, 839 F Supp 1274 (N Dist. OH 1993); and Lindstedt v. US, 78 AFTR 2d (RIA) 96-6211 (F Claim Ct. 1996). 126 United States v. Putzel, 2008 US Dist. LEXIS 50461 (Dist. Ct. NH 2008); aff’d on summary judgment 2008 US Dist. LEXIS 64102 (Dist. Ct. NH 2008) (a separated couple, filing separate returns for the first time since their separation, double-claimed a portion of amounts that they had remitted over the course of the year). 127 See Caselton, supra note 122, at 650; and US v. Egypt Planning Co, 70 AFTR 2d (RIA) 92-5724 (N Dist. MS 1992), at 3-4. See also US v. Mount Sinai Med Ctr of Fla, 2005 US Dist. LEXIS 46531 (S Dist. FL 2005), at 5-7. 128 Egypt Planning, supra note 127, at 4. interest relief on income tax debts: canada versus the united states n 959

IRC Section 6404(f): Erroneous Written Advice by the IRS IRC section 6404(f ) requires that the secretary abate any penalty (that is, not interest) “attributable to erroneous advice furnished to the taxpayer in writing by an officer or employee of the Internal Revenue Service,” provided that “the written advice was reasonably relied upon by the taxpayer” and was “in response to a specific written request of the taxpayer.” The fact that section 6404(f ) is limited to penalties implies, a contrario, that abatement of interest is not possible when a taxpayer makes a written request to the IRS for advice, the IRS provides the advice, and the taxpayer reasonably relies on it and underpays taxes as a result. This understanding of section 6404(f) is confirmed in Treas. reg. section 301.6404.3 and in the case law.129 This is a remarkable limitation on the circumstances in which interest relief is available and a sharp contrast with Canadian practice.130 That said, as discussed below, reliance on erroneous advice by the IRS is recognized as a ground for potential relief under the OIC regime.

IRC Section 6404(g): Audit Delays IRC section 6404(g), which concerns “audit delays,” mandatorily suspends the accrual of interest on income tax liabilities starting three years after the latest of three dates: (1) the filing of a taxpayer’s return; (2) the due date for filing; and (3) in some cases, the filing of additional documents following the filing of a return. The interest suspension lasts until the secretary provides “a notice to the taxpayer specifically stating the taxpayer’s liability and the basis for the liability.” The provision applies only to individuals (that is, not corporations)131 who file their tax returns on time (taking into account any extensions) and is subject to six broad exceptions: (1) liabilities shown on the tax return, (2) certain penalties, (3) fraud, (4) “gross misstatements,”132 (5) criminal penalties, and (6) “reportable” or “listed” transactions (that is, certain tax shelters).133 Section 6404(g) was enacted as part of the IRS Restructuring and Reform Act of 1998,134 a suite of reforms that aimed to provoke a shift in culture at the IRS toward

129 Howell, supra note 105, at 12. 130 See IC 07-1R1, supra note 11, at paragraph 26(b), which recognizes that interest relief may be available because of “errors in [CRA] material available to the public, which led taxpayers to file returns or make payments based on incorrect information.” 131 This limitation was applied in CreditGuard of America, Inc v. CIR, 149 TC 370 (2017), at 374, note 3. 132 “Gross misstatements” are defined in Treas. reg. section 301.6404-4(b)(4) and include, inter alia, a misstatement equal to or greater than 25 percent of the “gross income” reported on a return. See CCCM 20133702F, August 15, 2013, at 5-7 (www.irs.gov/pub/irs-lafa/20133702f.pdf ). 133 For an overview of “reportable” and “listed” transactions, see CCCM 20133702F, supra note 132, at 3-5; and Internal Revenue Service, “Instructions for Form 8886 (Rev. December 2019), Reportable Transaction Disclosure Statement,” December 20, 2019 (www.irs.gov/pub/ irs-pdf/i8886.pdf ). 134 Pub. L. no. 105-206 (herein referred to as “the RRA 98”). 960 n canadian tax journal / revue fiscale canadienne (2020) 68:4 greater efficiency and transparency with taxpayers.135 Section 6404(g) aimed in particular to address the allegation (which emerged during the congressional hearings leading to the adoption of the RRA 98) that the IRS was systematically victimizing taxpayers by intentionally dragging out audits.136 As originally enacted, section 6404(g) essentially gave the IRS one year (to take effect after a transitional period) to conduct its audits, and if an audit started or stretched beyond that period, no interest would accrue on any tax debts that the audit may reveal until the taxpayer was formally notified of them. The provision was a big-ticket item that was pre- dicted to reduce federal revenues by more than $2 billion in the 10 years following its adoption.137 Reality soon set in, however, when the IRS was confronted with auditing an array of highly complex and widely distributed tax shelters involving thousands of taxpayers, in which a common practice was to “inundate the IRS with numerous tax and information returns in an attempt to muddle the true source or amount of losses, or the basis in particular assets.”138 In response to the challenge of auditing these tax shelters and dealing properly with all the taxpayers concerned, Congress amended section 6404(g) to add the exceptions for gross misstatements, listed transactions, and reportable transactions, and also to extend the IRS’s audit period to 18 months and later three years (coinciding with the maximum period during which the IRS can audit a taxpayer and propose adjustments).139 The cumulative effect of all these amendments was to leave section 6404(g)—as at least one practitioner lamented— “moribund.”140 Congress gave section 6404(g) a new (albeit transient) vocation in 2005, after the IRS announced a “settlement initiative” that dealt with 21 types of tax shelter transactions that it considered to be abusive, including debt straddles, intellectual property donations, and offsetting foreign-currency contracts.141 Under the settlement initiative, taxpayers would concede 100 percent of the transaction’s tax benefits and pay either 25 percent or 50 percent of the related penalties, depending on the transaction.

135 For a history of IRC section 6404(g), see Hale E. Sheppard, “From the Cradle to the Grave: What Remains of Interest Suspension Under Code Sec. 6404(g)?” (2008) 10:1 Journal of Tax Practice & Procedure 41-64. 136 See Hale E. Sheppard, “Applying Old Theories in New Contexts: Interest Suspension Upheld Where Government Fails To Prove Fraud” (2007) 9:3 Journal of Tax Practice & Procedure 15-18 and 47-48, at 18. 137 HR conf. rep. no. 105-599, 105th Cong., 2d sess. (1998), at 361 (www.congress.gov/105/crpt/ hrpt599/CRPT-105hrpt599.pdf ). 138 Sheppard, supra note 136, at 47. 139 To use Canadian terminology, three years is the “normal reassessment period” in the United States, after which the year becomes statute-barred. 140 Sheppard, supra note 135, at 41. 141 Internal Revenue Service, “IRS Settlement Initiative,” Fact Sheet FS-2005-17, October 17, 2005 (www.irs.gov/pub/irs-news/fs-05-17.pdf ). interest relief on income tax debts: canada versus the united states n 961

To induce taxpayers to participate in the settlement initiative, Congress enacted “off-Code” legislation (buried in the Gulf Opportunity Zone Act of 2005)142 that provided that taxpayers who elected to take part in the settlement initiative by Janu- ary 23, 2006 would be granted the interest relief contemplated by section 6404(g) through to October 3, 2004. The legislation also authorized the secretary, on an ongoing basis, to grant interest suspension through to October 3, 2004 to taxpayers who participated in listed and reportable transactions “reasonably and in good faith.” With the passage of time, however, as the number of active disputes involving taxation years prior to 2004 has decreased, these GOZA provisions have become less relevant. Other than in respect of procedural or jurisdictional issues, the case law on section 6404(g) is limited and has dealt primarily with the application of the “fraud” exception.143 There have also been a handful of reported cases in which the IRS has conceded the taxpayer’s entitlement to interest suspension under section 6404(g).144

IRC Section 6404(i): Federally Declared Disaster or Terroristic or Military Action IRC section 6404(i), which also traces its origins to the RRA 98,145 is today a cross- reference to IRC section 7508A,146 which authorizes the secretary to postpone deadlines for taxpayers affected by a “federally declared disaster”147 or a “terroristic or military action.”148 The secretary is authorized to “specify a period of up to 1 year that may be disregarded” in determining, inter alia, whether various listed acts (including the filing of returns and payment of amounts due) were performed on time, as well as “the amount of any interest . . . for periods after such date.” The

142 Pub. L. no. 109-135, at section 303 (herein referred to as “the GOZA”). 143 Sala v. US, 552 F Supp 2d 1157 (Dist. Ct. CO 2007); Bolton v. US, 2014 US Dist. LEXIS 158615 (W Dist. TN 2014); and In re Wyly, 552 BR 338 (Bankr. N Dist. TX 2016). For a case comment on Sala, see Sheppard, supra note 136. 144 Cristo v. CIR, TC Memo 2017-239, at 5-6, note 5; aff’d 775 F Appx. 303 (9th Cir. 2019), motion for en banc rehearing denied, cert. denied; and Welch v. CIR, TC Memo 2017-229, at 21, note 23. 145 The provision that is now IRC section 7508A was originally enacted as section 6404(h) as part of the RRA 98. OCCM PMTA 00249 (www.irs.gov/pub/lanoa/pmta00249_7039.pdf ). 146 A cross-reference in US legislation apparently has no legal effect, but is enacted for the convenience of the reader (R. Michael Teper, personal communication, June 30, 2020). 147 A “federally declared disaster” is defined as a disaster “determined by the President of the United States to warrant assistance by the Federal Government under the Robert T. Stafford Disaster Relief and Emergency Assistance Act”: IRC section 165(i)(5)(A). 148 A “terroristic or military action” is defined broadly so as to include “any terroristic activity which a preponderance of the evidence indicates was directed against the United States or any of its allies” or “any military action involving the Armed Forces of the United States and resulting from violence or aggression against the United States or any of its allies (or threat thereof )”: IRC section 692(c)(2). “Training exercises” are excluded from the definition. 962 n canadian tax journal / revue fiscale canadienne (2020) 68:4

IRS regularly provides relief under section 7508A and maintains a web page that lists federal announcements of disasters.149 A key limitation on section 7508A, set out in Treas. reg. section 301.7508A-1, is that it relieves interest only on tax debts that arise during the postponement period. For tax debts that may have arisen prior to the postponement period, interest continues to accrue at all times—even in the case of payments that are made pursuant to a negotiated payment plan and that come due during the postponement period.150 In Canada, analogous relief under subsection 220(3.1) of the ITA in the event of a natural disaster is regularly granted by the CRA,151 with a key difference being that the CRA does not categorically exclude the granting of relief for interest that accrues on tax debts arising prior to the disaster.

The OIC Regime The OIC regime fills many of the gaps left byIRC section 6404 and arguably represents a closer US counterpart to subsection 220(3.1) of the ITA.152 IRC section 7122 (whose origins date back to tax legislation in the Civil War era)153 authorizes the secretary and the attorney general to “compromise” any civil or criminal case arising under the internal revenue laws. The secretary is authorized to enter into written agreements with taxpayers over their liabilities for any taxable period. Such agreements are “final and conclusive” and may not be reopened by any officer, employee, or agent of the United States “except upon a showing of fraud or malfeasance, or misrepresentation of a material fact.”154 The IRS (and its predecessor entities) long construed its power to compromise tax debts very narrowly, such that it would exercise that power only where there was

149 Internal Revenue Service, “Tax Relief in Disaster Situations” (www.irs.gov/newsroom/tax -relief-in-disaster-situations). 150 Treas. reg. section 301.7508A-1(f ), examples 6 and 8. The IRM states that “[t]he disaster period interest suspension should only apply to tax returns having an original return due date that falls within the disaster period,” and “[t]ax years with pre-existing liabilities that are due before or after a disaster period do not qualify for disaster interest relief under IRC 7508A”: IRM, supra note 67, at section 20.2.7.13.1(1). 151 See Sherman, supra note 5, notes on subsection 220(3.1) for a list of CRA press releases and communications. 152 For a recent overview of the OIC regime from a Canadian perspective, including a comparison with subsection 220(3.1) of the ITA, see Colin Jackson, “The Case for a Canadian Offer-in- Compromise Program” (2015) 40:2 Queen’s Law Journal 643-78, at 648-56. 153 Shu-Yi Oei, “Getting More by Asking Less: Justifying and Reforming Tax Law’s Offer-in- Compromise Procedure” (2012) 160 University of Pennsylvania Law Review 1071-1137, at 1101; and I. Jay Katz, “An Offer in Compromise You Can’t Confuse: It’s Not the Opening Bid of a Delinquent Taxpayer To Play Let’s Make a Tax Deal with the Internal Revenue Service” (2012) 81:7 Mississippi Law Journal 1673-1755, at 1681. 154 IRC section 7121. interest relief on income tax debts: canada versus the united states n 963

“uncertainty of liability or a questionable prospect for recovery.”155 Hardship or other equitable grounds were not considered suitable grounds for a compromise.156 An attorney general’s opinion issued in 1933, in the midst of the Great Depression, held that compromise of an established liability for tax, interest, or penalties was never possible if the taxpayer had the means to pay the liability in full—even when doing so would drive the taxpayer into bankruptcy or entail the shutdown of his or her business.157 This opinion was regarded as definitive until Congress intervened with the enactment of the RRA 98, which mandated a major overhaul and expansion of the OIC regime.158 Among other things, the RRA 98 required the IRS to publish guidance about when compromise would be available, to suspend all collection action when an OIC was proffered by a taxpayer, and to set up mechanisms for administrative review of refusals to accept an OIC.159 The congressional record also documented Congress’s “expectation” that

n the IRS would expand the grounds for compromising tax-related liabilities so as “to take into account factors such as equity, hardship, and public policy where a compromise of an individual taxpayer’s income tax liability would promote effective tax administration”; n “the IRS may utilize this new authority to resolve longstanding cases by forgo- ing penalties and interest which have accumulated as a result of delay in determining the taxpayer’s liability”; and n “the IRS will adopt a liberal acceptance policy for offers-in-compromise to provide an incentive for taxpayers to continue to file tax returns and continue to pay their taxes.”160

The IRS, at least initially, took Congress’s directive seriously and followed up by enacting Treas. reg. section 301.7122-1.161 Departing from the attorney general’s

155 Daniel T. Maggs, “Section 7122 of the Internal Revenue Code: The Offer in Compromise” (1976) 11:2 Gonzaga Law Review 481-504, at 483. See also Katz, supra note 153, at 1679 and 1683-85; and Jackson, supra note 152, at 648. 156 Katz, supra note 153, at 1679, 1686-94, 1717, and 1725. 157 United States, Attorney General Opinion, October 24, 1933, 38 US Op Atty Gen 94, Compromise of Claims Under Sections 3469 and 3229 of the Revised Statutes (www.justice.gov/ archives/usam/civil-resource-manual-9-attorney-general-opinion-october-24-1933). See the discussion in Katz, supra note 153, at 1687-88 and 1725. 158 See Katz, supra note 153, at 1725 et seq.; and Oei, supra note 153, at 1103 et seq. 159 See, for example, IRC sections 7122(d) and (e). For a discussion of the RRA 98’s reforms to the OIC regime, see Katz, supra note 153, at 1679-80. 160 HR conf. rep. no. 105-599, supra note 137, at 288-89; and S rep. no. 105-174, 105th Cong., 2d sess. (1998), at 88-90 (www.congress.gov/105/crpt/srpt174/CRPT-105srpt174.pdf ). 161 On the history of Treas. reg. section 301.7122-1, see Ricky Thomas, “An Offer You Can’t Refuse: Acceptance Under the Final Code Sec. 7122 Regulations” (2004) 6:2 Journal of Tax Practice & Procedure 41-56; and Katz, supra note 153, at 1727-37. 964 n canadian tax journal / revue fiscale canadienne (2020) 68:4

1933 opinion, the regulation sets out three grounds on which the secretary may accept an OIC: the two traditional grounds of “doubt as to liability”162 and “doubt as to collectability,”163 and the new ground “to promote effective tax administration” (known as “ETA”).164 ETA encompasses two subcategories: (1) the prevention of economic hardship and (2) circumstances “where compelling public policy or equity considerations identified by the taxpayer provide a sufficient basis for compromising the liability.”165 The regulation further provides that when a taxpayer makes an OIC, the IRS must undertake an individualized review taking into account the totality of the circumstances.166 The regulation goes on to list factors that can militate for or against compromising a tax liability under ETA, and provides examples of circumstances in which compromise may be warranted.167 The IRM expands further on those factors and provides additional illustrations.168 Relief under the “economic hardship” prong of ETA is available only to individuals (that is, not to corporations, trusts, partnerships, or other entities).169 Hardship is defined as “caus[ing] an individual taxpayer to be unable to pay his or her reasonable basic living expenses”;170 it typically involves situations where a taxpayer’s financial condition has become precarious owing to such events as a medical calamity,

162 Treas. reg. section 301.7122.1(b)(1). The “doubt as to liability” prong generally applies when a taxpayer misses a deadline to appeal a disputed deficiency to the US Tax Court. An OIC based on doubt as to liability is essentially an opportunity for the taxpayer to convince the IRS that its assessing position is incorrect and to propose a settlement without having to pay the disputed amounts and initiate refund proceedings. See Oei, supra note 153, at 1080. 163 Treas. reg. sections 301.7122.1(b)(2) and (c)(2). For a discussion of the “doubt as to collectability” prong, see Jackson, supra note 152, at 650-53; and Oei, supra note 153, at 1078-79. 164 Treas. reg. sections 301.7122-1(b)(3) and (c)(3). 165 Treas. reg. section 301.7122-1(b)(3)(ii). 166 Treas. reg. section 301.7122-1(c)(1). 167 Treas. reg. section 301.7122-1(c)(3). 168 IRM, supra note 67, at sections 5.8.11.3.1 and 5.8.11.3.2.1(1). 169 Ibid., at section 5.8.11.3.1(2). This restriction has been upheld in the case law: Lindsay Manor Nursing Home, Inc. v. CIR, 148 TC 235 (2017), at 242-46; vacated on other grounds 725 F Appx. 713 (10th Cir. 2018). See also Pazzo Pazzo, Inc. v. CIR, TC Memo 2017-12, at 28. As originally proposed, Treas. reg. section 301.7122 contemplated non-individual taxpayers making hardship claims; however, those provisions were removed in the final version of the regulations issued in 2002 (Katz, supra note 153, at 1729-30). That said, corporations or other non-individual taxpayers in financial distresscan make OICs based on “doubt as to collectability,” which can take into account hardship considerations; see, for example, Mayer Investment Company v. CIR, TC Memo 2010-52, at 14-15. 170 Treas. reg. sections 301.7122-1(b)(3)(i) and (c)(3)(i), read with Treas. reg. section 301.6343- 1(b)(4)(i). interest relief on income tax debts: canada versus the united states n 965 a natural disaster, or the need to provide full-time support to a dependant with a long-term illness.171 Relief on the basis of the “public policy and equity” prong of ETA may be considered in a variety of situations—for example, where

n the taxpayer was incapacitated by a long, serious illness requiring extended periods of hospitalization that prevented compliance with tax laws; n the taxpayer relied on erroneous guidance from the IRS; n the IRS made processing errors that resulted in the taxpayer losing entitle- ments to credits or deductions; n the IRS unreasonably delayed resolution of a taxpayer’s case in circumstances where abatement under IRC section 6404 was not available; n the tax liability is the result of a criminal or fraudulent act of a third party, provided that the taxpayer took “reasonable precautions to prevent the criminal or fraudulent acts at issue”; or n failure to compromise on the tax liability “would have a significantly negative impact on the community in which the taxpayer lives or does business.”172

The US Tax Court has confirmed that the examples set out in the regulations and the IRM are not exhaustive.173 Relief on the basis of public policy and equity will generally not be granted where a taxpayer is caught by a “tax trap”—that is, where grossly disproportionate amounts of tax are imposed as a result of presumably unintended effects of tax legislation. The IRS’s view, which has been upheld in the courts, is that any relief in such situations must come from Congress.174 This is another point of contrast with Can- ada, where unintended consequences of legislation are recognized as grounds for obtaining a remission order, as discussed above.

171 Treas. reg. section 301.7122-1(c)(3)(iii); and IRM, supra note 67, at sections 5.8.11.3.1(5) through (7). 172 Treas. reg. section 301.7122-1(c)(3)(iv); and IRM, supra note 67, at section 5.8.11.3.2.1(1). 173 Bogart v. CIR, TC Memo 2014-46, at 11-12. 174 IRM, supra note 67, at section 5.8.11.3.2.1(7). An archetypal case on such situations is Speltz v. CIR, 124 TC 165 (2005); aff ’d 454 F 3d 728 (8th Cir. 2006), which concerned a taxpayer who faced unintended consequences of the IRC’s alternative minimum tax regime. Specifically, the IRC’s treatment of stock options granted by the taxpayer’s employer for shares that subsequently collapsed in value resulted in an income tax bill in 2000 far in excess of the taxpayer’s actual income for the year. The IRS refused to compromise on the tax liability on the basis that any remedy must come from Congress—a decision affirmed by the US Tax Court and the Eighth Circuit. For a discussion of this case, see Katz, supra note 153, at 1746-48. See also Wai v. CIR, TC Memo 2006-179 (dealing with a similar situation). Following these cases, the national taxpayer advocate, in her 2006 annual report, described the alternative minimum tax regime as “[t]he most serious problem facing taxpayers today” and recommended its abolition for individuals. National Taxpayer Advocate, National Taxpayer Advocate, 2006 Annual 966 n canadian tax journal / revue fiscale canadienne (2020) 68:4

The regulations further provide that compromise under either subcategory of ETA is not permitted “if compromise of the liability would undermine compliance by taxpayers with the tax laws.”175 Consequently, a taxpayer can obtain relief only if

n “[t]he taxpayer has remained in compliance since incurring the liability and overall their compliance history does not weigh against compromise”; n “[t]he taxpayer must have acted reasonably and responsibly in the situation giving rise to the liabilities”; and n “[t]he circumstances of the case must be such that the result of the compromise does not place the taxpayer in a better position than they would occupy had they timely and fully met their obligations, unless special circumstances justifying the compromise are present.”176

Taxpayers disqualify themselves from relief if they have “an overall history of non- compliance” with the IRC, they have “taken deliberate actions to avoid the payment of taxes,” or they have “encouraged others to refuse to comply with the tax laws.”177 The IRS will not compromise, on the basis of ETA, “tax liabilities associated with the taxpayer’s participation in abusive tax avoidance transactions.”178 Similarly, the IRS does not compromise tax liabilities or related interest in the context of its voluntary disclosure program, which only provides relief from criminal prosecution and, to a limited extent, penalties.179 In addition, as a matter of policy, the IRS will not settle tax disputes on the basis of “nuisance value,” which is defined as “any concession made solely to eliminate the inconvenience or cost of further negotiations or litigation and is unrelated to the merits of the issues.”180

Report to Congress, vol. 1 (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2006), at 3-5 (www.irs.gov/pub/tas/2006_arc_vol_1_cover__section_1.pdf ). For another example of a “tax trap” situation that the IRS declined to resolve through compromise, see West v. Sec’y of the United States Treasury, 2006 US Dist. LEXIS 47359 (Dist. Ct. NH 2006) (involving a mismatch between stock market losses and gains, resulting in a tax liability of more than $90,000 on non-existent income). 175 Treas. reg. sections 301.7122-1(b)(3)(iii) and (c)(3)(ii). 176 IRM, supra note 67, at section 5.8.11.3.2(4). 177 Ibid., at section 5.8.11.3.3(2). 178 Ibid., at section 5.8.11.3.2(4) (with regard to the “public policy and equity” prong). See also Gustashaw v. CIR, TC Memo 2018-215, at 17-18, where the US Tax Court held that participating in a tax shelter justified refusing an OIC on the basis of the “hardship” prong of ETA. 179 IRM, supra note 67, at section 9.5.11.9(3); and Internal Revenue Service, Deputy Commissioner for Services and Enforcement, “Updated Voluntary Disclosure Practice,” Memorandum for Division Commissioners Chief, Criminal Investigation, November 20, 2018, at 1 and 3-5. 180 IRM, supra note 67, at section 8.6.4.2.4(1). interest relief on income tax debts: canada versus the united states n 967

Numerous procedural requirements (mostly enacted in 2005) govern the submission of OICs.181 The taxpayer must apply in writing using IRS form 656, provide all information requested by the IRS, sign under penalty of perjury,182 include a partial payment equal to 20 percent of the amount owing (or, if an instalment plan is proposed, the first payment),183 and pay an application fee.184 The IRS has, administratively, set out additional prerequisites to its consideration of a taxpayer’s OIC, including in particular that the taxpayer’s filings be up to date and that the taxpayer be in compliance with current reporting obligations and instalment payments.185 The courts have repeatedly held that failure by a taxpayer to comply with these various procedural requirements—whether statutory or administrative—justifies a rejection of, or even a refusal to process, an OIC by the IRS.186 Any remedy sought in an OIC must be tailored to the particular circumstances, which in some cases will result in cancellation of interest. For example, in cases of relief justified byIRS error or delay, the relief sought should put the taxpayer in the position that he, she, or it would have been in but for the error or delay.187 In cases of relief justified by errors or omissions caused by a third party, the IRS will generally agree to compromise interest and penalties, but will insist on payment of the underlying tax liability.188 If an OIC is accepted by the IRS, interest is suspended as of the date of acceptance “provided the taxpayer follows the agreement made with IRS, such as paying the agreed amount or making monthly payments.”189 As discussed above, the CRA has an analogous, although more limited, policy with regard to negotiated payment

181 Tax Increase Prevention and Reconciliation Act of 2005, Pub. L. no. 109-222. See Katz, supra note 153, at 1680 and 1753-54; and Oei, supra note 153, at 1107-9. 182 Treas. reg. section 301.7122-1(d)(1). 183 IRC section 7122(c)(1). 184 Treas. reg. section 300.3. The requirement to pay an application fee was originally enacted in 2003. Low-income taxpayers are exempt from the requirement. Oei, supra note 153, at 1107. 185 IRS form 656-B, “Offer in Compromise Booklet,” March 2009, at 3 (www.publiccounsel.org/ tools/assets/files/f656b-OIC-booklet.pdf ). 186 See, for example, Laurins v. CIR, 889 F 2d 910 (9th Cir. 1989), at 912; Olsen v. US, 414 F 3d 144 (1st Cir. 2005), at paragraphs 25-34; Living Care Alternatives of Utica, Inc. v. US, 411 F 3d 621 (6th Cir. 2005), at 630; Christopher Cross, Inc. v. US, 461 F 3d 610 (5th Cir. 2006), at 613; Rodriguez v. CIR, TC Memo 2003-153; Aaron v. CIR, TC Memo 2006-63, at 711-13; Kindred v. CIR, 454 F 3d 688 (7th Cir. 2006), at 697; Reed v. CIR, 141 TC 248 (2013), at 16-17, reconsideration denied TC Memo 2014-41; Bergdale v. CIR, TC Memo 2014-152, at 9-13; Crosswhite v. CIR, TC Memo 2014-179, at 24-25; Cunningham v. CIR, TC Memo 2014-200, at 18-19; Abu-Dayeh v. CIR, TC Memo 2015-136, at 9-11; Pazzo Pazzo, supra note 169, at 24-25 and 27; and Gentile v. CIR, TC Memo 2013-175, at 9; aff ’d 592 F Appx. 824 (7th Cir. 2014). 187 IRM, supra note 67, at section 5.8.11.5.3(2)(a). 188 Ibid., at section 5.8.11.5.3(2)(c). 189 Ibid., at section 20.2.11.11(3). 968 n canadian tax journal / revue fiscale canadienne (2020) 68:4 plans. In addition, and unlike the CRA, the IRS makes acceptance of an OIC conditional on the taxpayer’s remaining compliant with his, her, or its filing obligations for five years, failing which the entire liability, plus penalties and interest, may be reinstated.190 Research from the national taxpayer advocate demonstrates that settling cases with taxpayers through OICs tends to improve both the amount of revenue eventually collected from those taxpayers and their long-term compliance.191 In the years immediately following the enactment of the RRA 98, the IRS enthusi- astically encouraged taxpayers to submit OICs and reassigned staff to process them. A flood of offers followed, which in turn gave rise to a flood of administrative- ap peals and the accumulation of a large processing backlog.192 Eventually, owing in part to the “stock market collapse, the tax shelter and promoter scandals, and the need to fund a war on terrorism,” the IRS shifted resources back to enforcement and adopted a restrictive and formalistic approach to processing OICs in the early 2000s.193 Defying Congress’s instructions, the IRS refused to settle longstanding tax shelter cases by compromising on interest and penalties—a refusal blessed by the courts on the basis that Congress’s guidance was not binding.194 By 2004, the IRS was compromising virtually no cases on the basis of public policy and equity.195 The courts generally endorsed this parsimonious approach to OICs, resulting in a further hardening of the IRS’s attitude.196 However, a litany of complaints and damning reports from the national taxpayer advocate,197 the US Government Accountability

190 Form 656-B, supra note 185, at 18; “A Study of the IRS Offer in Compromise Program,” in National Taxpayer Advocate, National Taxpayer Advocate, 2017 Annual Report to Congress, vol. 2 (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2017), 42-60, at 42-43 and 46 (https://taxpayeradvocate.irs.gov/reports/2017-annual-report-to-congress/ full-report); and Keith Fogg, “Failing To Keep Current After Obtaining an Offer in Compromise,” Procedurally Taxing blog, August 7, 2020 (https://procedurallytaxing.com/ failing-to-keep-current-after-obtaining-an-offer-in-compromise). For cases in which the failure of a taxpayer to remain in compliance with post-OIC filing and payment obligations resulted in the voiding of a settlement and the reinstatement of a past tax liability, see Robinette v. CIR, 439 F 3d 455 (8th Cir. 2006); and Sadjadi v. CIR, no. 19-60633 (5th Cir. 2020). 191 “A Study of the IRS Offer in Compromise Program,” supra note 190, at 44, 47-48, and 53-59. 192 Wayne R. Johnson, “Is the Federal Offer-in-Compromise Program Still Viable” (2005) 7:6 Journal of Tax Practice & Procedure 47-62, at 48; and Oei, supra note 153, at 1109-12. 193 Johnson, supra note 192, at 48 and 52. See also Katz, supra note 153, at 1729 et seq.; and Oei, supra note 153, at 1112-16. 194 See Katz, supra note 153, at 1739-43; Fargo v. CIR, 447 F 3d 706 (9th Cir. 2006), at 711-12; and Keller v. CIR, 568 F 3d 710, at 720 (9th Cir. 2009). 195 National Taxpayer Advocate, National Taxpayer Advocate, 2004 Annual Report to Congress (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2004), at 328 (www.irs.gov/advocate/national-taxpayer-advocates-2004-annual-report-to-congress); and Johnson, supra note 192, at 48. 196 Katz, supra note 153, at 1737-52. 197 National Taxpayer Advocate, 2004 Annual Report to Congress, supra note 195, at 313-41, 433-34, and 450; National Taxpayer Advocate, National Taxpayer Advocate, 2007 Annual Report to interest relief on income tax debts: canada versus the united states n 969

Office,198 academics,199 and practitioners,200 as well as two interventions by Congress (the enactment of the GOZA, discussed above, followed by a legislated directive ordering the IRS to “give special consideration to an offer-in-compromise from a taxpayer who has been the victim of fraud by a third party payroll tax preparer”),201 have helped to prompt a change in attitude by the IRS. As a result, acceptance rates have significantly risen since 2012 and now stand at around 35 to 40 percent.202 Nevertheless, the national taxpayer advocate continues to express the view that the OIC program is underused and that the IRS needs to show greater flexibility in processing applications.203 Unlike in Canada, if a taxpayer makes an OIC and it is refused, an administrative appeal is available.204 Moreover, as discussed below, judicial review of a refusal of an OIC is available at the US Tax Court in the context of “collection due process” (CDP) proceedings.

Congress, vol. 1 (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2007), section 1, at 374-87 (www.irs.gov/advocate/national-taxpayer-advocates-2007-annual -report-to-congress); National Taxpayer Advocate, National Taxpayer Advocate, 2009 Annual Report to Congress, vol. 1 (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2009), section 1, at 196-216 (www.irs.gov/advocate/national-taxpayer-advocates-2009 -annual-report-to-congress); National Taxpayer Advocate, National Taxpayer Advocate, 2010 Annual Report to Congress, vol. 1 (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2010), section 1, at 311-18 (www.irs.gov/pub/tas/2010arcvol_1_status_update.pdf ); and National Taxpayer Advocate, National Taxpayer Advocate, 2013 Annual Report to Congress, vol. 1 (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2013), at 206-17 and 219-24 (https://www.taxpayeradvocate.irs.gov/reports/2013-annual-report-to-congress/ full-report/). 198 United States, General Accounting Office,Tax Administration: IRS Should Evaluate the Changes to Its Offer in Compromise Program, Report to the Chairman and Ranking Minority Member, Committee on Finance, U.S. Senate, GAO document no. GAO-02-311 (Washington, DC: GAO, March 2002) (www.gao.gov/products/GAO-02-311); and United States, General Accounting Office,IRS Offers in Compromise: Performance Has Been Mixed; Better Management Information and Simplification Could Improve the Program, Report to the Committee on Finance, U.S. Senate, GAO document no. GAO-06-525 (Washington, DC: GAO, April 2006) (www.gao.gov/ products/GAO-06-525). 199 See Katz, supra note 153, at 1730 et seq.; and Oei, supra note 153, passim. 200 See Johnson, supra note 192, passim. 201 Consolidated Appropriations Act, 2014, Pub. L. no. 113-76, division E, title I, section 106. See also supra note 142 and the accompanying text. 202 National Taxpayer Advocate, 2017 Annual Report to Congress, supra note 190, at 49; and National Taxpayer Advocate, National Taxpayer Advocate, 2019 Annual Report to Congress (Washington, DC: Internal Revenue Service, Taxpayer Advocate Service, 2019), at 307 (https://taxpayeradvocate.irs.gov/reports/2019-annual-report-to-congress/full-report). 203 National Taxpayer Advocate, 2019 Annual Report to Congress, supra note 202, at 89-96. 204 IRC section 7122(e); Treas. reg. section 301.7122-1(f)(5). 970 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Other Relieving Provisions The IRC contains a number of other interest-relieving provisions that reduce or obviate the need for discretionary interest abatement, some with counterparts in Canada, others without. IRC section 6621(d), known as the “interest-netting” provision, provides that during periods when a taxpayer and the secretary owe interest to each other on mu- tually offsetting overpayments and underpayments, “the net rate of interest under this section on such amounts shall be zero for such period.”205 This provision aims to reduce the imbalances caused by the application of different interest rates on underpayments and overpayments,206 and by the different treatment of interest paid to the IRS, which is non-deductible, and interest paid by the IRS, which is includible in income.207 Section 6621(d) was enacted as part of the RRA 1998, largely in response to concerns voiced by Congress that taxpayers were being systematically assessed interest when they owed no tax.208 While simple enough in concept, interest netting is administratively challenging for the IRS.209 Because IRS computers “cannot readily identify situations in which netting would be available,”210 taxpayers who wish to benefit from interest netting generally have to ask theIRS to make the necessary calculations and adjustments.211 As discussed above, the ITA contains a provision vaguely analogous to section 6621(d)—section 161.1—which was adopted around the same time; however, the Canadian provision applies only to corporations, and it allows for netting to take place only by written application from the

205 For an extensive overview of how the IRS implements the interest-netting regime, see IRM, supra note 67, at section 20.2.14. 206 In the United States, this differential applies only to corporations. Since January 1, 1999, there has been no interest differential for overpayments and underpayments for non-corporate taxpayers. IRC section 6621(a); and IRM, supra note 67, at section 20.2.14.1(6). 207 For discussion, see HR conf. rep. no. 105-599, supra note 137, at 256-57 and 360. 208 For a review of the history of IRC section 6621(d), including a summary of congressional deliberations leading to its adoption, see Wells Fargo & Co. v. US, 827 F 3d 1026 (F Cir. 2016), at 1035 et seq. See also Bob Probasco, “A Question of Identity—Interest Netting, Part 1,” Procedurally Taxing blog, January 8, 2019 (https://procedurallytaxing.com/a-question-of -identity-interest-netting-part-1). 209 As is evidenced by, inter alia, the sheer length of the IRM section on interest netting: IRM, supra note 67, at section 20.2.14. 210 Probasco, supra note 208, under the heading “Why Interest Netting?” 211 Unhelpfully, the IRS has not created a form for this purpose. Rather, taxpayers are expected to complete and submit IRS form 843, which is normally used for interest abatement requests, and to write at the top, “Request for Net Interest Rate of Zero under Rev. Proc. 2000-26.” See Internal Revenue Service, “Instructions for Form 843 (12/2019), Claim for Refund and Request for Abatement” (www.irs.gov/instructions/i843), under the heading “How To Request a Net Interest Rate of Zero”; see IRM, supra note 67, at section 20.2.14.6.3(1)(a). interest relief on income tax debts: canada versus the united states n 971 taxpayer within a 90-day window of time, the starting point of which is not always obvious.212 IRC section 6601(d), the US counterpart to subsection 161(7) of the ITA, deals with loss and credit carrybacks. Generally, it provides that when a tax liability is offset by a carryback from a subsequent year, interest runs on the liability from the year in which it arises until the filing date for the year in which the offsetting carry back arises.213 In other words, in contrast to the ITA, section 6601(d) does not contemplate imposing interest on tax debts during the period when they are completely offset. IRC section 6601(e)(3), in a manner similar to section 161.2 of the ITA, provides a grace period for payment when the secretary notifies a taxpayer that a specified amount of tax is due.214 As in Canada, interest does not accrue during the grace period if the amount is paid in full. The IRC also includes the following miscellaneous interest-relieving provisions, which have no Canadian equivalent:

n IRC section 692, which traces its origins back to the Second World War,215 provides for the forgiveness of unpaid taxes and interest for members of the armed forces killed or mortally wounded in combat, as an expression of “Congressional gratitude.”216 It has been expanded over the years217 and today provides various degrees of tax relief to military and civilian employees of the armed forces who die in non-combat operations, astronauts who die in the line of duty, and victims of specified acts of terror, namely, the 1995 Oklahoma bombing, the September 11, 2001 attacks, and the 2001 anthrax attacks. n IRC section 6151(b)(1) applies when a taxpayer elects under IRC section 6014 to have the secretary calculate how much tax is due in a particular year. Section 6014 is a measure that applies only to lower-income taxpayers (that is, those with a gross income below $10,000) whose income comes almost exclusively from wages. When such a taxpayer so elects, interest that would normally

212 See supra note 33 and the accompanying text. One might speculate that Parliament elected to give section 161.1 of the ITA more limited scope than IRC section 6621(d) in light of the significant operational challenges in implementing comprehensive interest netting. 213 This rule is reiterated in the accompanying Treas. reg. section 301.6601-1(e). 214 The grace period is 21 calendar days for amounts under $100,000 or 10 business days for higher amounts. According to the IRM, supra note 67, at section 20.2.7.10(6), this delay is extended to 45 days in the case of US taxpayers living overseas or in Puerto Rico. The statutory basis for the extension is not clear. There is an analogous provision for interest on penalties in IRC section 6601(2)(A). 215 For a discussion of the history of the provision, see Lupia’s Estate et al. v. Marcelle, 214 F 2d 942 (2d Cir. 1954), at 944-45. 216 Cane, supra note 69, at 848. 217 Lupia’s Estate, supra note 215, at 945: “[T]he march of this remedial legislation has ever been forward.” 972 n canadian tax journal / revue fiscale canadienne (2020) 68:4

accrue from the due date of the return is effectively suspended until the IRS determines the amount of tax that must be paid and sends a notice and demand to the taxpayer. n IRC section 6601(c) “suspends” the running of interest on tax deficiencies if the taxpayer waives, pursuant to IRC section 6213(d), the statutory bar on assessment and collection of a deficiency during the appeal period before the US Tax Court.218 If the taxpayer executes a section 6213(d) waiver and the secretary fails to make a notice and demand for payment within 30 days, interest is suspended on the deficiency until the notice and demand for payment is issued. The provisions found today in sections 6213(d) and 6601(c) both date back to 1926, when they were enacted to provide a vehicle for taxpayers to pay amounts under dispute before the US Tax Court in deficiency proceedings, and to stop the interest clock.219 Today, section 6213(d) waivers are typically given in the context of settlement discussions and include a waiver of the right to appeal to the US Tax Court.220 In such cases, section 6601(c) serves “to encourage the [IRS] to promptly process consents to assess” tax in accordance with the agreement with the taxpayer.221 n IRC section 7508 provides that the passage of time for a wide range of purposes is “disregarded” during times when a taxpayer is serving in the armed forces or in a related supporting role in a combat zone (as well as, if wounded while serving in the combat zone, for a period of “qualified hospitalization” thereafter). Unlike IRC section 7508A, section 7508 automatically suspends the accrual of interest on any unpaid tax liabilities that arise prior to deployment to a combat zone.222 In addition, members of the armed forces not serving in combat zones may qualify for a suspension of interest on outstanding tax liabilities, or a reduction in the interest rate, under section 3937 or section 4000 of the Servicemembers’ Civil Relief Act.223

Finally, it bears noting that, as in Canada, the US Bankruptcy Code provides for suspension of interest on tax liabilities following a bankruptcy petition, subject to some significant exceptions, including in particular tax liabilities secured by liens.224

218 This waiver is contemplated in IRC section 6213(d). The equivalent in Canada would be for a taxpayer to “waive” the collection restrictions found in section 225.1 of the ITA—a procedure that is not contemplated. 219 For a review of the history and purpose of the provisions, see US v. Price, 361 US 304 (1960), at 307-13. See also Ulrich v. CIR, 585 F 3d 1235 (9th Cir. 2009). 220 US v. Evseroff, 2002 US Dist. LEXIS 15703 (E Dist. NY 2002), at 16. 221 OCCM PMTA 2009-003, January 6, 2009, at 3 (www.irs.gov/pub/lanoa/pmta2009-003.pdf ); and OCCM PMTA 2012-09, March 29, 2012, at 1 (www.irs.gov/pub/lanoa/pmta_2012-09.pdf ). 222 IRM, supra note 67, at section 20.2.7.11(1). 223 50 US Code, sections 3937 and 4000. These programs are discussed in the IRM, supra note 67, at section 20.2.7.12. 224 IRM, supra note 67, at section 20.2.11.6.1. interest relief on income tax debts: canada versus the united states n 973

JUDICIAL RECOURSES AGAINST A REFUSAL TO SUSPEND OR ABATE INTEREST Key Differences in Income Tax Administration To facilitate an apples-to-apples comparison between the United States and Canada with regard to judicial recourses against a refusal by the revenue authority to waive, cancel, suspend, abate, or disregard interest, it is useful to review a few key differences in the income tax administration systems and the terminology employed in the two countries. In Canada, a “notice of assessment” issued by the minister (1) fixes a taxpayer’s liability for tax in a particular taxation year in a manner that is “valid and binding”; (2) entitles the taxpayer to access dispute resolution processes; and (3) entitles the minister to initiate collection action, subject to section 225.1, which generally suspends collection until the dispute resolution process has run its course.225 The Tax Court of Canada has exclusive jurisdiction—which Canadian courts have jealously defended226—to hear appeals of assessments, and in the context of such an appeal, the Tax Court has jurisdiction to determine the correctness of the amounts of tax, interest, and penalties imposed under the ITA. The Tax Court does not, however, have jurisdiction to review the minister’s discretionary acts,227 including decisions to waive or cancel interest. As discussed above, such matters fall within the exclusive purview of the Federal Court. In the United States, unlike in Canada, an assessment immediately opens the door to collection action. Taxpayers are generally obligated to pay taxes once assessed, and if they believe that there is an error in the assessment, their remedy is to sue the United States for a refund of any overpayment in either the US district court or the United States Court of Federal Claims.228 To attenuate the potential hardship and unfairness to taxpayers of having to pay all disputed taxes in order to dispute them, the IRC provides for an intermediate step—for which there is no equivalent in Canada—applicable to income, estate, gift, and certain excise taxes. If the IRS disputes the income reported on the taxpayer’s return (other than amounts attributable to “mathematical or clerical” error),229 then, prior to issuing an assessment, the IRS must issue a “notice of deficiency” setting out

225 ITA subsections 152(8) and 165(1), and section 222 et seq. 226 See Guy Du Pont and Michael H. Lubetsky, “The Power To Audit Is the Power To Destroy: Judicial Supervision of the Exercise of Audit Powers” (2013) 61, Supplement Canadian Tax Journal S103-21, at S108-11. 227 For a discussion of this rule and whether ITA subsection 247(10) is an exception to it, see Daniel Sandler and Lisa Watzinger, “Disputing Denied Downward Transfer-Pricing Adjustments” (2019) 67:2 Canadian Tax Journal 281-308, at 291 et seq. 228 IRC section 1346(a)(1). 229 IRC section 6211. 974 n canadian tax journal / revue fiscale canadienne (2020) 68:4 the disputed amounts. A taxpayer may “petition” the US Tax Court for a “redeter mination of the deficiency,”230 a grant of power known as the US Tax Court’s “deficiency jurisdiction.” Only when theUS Tax Court’s decision becomes final can the IRS issue an “assessment” for the deficiency.231 “Deficiency” is defined as referring only 232 totax. Consequently, the US Tax Court’s jurisdiction to review deficiencies does not extend to any related interest calculations.233 If a taxpayer believes that interest has been incorrectly calculated in an assessment, the taxpayer must pay the disputed interest and file a refund action in the district court or the Court of Federal Claims. A limited exception to this rule appears in IRC section 7481(c), which provides that after the US Tax Court has dis- posed of a deficiency claim and the secretary has assessed the taxpayer accordingly, the taxpayer may, within one year, petition the US Tax Court to review the interest calculation. However, the taxpayer must pay all disputed interest as a condition precedent to seeking this review.234 The IRC confers jurisdiction on the US Tax Court to hear a variety of other types of tax-related disputes, such as those involving spousal relief from joint and several liability,235 various declaratory judgments,236 partnership items,237 worker classification,238 awards to whistleblowers,239 IRS passport certifications,240 and so forth. Two grants of jurisdiction relevant to discretionary interest relief are those found in IRC section 6404(h) (review of decisions to refuse abatement) and sections 6320 and 6330 (CDP hearings), both discussed below. Finally, a point sometimes overlooked when discussing the US tax dispute resolution system is that the US bankruptcy courts have broad jurisdiction under the Bankruptcy Code to

determine the amount or legality of any tax, any fine or penalty relating to a tax, or any addition to tax, whether or not previously assessed, whether or not paid, and whether or not contested before and adjudicated by a judicial or administrative tribunal of competent jurisdiction.241

230 IRC section 6213. 231 IRC section 6213(a). 232 IRC section 6211. 233 For examples of taxpayers who unsuccessfully attempted to litigate interest abatement in the context of a deficiency proceeding, seeBourekis v. CIR, 100 TC 20 (1998), at 8-12; and Wienke v. CIR, TC Memo 2020-143, at 21-22, note 8. 234 IRC section 7481(c)(3). For discussion, see Frantz v. CIR, 4 F 3d 990 (5th Cir. 1993). 235 IRC section 6015. 236 IRC sections 7476 through 7479. 237 IRC section 6234. 238 IRC section 7436. 239 IRC section 7623. 240 IRC section 7345. 241 11 US Code, section 505. interest relief on income tax debts: canada versus the united states n 975

This includes the power to determine “the amount or legality” of an interest assessment, although it does not include the power to grant discretionary relief of interest.242

Mandatory Interest Relief When the IRS improperly assesses interest in situations where the IRC requires that the interest be abated, offset, suspended, or disregarded under the various provisions discussed in the second section of this article, a taxpayer’s default recourse is to pay the disputed interest and institute a refund action in the district court or the Court of Federal Claims.243 Similarly, if the United States institutes civil proceedings to collect the interest, a taxpayer may generally raise a mandatory interest relief provision as a ground of defence.244 In limited situations, the US Tax Court also has jurisdiction to hear disputes over the failure of the IRS to properly relieve, suspend, or offset interest in accordance with the IRC’s mandatory interest relief provisions, including

n when interest has been assessed to implement a decision of the US Tax Court following a deficiency proceeding (discussed above); n in the context of CDP hearings (discussed below); or n if a taxpayer has requested abatement of interest under IRC section 6404 and the IRS refuses (also discussed below).

In addition, although the jurisdictional basis is unclear, there have been a number of cases where the US Tax Court, in the course of reviewing a refusal of interest abatement under section 6404, has found that a taxpayer was overcharged interest under IRC section 6601(c) and has ordered relief accordingly.245

242 In re Matter of Carlson v. US, 126 F 3d 915 (7th Cir. 1997), at 920; Hromadka v. US, 1999 Bankr LEXIS 558 (W Dist. TX 1999), at 4-5; and In re Vendell Healthcare, Inc., 222 BR 564 (Bankr M Dist. TN 1998), at 567-71. 243 For example, see Converse, supra note 125 (IRC section 6404(e)(2)); Lindstedt, supra note 125 (IRC section 6404(e)(2)); Putzel, supra note 126 (IRC section 6404(e)(2)); Sala, supra note 143 (IRC section 6404(g)); Farrell v. US, 484 F Supp 1097 (E Dist. AR 1980) (IRC section 6601(e)(3)); and Rockefeller v. US, 572 F Supp 9 (E Dist. AR 1982) (IRC section 6601(e)(3)). Similar refund claims that were adjudicated on their merits against the taxpayer include Bolton v. US, 114 AFTR 2d (RIA) 2014-6407 (W Dist. TN) (IRC section 6404(g)); and Kalantari v. US, Case no. EDCV 14-02580-VAP (SPx) (C Dist. CA 2015) (IRC section 6404(g)). 244 See US v. Sullivan, 2007 US Dist. LEXIS 26153 (W Dist. OK 2007) (the taxpayer unsuccessfully argued that IRC section 6404(e)(2) applied to a refund that it had received in error). See also Larson v. US, 2011 US Dist. LEXIS 74319 (N Dist. CA 2011) (adjudication of a claim under IRC section 6404(g) in the context of judicial review of a jeopardy assessment); and Wyly, supra note 143 (adjudication of a claim under IRC section 6404(g) in the context of a bankruptcy proceeding). 245 Hancock, supra note 117, at 2 (IRC section 6601(c)) (on consent); and Corson, supra note 73, at 16 (IRC section 6601(c)). Similarly, in Hull, supra note 94, at 10-13, the US Tax Court decided on the merits that IRC section 6601(c) did not apply to the taxpayer’s situation. 976 n canadian tax journal / revue fiscale canadienne (2020) 68:4

It bears noting that a number of US district court decisions have construed the Hinck decision from the US Supreme Court,246 which held that the US Tax Court’s jurisdiction to review refusals to abate interest under IRC section 6404(e)(1) is exclusive, as meaning that the US Tax Court’s jurisdiction to review refusals to suspend or abate interest under IRC sections 6404(e)(2) and 6404(g) also is exclusive.247 This outcome—which is not consistent with the reasoning of Hinck and is presumably not what the court intended248—has left ongoing confusion and uncertainty as to whether taxpayers can continue to rely on sections 6404(e)(2) and 6404(g) in refund actions or in civil actions brought by the IRS to collect tax debts.

IRC Section 6404(h) IRC section 6404(h) reads in part as follows:

(H) JUDICIAL REVIEW OF REQUEST FOR ABATEMENT OF INTEREST (1) IN GENERAL The Tax Court shall have jurisdiction over any action brought by a taxpayer who meets the requirements referred to in section 7430(c)(4)(A)(ii) to determine whether the Secretary’s failure to abate interest under this section was an abuse of discretion, and may order an abatement, if such action is brought— (A) at any time after the earlier of— (i) the date of the mailing of the Secretary’s final determination not to abate such interest, or (ii) the date which is 180 days after the date of the filing with the Secretary (in such form as the Secretary may prescribe) of a claim for abatement under this section, and (B) not later than the date which is 180 days after the date described in subpara graph (A)(i).

This provision was first added as part of the Taxpayer Bill of Rights 2 in 1996,249 largely in response to a series of decisions that held that no judicial recourse of any

246 Hinck v. US, 550 US 501 (2007). 247 US v. Bates, 2015-2 US TC (CCH) P50, 575 (M Dist. FL 1995); and Michael D. v. IRS, 2009 US Dist. LEXIS 85948 (Dist. Ct. OR 2009). 248 A key aspect of the reasoning in Hinck was that Congress enacted the provision now found in IRC section 6404(h) in response to case law that held that interest abatement decisions under section 6404(e)(1) were not subject to any form of judicial review. Conferring jurisdiction on the US Tax Court to review refusals was thus remedial; the aim was to create a remedy for taxpayers where none had previously existed, not to curtail the existing jurisdiction of the US district courts or the Court of Federal Claims. 249 Supra note 91. The provision was originally enacted as IRC section 6404(g), but was subsequently redesignated as section 6404(i) and then section 6404(h). See McGaughy, supra note 94, at 9-10, note 5; and Foote, supra note 106 (USTC), at 15-16, note 10. interest relief on income tax debts: canada versus the united states n 977 kind was available against refusals of interest abatement under IRC section 6404(e)(1).250 As the provision was originally enacted, a taxpayer could petition the US Tax Court only after receiving a final decision from theIRS on an interest abatement.251 Subparagraph (A)(ii), which allows a taxpayer to petition the US Tax Court after 180 days have passed following the date of application, was added in 2016.252 Taxpayers can avail themselves of this right to petition the US Tax Court only if they meet “the requirements referred to in section 7430(c)(4)(A)(ii),” which is a net worth requirement. Essentially, the right to petition the US Tax Court to review a decision not to abate interest is limited to individuals whose net worth does not exceed $2 million and owners of businesses or corporations with no more than 500 employees, whose net worth does not exceed $7 million.253 This net worth requirement (which also applies in other contexts) has been the object of much criticism and litigation in the United States, both for its inherent unfairness and for its complexity, but its constitutionality has been recognized.254 As specified in section 6404(h), the standard of review applicable to IRS decisions not to abate interest is “abuse of discretion.” This means that “[i]n order to prevail, a taxpayer must demonstrate that . . . the Secretary exercised his discretion arbi- trarily, capriciously, or without sound basis in fact or law.”255 It is a highly deferential

250 Hinck, supra note 246, at 503-4. These decisions included Clinton Street Corp. v. CIR, 89 TC 352 (1987), at 354-57; Selman v. US, 941 F 2d 1060, at 1064 (10th Cir. 1991); Horton Homes, Inc. v. US, 936 F 2d 548, at 554 (11th Cir. 1991); Bax, supra note 73, at 58; Argabright v. US, 35 F 3d 472, at 476 (9th Cir. 1994); and Carlson v. US ( In re Carlson), 126 F 3d 915, at 920 (7th Cir. 1997). 251 See Ho v. CIR, TC Memo 1998-363; Gilmer v. CIR, TC Memo 2009-296; and Ward v. CIR, TC Memo 2007-374, at 9 (“[F]ailure to act on a request [for abatement] within a reasonable time does not constitute a final determination for section 6404(h) purposes”). 252 Consolidated Appropriations Act, 2016, Pub. L. no. 114-113, division Q, section 421(a). 253 For an example of a case where the taxpayers’ petition was rejected because of the net worth requirement, see Vercel v. CIR, TC Memo 2014-20. 254 See, for example, Estate of Kunze v. CIR, 233 F 3d 948, at 954 (7th Cir.); Sanford J. Boxerman, “You Can Ask the IRS To Abate Interest on a Tax Liability—And If You’re Not Too Rich, You Can Obtain Judicial Review, Too” (2014) 17:4 Journal of Passthrough Entities 65-68, at 66; and Gamino, supra note 92, at 45-47. 255 See Lee, supra note 105, at 149; Pettyjohn, supra note 94, at 16; Berry, supra note 105, at 10; Hawksley, supra note 94, at 18; Bucaro, supra note 94, at 18; Nelson, supra note 108, at 7-8; Jean, supra note 99, at 7; Landvogt, supra note 118, at 9; Guerrero, supra note 94, at 8; Wright, supra note 94 (5th Cir.), at 549; Mekulsia, supra note 94 (USTC), at 22; Beagles, supra note 105, at 6-7; Matthews, supra note 94, at 13-14; Allcorn, supra note 93, at 57; Swanson, supra note 105, at 11; Paneque, supra note 115, at 18-19; McGaughy, supra note 94, at 9-10; Goettee, supra note 73 (4th Cir.), at 218; Howell, supra note 105, at 8; Corson, supra note 73, at 8-9; Telesmanich, supra note 108, at 4-5; Hull, supra note 94, at 16; Foote, supra note 106 (USTC), at 16; Hornbacker, supra note 106, at 17; Prakash, supra note 117, at 9; Santana, supra note 117, at 18-19; Mathia, supra note 99, at 33; and Sandberg, supra note 110, at 12-13. 978 n canadian tax journal / revue fiscale canadienne (2020) 68:4 standard256 that limits judicial intervention to situations where a decision is based “on an erroneous view of the law or a clearly erroneous assessment of the facts.”257 There does not yet seem to be any case law on how the “abuse of discretion” standard applies when a taxpayer petitions under section 6404(h) before the IRS makes a decision on an interest abatement application, in reliance on the recently enacted section 6404(h)(1)(A)(ii). The overwhelming majority of petitions decided pursuant to section 6404(h) have involved reviews of refusals of abatement based on IRC section 6404(e)(1). In such cases, the US Tax Court has repeatedly endorsed a three-part test that must be satisfied for the court to grant relief.258 That test requires

1. an error or delay by the IRS in performing an ministerial or managerial act; 2. a correlation between that error or delay and a specific period of delay in payment; and 3. that the taxpayer would have paid the tax liability but for the IRS’s error or delay.

An ironic (if not Kafkaesque) aspect of this test is that, to satisfy the third prong, a taxpayer must show that he, she, or it had sufficient financial resources available at the relevant time to pay the tax liability at issue. Frequently, if the taxpayer was impecunious during the period of IRS delay, the US Tax Court will not find it abusive for the IRS to refuse to abate interest.259 Moreover, the US Tax Court has repeatedly observed that “[t]he mere passage of time does not establish error or delay” on the part of the IRS,260 and has held that

256 See, for example, Yeomans v. CIR, TC Memo 2009-216, at 19: “If this was a de novo proceeding, and had the statute not specified an abuse of discretion standard, this Court might have been more generous, but by law that is not the standard of review to be employed here.” 257 Fargo, supra note 194, at 709. 258 The test seems to have originated in CIR v. Braun, TC 2005-221, at 12-15, and has been applied in Hancock, supra note 117, at 9; Paneque, supra note 115, at 18; Hull, supra note 94, at 15-16; Foote, supra note 106 (USTC), at 15; Hornbacker, supra note 106, at 16; Prakash, supra note 117, at 9; Santana, supra note 117, at 18; Krehnbrink v. CIR, TC Memo 2019-56, at 14-15 (“Krehnbrink #2”); Sher v. CIR, TC Memo 2009-86; aff’d 381 F Appx. 62 (2d Cir. 2010); and Coleman v. CIR, TC Memo 2012-116, at 11-12; aff’d 594 F Appx. 417 (9th Cir. 2015). Note that this is the test for the US Tax Court to intervene in an IRS decision not to abate interest; it is not necessarily the test that the IRS has to follow. 259 See Hancock, supra note 117, at 13-15. 260 See Lee, supra note 105, at 150; Bucaro, supra note 94, at 18; Pettyjohn, supra note 94, at 19; Wright, supra note 94 (USTC), at 13; Dadian, supra note 99, at 12; Landvogt, supra note 118, at 9; Guerrero, supra note 94, at 7; Hanks, supra note 105, at 7; Mekulsia, supra note 94 (USTC), at 23; T. Larkin, supra note 105, at 11-12 (“The Tax Court has acknowledged in previous opinions that the press of business and the Commissioner’s internal processing procedures require a certain amount of time”); Mathia, supra note 99, at 35; Corson, supra note 73, at 15; Jaffe, supra note 99 (USTC), at 14; Roudakov, supra note 105, at 9; Foote, interest relief on income tax debts: canada versus the united states n 979 the taxpayer must “identify a specific period of time over which interest should be abated as a result of such error or delay.”261 The court has been particularly critical of taxpayers who seek abatement of “all interest” in an assessment, characterizing such a claim as a request to be exempt from interest as opposed to a request for an abatement of interest.262 In a petition under section 6404(h), the burden of proof is on the taxpayer. This requirement can impose evidentiary challenges, given that taxpayers are generally not privy to all the inner workings of the IRS in the treatment of their files.263 More- over, once in litigation, taxpayers may not be able to gather sufficient information to reconstruct the IRS’s processing of their file owing to the US Tax Court’s relatively informal discovery procedures.264 To address this information imbalance, the court has recognized that because “[t]he [Secretary] is the best person to know what actions were taken by IRS officers and employees,”

[w]here the administrative record is silent regarding the actions taken on a taxpayer’s matter and the [Secretary] does not come forth with evidence to show that the employees assigned to the matter or involved in its review were actively working on it, . . . the unsupported determination may constitute an abuse of discretion.265

The court has also recognized a “duty of consistency” that requires the IRS to treat taxpayers in similar situations similarly, and has held that failure to do so can, in certain instances, constitute an abuse of discretion.266 Given the narrowness of circumstances where relief can be provided under section 6404(e)(1), coupled with the burden of proof and the standard of review, it is perhaps no surprise that taxpayers have rarely been successful before the US Tax

supra note 106 (USTC), at 23 and 27; Hornbacker, supra note 106, at 16 and 22; Howell, supra note 105, at 10; McGaughy, supra note 94, at 16; Krehnbrink #2, supra note 258, at 13-14; and Cosgriff, supra note 118, at 7. 261 Donovan, supra note 94, at 5-6; Guerrero, supra note 94, at 8-9; T. Larkin, supra note 105, at 13-14; and Goettee, supra note 73 (4th Cir.), at 218. 262 Hawksley, supra note 94, at 20, note 15; Corson, supra note 73, at 15; Donovan, supra note 94, at 6; Pettyjohn, supra note 94, at note 10; Guerrero, supra note 94, at 8; Hornbacker, supra note 106, at 19-20; and Spurgin v. CIR, 82 TCM 2001, at 846, note 13. 263 Tax Court Rules, rule 142(a). For a discussion of the burden of proof, see Hawksley, supra note 94, at 18, note 13. See also Corson, supra note 73, at 7; Hancock, supra note 117, at 8-9; Prakash, supra note 117, at 8; Sandberg, supra note 109, at 11; and Mekulsia, supra note 94 (USTC), at 22. 264 Gamino, supra note 92, at 48-49. 265 Ibid., at 49-50. See Bucaro, supra note 94, at 18-19; Jacobs, supra note 99, at 21; Mathia, supra note 99, at 39-44 (review of case law; five years of relief granted “[b]ecause the delay in countersigning the decision document is not explained by credible evidence in the stipulated record”); and Corson, supra note 73, at 8-9. 266 See Jaffe, supra note 99 (USTC), at 10-11. 980 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Court in challenges to a refusal by the IRS to abate interest. The few cases in which taxpayers have been wholly or partially successful have, for the most part, involved situations where either a taxpayer conscientiously sought to pay off a tax deficiency but was given incorrect factual information by the IRS,267 or the taxpayer’s file was misplaced or languished unattended for reasons that the IRS was unable to explain.268 There have also been cases in which the IRS has consented to abatement over the course of the US Tax Court proceedings.269 It bears noting that reviews of abatement decisions that may occur pursuant to section 6404(h) are sometimes combined with reviews of determinations made following CDP hearings, discussed below.270

IRC Sections 6320 and 6330 (CDP Hearings) IRC sections 6320 and 6330, which were among the marquee items enacted by the RRA 98, set out a regime for administrative and judicial review of collection actions by the IRS. Essentially, whenever the secretary files a notice of lien or before the secretary makes a levy on the taxpayer’s property or rights to property,271 a taxpayer must be offered an administrative hearing (that is, a CDP hearing) before an IRS official with no prior involvement with respect to the unpaid tax at issue. Section 6330(c)(2) lists the issues that may be considered at a CDP hearing, which include “offers of collection alternatives, which may include . . . an offer-in-compromise.”272

267 See Douponce, supra note 117, at 7-8; Kincaid v. CIR, TC Memo 1999-419, at 9 (taxpayers were incorrectly advised that interest was tolled pending related litigation); Harbaugh v. CIR, 86 TCM 596 (2003), at 598 (“The act . . . of misinforming petitioners about what their total liability would ultimately be was ministerial”); and Hancock, supra note 117, at 9 et seq. (the IRS’s actions were “pervaded” by “confusion and inconsistency”). See also Krugman, supra note 94, at 16-17; Hawksley, supra note 94, at 27-28 (abatement granted by the IRS when it miscalculated interest and failed to inform taxpayers of the full extent of their liability—at least until it corrected the error); and Goldberg v. CIR, TC Memo 2020-38, at 145-47. 268 Palihnich, supra note 103, passim (the IRS lost taxpayers’ returns for 11 years); Mathia, supra note 99, at 39-44; Goettee, supra note 73 (USTC), at 45-51; Dadian, supra note 99, at 16 and 18; Jacobs, supra note 99, at 21-22; and Bucaro, supra note 94, at 25-32. See also Bo, supra note 99, at 14 and 21-23 (an IRS officer refused to process an offer in compromise because of an out-of-date code in the IRS’s computer system). 269 See, for example, Geiger, supra note 108, at 7; Carpentier v. CIR, TC Memo 2014-75, at 2; and Goldberg, supra note 267, at 145-47. 270 Gray v. CIR, 138 TC 295 (2012), at 303-5. Examples of reviews of interest abatement refusals in the context of a CDP hearing include A-Valey Engineers, Inc. v. CIR, TC Memo 2012-199; Garavaglia v. CIR, TC Memo 2017-131; Roudakov, supra note 105; Santana, supra note 117; Love v. CIR, TC Memo 2019-92; Krehnbrink #2, supra note 258; and Goldberg, supra note 267, at 16 and 40-41. 271 The “lien” is the general US equivalent of a “certificate” under section 223 of the ITA, while the “levy” is the counterpart to a “garnishment” or “seizure” under section 224 or 225 of the ITA, respectively. 272 IRC section 6330(c)(2)(A)(iii). This list also applies to CDP hearings under IRC section 6320 by virtue of section 6320(c). interest relief on income tax debts: canada versus the united states n 981

If a taxpayer is dissatisfied with the final determination of the CDP hearing, the taxpayer may petition the US Tax Court for a review.273 The combined result of these provisions is that the US Tax Court effectively has jurisdiction to review the refusal by the IRS of an OIC made by a taxpayer over the course of a CDP hearing. If such an OIC seeks discretionary cancellation of interest on the basis of hardship, excep- tional circumstances, IRS error or delay, or other public policy reasons as discussed above, the US Tax Court thus has jurisdiction to review the IRS’s final decision. Neither section 6320 nor section 6330 specifies the standard of review to be used by the US Tax Court when reviewing the outcome of a CDP hearing. However, the court has repeatedly determined that any questions about the existence or quantum of an underlying tax liability are reviewed de novo,274 while all other issues (including the rejection of or refusal to consider an OIC) are reviewed for abuse of discretion.275 As with reviews of abatement decisions under IRC section 6404, the US Tax Court has proved to be highly deferential to the IRS with respect to its decisions to reject or refuse to process OICs, with decisions being “decidedly one-sided in favor of the Service.”276 The US Tax Court has repeatedly held that it will not “conduct an independent review of what would be an acceptable OIC” and will not “reweigh the equities” if “the Appeals officer follows all statutory and administrative guidelines and provides a reasoned, balanced decision.”277 Indeed, the court has repeatedly held that the IRS may decline to even consider an OIC unless all the statutory and administratively prescribed formalities (including payment of the application fee and compliance with minimum payment requirements) are scrupulously respected, or if the taxpayer does not provide all of the information requested by the IRS.278 Moreover, even if a taxpayer does comply with formalities and cooper- ates fully with the IRS, the IRS has no obligation to make any counteroffer or to enter into negotiations with the taxpayer.279

273 IRC sections 6320(c) and 6330(d)(1). 274 The taxpayer may also put at issue in a CDP hearing the “existence or amount of the underlying tax liability” if the taxpayer has not yet had an opportunity to contest it, such as in the course of a deficiency proceeding. IRC section 6330(c)(2)(B); see also section 6330(c)(4)(A). 275 The leading cases on the standard of review in CDP hearings are Goza v. CIR, 114 TC no. 12 (2000), at 9-10; and Sego v. CIR, 114 TC no. 37 (2000), at 10-11. Both of these cases are regularly cited by the US Tax Court. The “abuse of discretion” standard of review has been adopted by circuit courts in, inter alia, Olsen, supra note 186, at paragraphs 21-23; Murphy v. CIR, 469 F 3d 27 (1st Cir. 2006), at 32-33; Jones v. CIR, 338 F 3d 463 (5th Cir. 2003), at 466; Living Care, supra note 186, at 626 and note 5; Robinette, supra note 190, at 458-59; and Sadjadi, supra note 190, at 4. 276 Katz, supra note 153, at 1738 et seq. (review of case law). 277 Campbell v. CIR, TC Memo 2019-4, at 12; Wilson v. CIR, TC Memo 2012-229, at 21; and Link v. CIR, TC Memo 2013-53, at 12. 278 See the sources cited in note 186, supra. 279 Fargo, supra note 194, at 712-13; Keller, supra note 194, at 718-19; and Brombach v. CIR, TC Memo 2012-265, at 26-29 (review of case law). 982 n canadian tax journal / revue fiscale canadienne (2020) 68:4

As recently noted by the national taxpayer advocate, “CDP has been one of the federal tax issues most frequently litigated in the federal courts since 2001.”280 While the IRS wins the overwhelming majority of those cases (over 88 percent and 93 percent in 2018 and 2019, respectively),281 the national taxpayer advocate has noted that the availability of judicial review by the US Tax Court plays an important role in affirming taxpayers’ rights.282

OBSERVATIONS FOR CANADA The extended review in the preceding sections of this article reveals several differences between the Canadian and US regimes for discretionary interest relief that raise points for consideration on this side of the border. First, while discretionary relief of interest in the United States is generally limited to certain instances of IRS error or delay (IRC sections 6404(d), (e), and (g)) or to taxpayers affected by natural disasters (IRC section 7508A), this parsimoniousness is largely offset by a broader, and arguably more coherent, set of relieving provisions, including the interest-netting regime and provisions pertaining to carrybacks. As discussed above, the US interest-netting regime was enacted as part of the RRA 1998 to reflect the policy objective that “taxpayers should be charged interest only on the amount they actually owe.”283 Inasmuch as arrears interest aims to compensate the public treasury for the taxpayer’s use of its money, it is illogical—if not absurd and unfair—to charge interest on amounts that have been completely offset. While there are gaps in the US interest-netting regime that can lead to assessment of interest on phantom tax debts,284 such instances presumably arise less often in the United States than in Canada. Similarly, as also discussed above, the loss carryback provisions in IRC section 6601(d) provide that interest does not run on amounts that are offset by loss carry backs, starting from the filing-due date for the year in which the offsetting carryback arises. This longstanding rule is again consistent with the general principle that “taxpayers should be charged interest only on the amount they actually owe.”285

280 National Taxpayer Advocate, 2019 Annual Report to Congress, supra note 202, at 136. 281 Ibid., at 136-37. 282 Ibid., at 148. 283 Wells Fargo, supra note 208, at 1031. 284 As seen in M. Larkin, supra note 95 (USTC), at 21, where the court noted that “the interest provisions have an asymmetry that is not perfectly cured.” Other well-known situations where taxpayer payments do not stop the interest clock include cash-bond deposits (which were at issue in Ford Motor Company v. US, 768 F 3d 580 (6th Cir. 2014); cert. denied 135 S Ct. 2858 (2015)); in certain situations, companies involved in amalgamations (see David Berke, “More of the ‘Same’: Section 6621(d) in the Federal Circuit” (2018) 72:1 Tax Lawyer 201-22, passim; and Probasco, supra note 208, passim); and overpayments made by related taxpayers who do not file, or who are not permitted to file, a consolidated return (see the Treasury penalty and interest report, supra note 7, at 118). 285 Supra note 283. interest relief on income tax debts: canada versus the united states n 983

The IRC contains no counterpart to paragraph 161(7)(b), which bedevils Canadian taxpayers by assessing interest on tax liabilities over periods of time when those liabilities are completely offset by carrybacks, giving rise to applications for discretionary relief that are not always processed consistently. Learning from the US experience, Parliament could amend section 161.1 and paragraph 161(7)(b) of the ITA to provide for (1) comprehensive and automatic interest netting,286 and (2) not charging interest on tax debts over periods when those debts have been extinguished by carrybacks. Alternatively or additionally, the CRA could revise IC07-1R1 to include, as an additional ground for providing interest relief, wash transactions and other situations where the CRA effectively has the use of the amounts owing by the taxpayer. Such reforms would help to remedy one of the major gaps in Canada’s interest relief regime, discussed above. Indeed, as a more ambitious reform, Parliament and/or the government might commission a top-down review of the ITA’s interest provisions to ensure that they reflect the overarching policy objective that interest serves to appropriately compensate the government for the taxpayer’s use of the public’s money. The United States conducted several reviews of the IRC’s interest regime throughout the 1970s, 1980s, and 1990s, which ultimately led, in the RRA 98, to the equalization of the interest rate on overpayments and underpayments for non-corporate taxpayers, as well as to the adoption of the interest-netting regime.287 Following up on these reforms, the RRA 98 also mandated the Joint Committee on Taxation and the secretary of the Treasury to each conduct a study of the IRC’s interest and penalty provisions and their administration by the IRS, and to make “legislative and administrative recommendations . . . to simplify penalty or interest administration and reduce taxpayer burden.”288 Such a comprehensive study would be of use on this side of the border.

286 Of course, the significant administrative challenges in implementing a comprehensive netting rule will have to be addressed; see supra note 209 and the accompanying text. 287 Treasury penalty and interest report, supra note 7, at 32-35 and 115-24. 288 RRA 98, supra note 134, at section 3801. In response to this directive, both the Treasury department and the Joint Committee on Taxation (“the joint committee”) made detailed reports that were submitted to the Committee on Ways and Means, and were considered in a public hearing held on November 9, 1999 (the record of which is the SOCWM hearing, supra note 53). The Treasury department recommended that IRC section 6404(f) be expanded to include interest (the Treasury penalty and interest report, supra note 7, at 7-8 and 133-37). The joint committee made a similar recommendation as well as several others, including the harmonization of interest rates for individual and corporate taxpayers, the exclusion of interest from income for individual taxpayers, significant expansion of the circumstances in which interest abatement is possible, and the creation of a system of “dispute reserve accounts” that can stop the interest clock running on disputed amounts. See United States, Joint Committee on Taxation, Study of Present-Law Penalty and Interest Provisions As Required by Section 3801 of the Internal Revenue Service Restructuring and Reform Act of 1998 (Including Provisions Relating to Corporate Tax Shelters), vol. 1 (Washington, DC: US 984 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Second, the United States deals with tax delinquencies caused by hardship, extraordinary personal circumstances (such as medical expenses, bereavements, and so forth), or other unusual situations, not as matters for discretionary cancellation of interest per se, but rather under the aegis of the OIC regime. One advantage of this approach for taxpayers is that it takes into account not only interest, but also the underlying tax liability. Presumably, if a taxpayer is facing hardship from a tax liability, whether the liability consists of tax or interest is not relevant. Another potential advantage of dealing with interest relief under the aegis of the OIC regime is that relief can be made conditional on the taxpayer’s remaining in compliance with filing and payment obligations. This requirement has been found to be very effective in boosting taxpayer compliance. Learning from the US experience (and as others have suggested),289 Parliament could amend the ITA to allow the minister to compromise tax liabilities (including interest and penalties) along lines analogous to those set out by regulation in the OIC regime. A less ambitious reform, which could potentially be accomplished without legislation but would require the development of monitoring procedures within the CRA, would be for the CRA to start making interest relief conditional on compliance by the taxpayer with future filing and payment obligations. One might envisage that the CRA would be more willing to grant interest relief to delinquent taxpayers if doing so were seen as an effective step for bringing the taxpayer back into long-term compliance. In a related vein, although the IRS has historically been reluctant to use interest relief as a basis for settling longstanding tax disputes, Congress has repeatedly sought to push the IRS to be more flexible in this regard, including with the enactment of the RRA 98 (which reformed the OIC regime) and of the GOZA (which expanded the use of section 6404(g) to entice taxpayers to participate in the 2005 settlement initiative). As discussed above, settling cases consensually has been shown in the United States to improve long-term taxpayer compliance as well as to improve overall revenue collection—particularly if the settlement is made conditional on the taxpayer’s remaining in compliance with future obligations. To the extent that greater flexibility on interest can facilitate the reaching of settlements, such flexibility can benefit both individual taxpayers and the national treasury. Third, in many (though not all) situations, taxpayers denied interest relief (whether pursuant to IRC section 6404 or under the aegis of the OIC regime) have a judicial remedy available at the US Tax Court. Taxpayers thus obtain the benefit of both the court’s expertise in tax matters and its procedural rules, which seek to ac- commodate the fact that many of its litigants are self-represented. By contrast, in

Government Printing Office, July 22, 1999), at 3-4, 76-80, and 89-110 (www.congress.gov/ 106/cprt/JPRT57655/CPRT-106JPRT57655.pdf ). None of these proposals, however, were acted upon. See Gamino, supra note 92, at 50-52. 289 See Jackson, supra note 152. interest relief on income tax debts: canada versus the united states n 985

Canada, the Tax Court of Canada’s lack of jurisdiction over interest relief is a longstanding source of confusion and irritation. As others have observed,290 expanding the Tax Court’s jurisdiction to review discretionary decisions by the minister relevant to a taxpayer’s liability (including decisions to waive or cancel interest) would enable many taxpayers with longstanding disputes with the CRA to obtain a complete resolution of their disputes in a single forum, rather than having to institute multiple proceedings before different courts.291 Finally, the United States has no counterpart to the 10-year limitation set out in subsection 220(3.1) of the ITA, which, as discussed above, was heavily criticized when enacted in 2004 and arguably serves no useful purpose today. Learning from the US experience, Parliament could repeal the 10-year rule. Evaluation of these policy initiatives requires consideration of the underlying rationale for discretionary interest relief within the broader fiscal framework set out in the ITA. The comparative analysis above reveals three distinct rationales, not all of which have received the same degree of attention on this side of the border. First, discretionary interest (as well as penalty) relief aims to address the fact that a one-size-fits-all approach to sanctioning non-compliance—in the form of prescribed rates of interest and fixed, mandated penalties—can lead to inequitable results that fail to take into account the totality of the taxpayer’s circumstances. As mentioned at the beginning of this article, this rationale seems to have been the driving factor behind the original enactment of subsection 220(3.1) of the ITA, and it arguably remains the dominant feature of Canadian discourse on interest relief, at least in the context of income tax. Second, discretionary interest (though not penalty) relief addresses the fact that interest serves to compensate the government for the use of its funds by the taxpayer. As a result, interest relief becomes appropriate particularly in circumstances where the taxpayer does not, in fact, have use of the funds at issue, or where the funds have been effectively foisted upon the taxpayer through the actions of revenue officials completely outside the taxpayer’s control. The IRC’s interest-relieving and interest abatement provisions aim largely (if imperfectly) to reflect these policy concerns. In Canada, however, recognition of this basic rationale for interest relief— especially in the case of income tax—has remained limited. Finally, discretionary relief of interest (as well as penalties or even taxes) can provide a vehicle for rehabilitating delinquent taxpayers and motivating future long-term compliance. This policy rationale features particularly prominently in

290 See David Sherman, Annotation to Pine Valley Enterprises (available on Taxnet Pro); and Du Pont and Lubetsky, supra note 226, at S120-21. 291 Presumably, a taxpayer would have to comply with procedural requirements prior to seeking a review at the Tax Court, including making an application to the minister to waive or cancel interest and obtaining a final decision. If a taxpayer first raises the question of interest relief in the context of an appeal before the Tax Court, the appeal can potentially be held in abeyance until the minister makes a final decision on the matter. 986 n canadian tax journal / revue fiscale canadienne (2020) 68:4 the conception and implementation of the US OIC regime, which makes even the consideration of an OIC conditional on the taxpayer being up-to-date with his, her, or its filings, and also makes any final agreement conditional on future compliance by the taxpayer. The CRA has made some use of this rationale in the context of its voluntary disclosures program, which offers interest and penalty relief as an incentive for non-compliant taxpayers to come forward. However, the US experience demonstrates that there is arguably much more that could be done. canadian tax journal / revue fiscale canadienne (2020) 68:4, 987 - 1007 https://doi.org/10.32721/ctj.2020.68.4.pham

Tax Literacy: A Canadian Perspective

Anthony Pham, Antoine Genest-Grégoire, Luc Godbout, and Jean-Herman Guay*

PRÉCIS La fiscalité est indéniablement un sujet complexe qui rebute bon nombre d’individus. Pourtant, la compréhension de la fiscalité est cruciale alors que la planification de la retraite, l’épargne-étude et une portion importante de nos politiques sociales sont imbriquées dans notre système fiscal. Cette étude utilise une enquête par sondage pour mesurer les connaissances des Canadiens quant aux principes de base de l’impôt sur le revenu, pour observer l’évaluation qu’ils font de leurs connaissances et pour étudier leur comportement en termes de déclaration d’impôt. Faire preuve de littératie fiscale consiste à disposer de connaissances fiscales adéquates, d’une bonne appréciation des limites de ses connaissances, le tout permettant de prendre des décisions éclairées, notamment au moment de faire sa déclaration de revenu. Nos résultats montrent que les Canadiens performent bien aux questions portant sur les concepts de base mais qu’ils éprouvent plus de difficulté pour la compréhension des sujets plus complexes liés à la progressivité de l’impôt. Les résultats étaient assez similaires d’une province à l’autre, quoique les Québécois se distinguent avec des notes légèrement meilleures à notre questionnaire accompagnées par une évaluation plus négative de leurs propres connaissances. Notre méthode de mesure permettra le développement de meilleurs modèles expliquant la conformité fiscale ou l’opinion publique en matière de fiscalité et de redistribution.

* Anthony Pham, Luc Godbout, and Jean-Herman Guay are of Université de Sherbrooke; Antoine Genest-Grégoire is of Carleton University and the corresponding author (e-mail: [email protected]). This research has been made possible by the ongoing financial support provided by Quebec’s Ministry of Finance to the activities of the Chaire de recherche en fiscalité et en finances publique (Research Chair in Taxation and Public Finance) at Université de Sherbrooke. Antoine Genest-Grégoire also acknowledges the support of the Fonds de recherche du Québec—Société et Culture and the Canadian government’s Vanier-Banting Secretariat. The authors would like to thank David Boisclair, Kevin Milligan, Jennifer Robson, and Gregory Eady for helpful comments on the original draft of the survey questionnaire.

987 988 n canadian tax journal / revue fiscale canadienne (2020) 68:4

ABSTRACT Tax systems are complex structures that can be difficult for individuals to navigate. Understanding the way taxes are calculated and liabilities are assessed matters a lot when personal saving for retirement and education, and much of the government’s social policy apparatus, are closely integrated with the tax system. This study uses a survey to measure individuals’ knowledge about basic elements of the personal income tax, their perception of their own tax knowledge, and their tax-filing behaviour. One would hope that tax-literate Canadians would have a high level of knowledge of the way taxes work, and a realistic appreciation of the limits of their knowledge, and thus that they could make informed decisions, for example, when filing their tax returns. The survey data show that Canadians have good knowledge of basic tax facts but struggle when asked more complex questions regarding the progressivity of the income tax. Results were generally consistent across provinces with the notable exception of respondents in Quebec, who had higher marks on the authors’ tax quiz but lower self-assessed tax knowledge. The measurement instrument employed in the study will allow for a refinement of research exploring the drivers of tax compliance as well as political attitudes toward taxes and redistribution. KEYWORDS: PERSONAL INCOME TAXES n SELF-ASSESSMENT n SURVEYS n BEHAVIOUR n FILING n FINANCIAL MANAGEMENT

CONTENTS Background 988 The Study of Tax Literacy—Theoretical and Empirical Foundations 989 Theory 989 Survey of Comparable Research 991 Survey and Sample 993 Tax Literacy Questions 994 Results 996 Literacy Scores 996 Self-Reported Tax Knowledge and Tax-Filing Behaviour 996 Regional Differences 996 Analysis 998 Conclusion 1003 Appendix Changes in the Tax Literacy of Quebecers Between 2015 and 2019 1006

BACKGROUND Governments are increasingly preoccupied with the complexity of financial markets and products. Financial literacy has thus become a new object of focus for policy makers wary of the dire consequences of misguided financial decisions by citizens. Following that trend, in 2007 Canada expanded the mandate of the Financial Consumer Agency of Canada (FCAC) to include financial education.1 Two years

1 Financial Consumer Agency of Canada, “Financial Literacy Background” (www.canada.ca/en/ financial-consumer-agency/programs/financial-literacy/financial-literacy-history.html). tax literacy: a canadian perspective n 989 later, the Department of Finance created the Task Force on Financial Literacy to further study the financial knowledge and capabilities of Canadians.2 In 2014, the Canadian Financial Capability Survey “identified gaps in Canadians’ basic financial literacy knowledge and skills in areas such as budgeting, understanding credit and debt . . . and financial planning.”3 The federal government responded by launching the national strategy for financial literacy in 2015.4 Its aims were to support the financial well-being of Canadians by helping them to manage money anddebt wisely, and to plan and save for the future, while protecting them against fraud and financial abuse. Tax is an important component of financial management. A large portion of personal income is dedicated to paying taxes, and tax law interacts with work and savings decisions in complex ways. Individuals’ understanding of personal income tax is thus an important element in furthering governments’ objectives regarding financial literacy. Understanding of tax also likely plays a part in the complicated relationships between taxpayers and tax authorities, which shape tax compliance. The study described in this article examines tax literacy in Canada in 2019. The next section covers the theoretical origins of the concept of tax literacy, surveys some research on the topic, and presents our measurement instrument. We then set out our main descriptive results, followed by a detailed analysis of the data. The article ends with our concluding comments.

THE STUDY OF TAX LITERACY—THEORETICAL AND EMPIRICAL FOUNDATIONS Theory Tax literacy is still a young area of research, and this creates the double challenge of having few points of comparison and no consensus on the definition of the concept. By extension from work on financial literacy (such as Lusardi and Mitchell’s research), 5 a study carried out by Genest-Grégoire, Godbout, and Guay in 20156

2 Task Force on Financial Literacy, Canadians and Their Money: Building a Brighter Financial Future, Report of Recommendations on Financial Literacy (Ottawa: Task Force on Financial Literacy, December 2010). 3 Financial Consumer Agency of Canada, Implementing the National Strategy for Financial Literacy—Count Me In, Canada: Progress Report 2015 - 2019 (Ottawa: FCAC, 2019), at 4. 4 Ibid. 5 Annamaria Lusardi and Olivia S. Mitchell, “Baby Boomer Retirement Security: The Roles of Planning, Financial Literacy, and Housing Wealth” (2007) 54:1 Journal of Monetary Economics 205 - 24. 6 Antoine Genest-Grégoire, Luc Godbout, and Jean-Herman Guay, Littératie fiscale : Exploration du concept et bulletin de la population québécoise, Working Paper no. 2016/03 (Sherbrooke: Université de Sherbrooke, Chaire de recherche en fiscalité et en finances publiques, March 2016) (https://cffp.recherche.usherbrooke.ca/wp-content/uploads/2018/12/ cr_2016 - 03_litteratie_fiscale_expl_du_concept.pdf ). For an English version of the same results, see Antoine Genest-Grégoire, Luc Godbout, and Jean-Herman Guay, The Knowledge 990 n canadian tax journal / revue fiscale canadienne (2020) 68:4 developed a definition of tax literacy. In simplest terms, tax literacy can be described as “having the knowledge, skills and confidence to make responsible tax decisions.”7 To be able to make responsible tax decisions, individuals must be able to understand and correctly predict the tax implications of their choices. Thus, tax literacy does not rely exclusively on knowledge about taxes. To be considered tax literate, individuals need to possess knowledge, be able to apply that knowledge appropriately to real-world situations, and have the confidence to do so. More recent work by Bornman and Wassermann8 aims to structure future discussions about tax literacy by categorizing the research already done on the subject. Reviewing academic articles about tax literacy published between 1996 and 2017, Bornman and Wassermann identify three theoretical constructs underlying tax literacy: tax awareness, contextual knowledge, and meaning making. Tax awareness encapsulates the ability to recognize the factors affecting a tax decision and the ability to assess the consequences of that decision. Contextual knowledge refers to the understanding of both the tax law and the process through which taxpayers inter act with tax authorities. Finally, meaning making describes the ability to make informed tax decisions, based on both awareness and knowledge. It should be noted that the study of tax literacy is related to but distinct from the field of behavioural public finance. Researchers in that field modify traditional economic models by incor porating insights about bounded rationality (usually derived from experimental research), and focus on issues such as tax salience and loss aversion,9 whereas tax literacy research is much closer to legal and accounting studies, and focuses on the beliefs, perceptions, and values of individuals as explanatory factors for behaviour. Because the desired consensus on the theoretical framework for tax literacy has not yet been achieved, and because we want to be able to compare our results with those reported in the earlier study by Genest-Grégoire et al., we have chosen to use the definition from that study. In the context of Bornman and Wassermann’s framework, we could say that this definition prioritizes knowledge—specifically, knowledge of the tax law.

Deficit About Taxes: Who It Affects and What To Do About It, C.D. Howe Institute Commentary no. 484 (Toronto: C.D. Howe Institute, July 2017). 7 Genest-Grégoire et al., The Knowledge Deficit About Taxes, supra note 6, at 4. 8 Marina Bornman and Marianne Wassermann, “Tax Literacy in the Digital Economy,” paper presented at the Thirteenth International Conference on Tax Administration, Sydney, April 5- 6, 2018. 9 See the following for examples of this tradition: Raj Chetty, Adam Looney, and Kory Kroft, “Salience and Taxation: Theory and Evidence” (2009) 99:4 American Economic Review 1145 - 77 (https://doi.org/10.1257/aer.99.4.1145); Amy Finkelstein, “E-ztax: Tax Salience and Tax Rates” (2009) 124:3 Quarterly Journal of Economics 969 - 1010 (https://doi.org/10.1162/qjec.2009.124.3 .969); Raj Chetty and Emmanuel Saez, “Teaching the Tax Code: Earnings Responses to an Experiment with EITC Recipients” (2013) 5:1 American Economic Journal: Applied Economics 1 - 31 (https://doi.org/10.1257/app.5.1.1); and Alex Rees-Jones, “Quantifying Loss-Averse Tax Manipulation” (2018) 85:2 Review of Economic Studies 1251 - 78 (https://doi.org/10.1093/restud/ rdx038). tax literacy: a canadian perspective n 991

Survey of Comparable Research Most current research on tax literacy is focused on the links between tax literacy and tax compliance.10 Since the aim of our study is to understand the drivers of tax literacy, we surveyed the literature to find points of comparison relating to the links between tax literacy and socioeconomic factors. An article by Latiff, Noordin, Omar, and Harjito published in 200511 details one of the first attempts to measure the tax literacy of a specific population. The authors defined tax literacy as the capacity to read and write, and to understand issues of taxation. Thus, being tax literate involves both knowledge about taxes and the capacity to file one’s taxes. A sample of 143 Malaysian individuals was selected and divided by age, job, and ethnicity. The authors measured the participants’ capacity to file their own tax return themselves, whether they were interested enough to double-check their return, and the number of errors found on it. To measure tax knowledge specifically, participants were presented with a questionnaire and asked whether certain items were taxable or not. The authors found that about 30 to 35 percent of the participants did not master tax concepts, and that professionals, army officers, and business owners tended to score higher than farmers. Bhushan and Medury defined tax literacy as “the knowledge which an individual should possess in order to manage the issues concerning personal taxation effectively”12 and tried to discover which sociodemographic variables were associated with such a trait. Using a sample of more than 500 respondents from the Indian state of Himachal Pradesh, they administered a survey questionnaire containing 13 questions about “basic concepts of income tax, computation of tax liability, assessment rates, [and] deductions.”13 According to their data, a significant positive relationship exists between tax literacy and each of three variables—income, education, and age. Male respondents also had higher literacy scores than female respondents. Chardon, Freudenberg, and Brimble14 studied tax literacy as an integral component of financial literacy. Their survey included 65 questions covering knowledge about tax, confidence in making decisions, and attitudes about taxation. Using data from 604 Australian respondents, their work shows that there is a strong link between an individual’s confidence about his or her knowledge of tax matters and that knowledge as measured by the questionnaire.

10 For an overview of that research, see Marina Bornman and Pusheletso Ramutumbu, “A Conceptual Framework of Tax Knowledge” (2019) 27:6 Meditari Accountancy Research 823 - 39. 11 Ahmed Razman Abdul Latiff, Bany Ariffin Amin Noordin, Mohamad Raflis Che Omar, and Dwipraptono Agus Harjito, “Tax Literacy Rate Among Taxpayers: Evidence from Malaysia” (2005) 9:1 Jurnal Akuntansi & Auditing Indonesia 1 - 10. 12 Puneet Bhushan and Yajulu Medury, “Determining Tax Literacy of Salaried Individuals— An Empirical Analysis” (2013) 10:6 IOSR Journal of Business and Management 76 - 80, at 76. 13 Ibid., at 78. 14 Toni Chardon, Brett Freudenberg, and Mark Brimble, “Tax Literacy in Australia: Not Knowing Your Deduction from Your Offset” (2016) 31:2Australian Tax Forum 321 - 62. 992 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Many researchers have found links between education and tax literacy. To study this pathway, Moučková and Vítek15 surveyed 150 bachelor degree students enrolled in a taxation and tax policy program. The students were asked questions about income and consumption taxes derived from either Czech tax law or practical problems. The authors sought to see whether the completion of a university course on income taxation or consumption taxation led to better results in those areas (an education effect) and whether students who had filed their first tax declaration were more knowledgeable than those who had not (an experience effect). They concluded that education has an impact on tax knowledge but that the effect of experience is muted. Blechová and Sobotovičová16 also conducted a similar experiment involving Czech students. A study of the links between tax compliance and tax literacy by Nichita et al.17 included a section devoted to differences in literacy between different groups. The authors administered an 11 -item questionnaire to 358 Romanian taxpayers, assessing their tax literacy. Their results show that men had higher tax literacy scores than women, that literacy rose with the respondent’s level of formal education, and that employers had better scores than employees. The authors did not find any association between tax literacy and income levels. Closer to home, the study by Genest-Grégoire et al. cited above18 used a sample of 1,000 adults in Quebec and asked questions about income tax and tax progres sivity, confidence, and tax-filing behaviour. Their analysis revealed positive links between tax knowledge and family income, education, and age. The authors found no gender effect, but having children at home was associated with lower scores and being a homeowner with higher scores. The last five years have seen a rise in the number of studies of tax literacy. Because there is still no consensus on a definition, researchers use their own definitions, and comparison between different studies is therefore limited. In addition to the meth- odological differences, differences in tax regimes from one country to another preclude the development of universal measurement instruments. There is more hope for a shared understanding of concepts than for shared empirical tools, and Bornman and Wassermann’s work may help in charting such a course.19

15 Michaela Moučková and Leoš Vítek, “Tax Literacy” (2018) 66:2 Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis 553 - 59 (https://doi.org/10.11118/actaun201866020553). 16 Beáta Blechová and Šárka Sobotovičová, “Tax Education as a Part of Financial Literacy” (2013) 7:14 Trends Economics and Management 17 - 24. 17 Anca Nichita et al., “We Learn Not for School but for Life: Empirical Evidence of the Impact of Tax Literacy on Tax Compliance” (2019) 57:5 Eastern European Economics 397 - 429 (https://doi.org/10.1080/00128775.2019.1621183). 18 Supra note 6. 19 See supra note 8 and Marina Bornman and Marianne Wassermann, “Tax Knowledge for the Digital Economy” (2020) 13:1 Journal of Economic and Financial Sciences 1 - 11. tax literacy: a canadian perspective n 993

Survey and Sample Following Genest-Grégoire et al. as well as Lusardi and Mitchell’s work on financial literacy, we designed our measurement instrument around quiz questions. This now-popular method was also used in the Canadian Financial Capability Survey.20 It sets up a playful context and then asks respondents multiple-choice or true/false questions about the subject of interest. Our questionnaire surveyed respondents’ tax knowledge, their confidence in that knowledge, and their tax-filing behaviour. Data were collected through a commercial survey administered by Leger in the spring of 2019. The firm operates a 420,000 -member Canadian web panel from which the sample was drawn. A majority of panel members are recruited by random phone sampling, and the rest are recruited through social media or third-party campaigns. The sample was stratified by province, and the survey was pre-tested with 35 respondents. The original sample consisted of 3,156 individuals, but analysis was done on a restricted sample of 2,739 (with a response rate of 14 percent according to the AAPOR definition of “RR1”).21 This conservative subsample excluded any respondent who answered “I don’t know” or “I prefer not to answer” to any of the sociodemographic questions used in conjunction with tax literacy questions. Such a drop in sample size is common in similar survey research. To ensure sufficient subsample sizes in the two provinces with a distinctive tax situation—Quebec, because of its parallel provincial and federal tax systems, and Alberta, because of its lack of a provincial sales tax and its use of a flat income tax for many years—the sample is not proportional to the provincial distribution of the Canadian population. However, all results presented are weighted to be representative of the Canadian population in terms of age, gender, region, language, education, and presence of children at home. Table 1 describes the sample used.

20 See supra notes 3 and 4, and the accompanying text. 21 The American Association for Public Opinion Research, Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys, 9th ed. (Washington, DC: AAPOR, 2016), at 61. The AAPOR defines RR1 as follows: RR1 = I/[(I + P) + (R + NC + O) + (UH + UO)], where I = Complete interview, P = Partial interview, R = Refusal and breakoff, NC = Non-contact, O = Other, UH = Unknown if household/occupied, and UO = Unknown, other. 994 n canadian tax journal / revue fiscale canadienne (2020) 68:4

TABLE 1 Sample Description

Share Share (percent of (percent of weighted weighted Characteristics sample) Characteristics sample)

Province of residence Family income British Columbia ...... 14 Less than $20,000 . . . . . 9 Alberta ...... 11 $20,000-$39,999 ...... 19 Saskatchewan ...... 3 $40,000-$59,999 ...... 20 Manitoba ...... 3 $60,000-$79,999 ...... 16 Ontario ...... 38 $80,000-$99,999 ...... 14 Quebec ...... 24 $100,000-$149,999 . . . . . 22 New Brunswick ...... 2 Education Nova Scotia ...... 3 High-school diploma . . . . 32 Prince Edward Island ...... 1 CEGEP or other junior Newfoundland and college, or professional Labrador ...... 1 diploma ...... 40 Gender University degree . . . . . 27 Male ...... 51 Home ownership Female ...... 49 Tenant ...... 37 Age Homeowner ...... 63 18- 24 ...... 9 Children 25- 34 ...... 17 Without children at home . . 73 35- 44 ...... 16 One child or more . . . . . 27 45- 54 ...... 19 55- 64 ...... 18 65 and over ...... 21

CEGEP = collège d’enseignement général et professionel (Quebec junior college).

Tax Literacy Questions We included questions related to two categories of tax knowledge: tax coverage and tax progressivity. Indirectly, these questions test the respondent’s understanding of how taxes are calculated and how they vary from one situation to another. Because tax literacy involves the capacity to take action, we have chosen not to reuse Genest- Grégoire et al.’s questions on consumption taxes and to focus solely on items related to income tax. Our reasoning is that individuals can change their behaviour in response to tax incentives, and they have discretion in choosing how they file their income tax returns, but consumption taxes affect their behaviour only through the price of goods they buy. The decision to eat at home rather than go to a restau- rant is certainly affected in part by the price differential that comes from exemptions to value-added taxes, but knowing the source of this gap is not likely to change an individual’s behaviour. We also chose to exclude tax issues related to businesses, in order to reach a broader segment of the population, though we can speculate that tax literacy: a canadian perspective n 995 individuals who are knowledgeable about taxes in a business setting are probably also knowledgeable about personal income tax, which our survey captures. We measured tax literacy by asking respondents to answer the following questions:

n To your knowledge, are these sources of income taxable or not? [Answers: Tax- able, Not taxable, or I don’t know] n Withdrawal from a Registered Retirement Savings Plan (RRSP) other than to study or to purchase a home [Taxable] n Withdrawal from a Tax-Free Savings Account [Not taxable] n Employment Insurance benefits [Taxable] n Lottery winnings [Not taxable] n Child support received (after 1997) [Not taxable] n Profits from selling a primary residence for more than it was original[ly] bought [Not taxable] n Taxpayer A and Taxpayer B, both single individuals, each deduct a $1,000 contribution to their Registered Retirement Savings Plan (RRSP). Taxpayer A has an annual income of $50,000 and Taxpayer B has an annual income of $100,000. Which of these statements is true? n Taxpayer A will see his taxes reduced by a larger amount than Taxpayer B. n Taxpayer B will see his taxes reduced by a larger amount than Taxpayer A. [Correct answer] n Both Taxpayers A and B will see their taxes reduced by the same amount. n Neither Taxpayers A nor B will see their taxes reduced. n I don’t know. n I prefer not to answer. n Couple A has someone with $100,000 in income and someone else who has no income. Couple B has two people who both make $50,000 in income. Which of the following statements is true? n The two couples will pay the same total amount of income tax. n Couple A will pay more income tax. [Correct answer] n Couple B will pay more income tax. n I don’t know. n I prefer not to answer.

These questions cover the taxable nature of different sources of income, the way credits and deductions differ, tax progressivity, and the unit of taxation. Specifically, we used the second and third questions, on RRSP deductions and family taxation respectively, as measures of respondents’ understanding of progressivity (though the last question also captures understanding of the unit of taxation used in Canada). The difference between a deduction and a credit stems mainly from the existence of tax brackets, as does the difference in total tax paid by couples in which the partners exhibit significant income gaps. 996 n canadian tax journal / revue fiscale canadienne (2020) 68:4

RESULTS Literacy Scores Table 2 shows the share of respondents who gave the correct answer to each of our eight survey items on tax literacy. The global score is the simple average of the eight items. The average global score is 50 percent. Men did better than women on average, as did homeowners compared to tenants. Age, education, and family income were also all associated with higher results on the quiz. Age, education, and gender were associated with higher literacy scores in Indian, Australian, and Romanian samples.22 Family income was associated with higher literacy scores only in Australia and India. Genest-Grégoire et al. found no effect between gender and their measure of tax literacy for Quebecers, which constitute about a third of our sample.23 However, in that earlier study, Genest-Grégoire et al. found the same links that we found regarding age, income, education, and home ownership.

Self-Reported Tax Knowledge and Tax-Filing Behaviour In addition to asking quiz questions to objectively assess knowledge about taxation, we asked respondents about their own perception of their knowledge of such matters. The results are summarized in table 3. Every descriptive variable seems to have a significant impact on self-reported tax knowledge with the exception of being a tenant or homeowner. These results are in line with those of Chardon et al. for Australia.24 Chardon et al. also found that men rated their knowledge as being higher than that of women and that this evaluation tended to rise with age, income, and education levels. Respondents were also asked how they filed their taxes for the previous tax year. Table 4 shows the distribution of their answers to this question. The variables listed in these tables (gender, age, family income, etc.) are analyzed in relation to literacy later in the text.

Regional Differences Our large sample allows us to estimate results for different regions of Canada. These regions share the same basic tax architecture but also experience important differences in their respective fiscal situations. To obtain statistically relevant estimates, we merged the Manitoba and Saskatchewan subsamples and grouped all respondents from the Atlantic provinces (New Brunswick, Nova Scotia, Prince Edward Island, and Newfoundland and Labrador) in a single region.

22 See, respectively, Bhushan and Medury, supra note 12; Chardon et al., supra note 14; and Nichita et al., supra note 17. 23 Genest-Grégoire et al., The Knowledge Deficit About Taxes, supra note 6. 24 Chardon et al., supra note 14. tax literacy: a canadian perspective n 997 50 52* 48* 36* 43* 48* 49* 57* 58* 36* 45* 50* 51* 52* 57* 46* 51* 53* 43* 54* 51* 48* score Global 34 37* 30* 33* 30* 32* 31* 38* 37* 23* 29* 34* 31* 34* 45* 28* 34* 39* 28* 37* 34 32 Tax unit Tax 20 25* 14* 18 22 19 15 22 20 12* 15* 19* 19* 23* 26* 15* 20* 24* 17* 21* 20 20 RRSP contribution 37 40* 33* 12* 23* 29* 32* 48* 58* 17* 25* 35* 41* 40* 52* 30* 37* 44* 21* 46* 40* 29* Primary residence 31 31 30 31* 33* 36* 33* 31* 22* 30 34 29 28 29 32 28 31 32 33 29 27* 40* Child support 64 69* 58* 33* 44* 56* 66* 78* 86* 45* 60* 63* 68* 69* 70* 64* 67* 60* 54* 70* 68* 53* Lottery EI 68 69 67 48* 58* 67* 72* 79* 73* 53* 62* 70* 73* 71* 72* 65* 71* 67* 61* 72* 67 70 benefits 69 68 70 56* 66* 71* 65* 72* 77* 53* 65* 72* 72* 68* 77* 63* 72* 74* 62* 73* 70 67 (Table 1 is concluded on the next page.) (Table TFSA withdrawal 77 77 77 55* 65* 75* 79* 88* 88* 58* 72* 79* 78* 84* 84* 72* 80* 80* 68* 83* 79* 73* RRSP withdrawal Question (Percent) Literacy Tax . . to Each Answers Correct . 24 . or professional diploma . Male Female . 18- 25- 34 . 35- 44 . 45- 54 . 55- 64 . 65 and over . Less than $20,000 . $20,000-$39,999 . $40,000-$59,999 . $60,000-$79,999 . $80,000-$99,999 . $100,000-$149,999 . High-school diploma . CEGEP or other junior college, University degree . . Tenant Homeowner . children at home Without One child or more All respondents . Gender Age Family income Education Home ownership Children TABLE 2 TABLE 998 n canadian tax journal / revue fiscale canadienne (2020) 68:4

TABLE 2 Concluded

CEGEP = collège d’enseignement général et professionel (Quebec junior college); EI = employment insurance; RRSP = registered retirement savings plan; TFSA = tax-free savings account. Notes: We make no distinction between answering “I don’t know” and providing a wrong answer. Asterisks indicate a significant ( p < 0.05) difference between categories, according to a chi-square or Fischer exact test. For example, knowledge about the non-taxation of profits from the sale of a primary residence varies by age, but knowledge about the treatment of RRSP contributions does not.

As shown in table 5, there were no significant regional differences in answers about the taxation of RRSP withdrawals, lottery gains, and child support payments. Respondents in Quebec, followed closely by those in British Columbia, did better than other Canadians when asked about the taxation of tax-free savings account (TFSA) withdrawals. Respondents in the Atlantic provinces did significantly better than other Canadians when asked about the taxation of employment insurance (EI) benefits. They also did significantly worse when asked about the sale of a primary residence, while Quebecers scored highest on this question. On the overall score, residents of Quebec and the Atlantic provinces were the only ones to distinguish themselves from the others: Quebecers had a significantly higher score than other Canadians, and respondents in the Atlantic provinces had a significantly lower score.

ANALYSIS We ran ordinary least-squares regressions on the global literacy score. The socioeconomic variables presented earlier were used as regressors, as well as the region of the respondent. The use of such a model allows us to try to disentangle the association between these individual characteristics and higher tax literacy. The estimates can be interpreted, for example, as the link between being from a specific province and having a high literacy score, regardless of gender or household income. Stan- dardized estimates allow us to compare the size of such associations in terms of standard errors. Thus, a rise in age equivalent to one standard deviation on the age scale is associated with a rise of a little less than a third (0.298) of a standard deviation on the tax literacy scale. As shown in table 6, age, education, and family income are all significantly associated with higher tax literacy in this multivariate framework. Both age and education can be associated with higher human capital in general, which can translate to a better understanding of a complex subject such as tax. Higher income involves more tax to pay because of tax progressivity; accordingly, higher-income individuals have stronger incentives to learn about the subject. Age seems to be the most important of these three interconnected factors. Being a woman is associated with lower literacy scores and being a homeowner with higher ones. The link with gender is tax literacy: a canadian perspective n 999

TABLE 3 Distribution of Answers to the Question “To What Extent Are You Familiar with the Tax Rules That Affect You Personally?” (Percent)

Answers Not familiar Not very Somewhat Very at all familiar familiar familiar

All respondents 9 29 52 10 Gender Male* ...... 8 27 54 11 Female* ...... 9 32 51 9 Age 18- 24* ...... 16 46 34 4 25- 34* ...... 10 35 48 8 35- 44* ...... 8 34 52 7 45- 54* ...... 9 31 49 11 55- 64* ...... 6 23 59 12 65 and over* ...... 7 18 61 14 Family income Less than $20,000* ...... 19 34 40 7 $20,000-$39,999* ...... 11 32 47 11 $40,000-$59,999* ...... 10 29 52 9 $60,000-$79,999* ...... 5 31 55 10 $80,000-$99,999* ...... 6 29 56 8 $100,000-$149,999* ...... 5 24 57 13 Education High-school diploma* ...... 12 34 47 7 CEGEP or other junior college, or professional diploma* ...... 7 28 54 11 University degree* ...... 7 25 55 13 Home ownership Tenant ...... 12 32 48 8 Homeowner ...... 7 28 55 11 Children Without children at home* ...... 9 27 53 11 One child or more* ...... 9 35 49 8

CEGEP = collège d’enseignement général et professionel (Quebec junior college). Notes: Asterisks indicate a significant ( p < 0.05) difference between categories, according to a chi-square or Fischer exact test. For example, the distribution for self-reported tax knowledge is significantly different from the distribution for respondents with or without children at home, but not from the distribution for home ownership. commonly seen when assessing financial literacy and can be explained by within- couple specialization that favours men’s experience with finances25 and, perhaps, with tax filing. These effects are still much smaller than the effect of age. There is

25 Raquel Fonseca, Kathleen J. Mullen, Gema Zamarro, and Julie Zissimopoulos, “What Explains the Gender Gap in Financial Literacy? The Role of Household Decision Making” (2012) 46:1 Journal of Consumer Affairs 90 - 106 (https://doi.org/10.1111/j.1745-6606.2011.01221.x). 1000 n canadian tax journal / revue fiscale canadienne (2020) 68:4

TABLE 4 Distribution of Answers to the Question “How Did You Prepare Your Income Tax Return for 2018?” (Percent)

Answers Pen and Tax Friend or Tax Did not paper software relative professional file

All respondents 7 38 14 39 2 Gender Male* ...... 8 42 12 36 2 Female* ...... 6 34 17 41 2 Age 18- 24* ...... 7 25 27 37 4 25- 34* ...... 6 39 18 35 2 35- 44* ...... 3 43 12 38 3 45- 54* ...... 7 38 9 41 4 55- 64* ...... 8 41 15 35 1 65 and over* ...... 10 35 11 44 0 Family income Less than $20,000* . . . . 13 27 17 39 4 $20,000-$39,999* . . . . . 11 35 11 40 3 $40,000-$59,999* . . . . . 7 37 13 42 2 $60,000-$79,999* . . . . . 6 43 14 35 2 $80,000-$99,999* . . . . . 5 37 17 39 1 $100,000-$149,999* . . . . 4 42 16 37 1 Education High-school diploma* . . . 9 32 19 38 3 CEGEP or other junior college, or professional diploma* ...... 7 39 12 40 2 University degree* . . . . 6 43 13 37 2 Home ownership Tenant* ...... 8 37 15 37 3 Homeowner* ...... 7 38 14 40 1 Children Without children at home* ...... 8 39 14 37 2 One child or more* . . . . 6 35 14 43 1

CEGEP = collège d’enseignement général et professionel (Quebec junior college). Note: Asterisks indicate a significant ( p < 0.05) difference between categories, according to a chi-square or Fischer exact test. no statistically significant effect associated with having children at home. The only regional effect we observe is for Quebec respondents: they seem to have significantly higher tax literacy than other Canadians (where the reference region is Ontario, Canada’s most populous province). Quebec is unique among Canada’s provinces in requiring its residents to file a separate income tax return with provincial tax authorities, in addition to filing a federal return. Perhaps this added complexity explains the higher scores of Quebec respondents on Canada-wide quiz items. tax literacy: a canadian perspective n 1001

TABLE 5 Correct Answers to Each Tax Literacy Question, by Region (Percent)

Manitoba and All Atlantic Saskatch- British regions provinces Quebec Ontario ewan Alberta Columbia (average)

RRSP withdrawal . . . 76 78 76 75 80 79 77 TFSA withdrawal* . . . 64 74 67 66 67 73 69 EI benefits* . . . . 77 70 67 63 69 66 68 Lottery . . . . . 59 66 66 65 63 59 64 Child support payments . . . . 34 34 29 26 28 30 30 Primary residence* . . . . 23 41 36 35 38 37 37 RRSP deduction . . 12 18 21 22 20 21 20 Tax unit . . . . . 31 31 33 35 38 38 34 Global score . . . 47* 52* 49 48 50 50 50

EI = employment insurance; RRSP = registered retirement savings plan; TFSA = tax-free savings account. Note: Asterisks indicate a significant ( p < 0.05) difference between categories.

We also ran a similar regression on the self-reported knowledge variable to assess whether it shares the same drivers. As shown in table 7, age, income, and education display significant positive effects in this context. Age again seems to be the strongest driver. Women have lower self-reported knowledge than men, as was the case for tax literacy, but home ownership does not seem to have any effect. As to regional effects, Quebecers are again the only Canadians that diverge from the others. How- ever, the effect is the opposite of the one observed for measured tax literacy: Quebecers had lower self-reported knowledge than all other Canadians. The higher tax complexity that they face does not explain this more negative assessment very well. The high correlation between measured and self-reported levels of tax knowledge (r = 0.313, p < 0.00) suggests that respondents are not totally blind to the limits of their own expertise in respect of taxes. Finally, we ran a multinomial logistic regression to assess the effect of tax literacy on tax-filing behaviour. This model highlights which factors are associated with different choices among qualitative options offered to respondents. The results are shown in table 8. Each column represents a different way to attend to tax-filing responsibilities, and the rows show how some of these options are associated with the individual characteristics of respondents, such as gender or age. Table 8 shows that, surprisingly, the score obtained on our tax quiz does not have much predictive value for the tax-filing behaviour of respondents. Respondents with higher measured literacy had much higher odds (shown in bold font in the table) of 1002 n canadian tax journal / revue fiscale canadienne (2020) 68:4

TABLE 6 Predictors of Tax Literacy—Ordinary Least Squares Regression

Standard Estimate Standardized error t value Pr(>|t|)

Intercept ...... 0.28 0 0.013 21.92 0 Region British Columbia . . . 0.022 0.03 0.012 1.836 0.066 Alberta ...... 0.002 0.004 0.013 0.179 0.858 Manitoba and Saskatchewan . . . . 0.009 0.01 0.016 0.567 0.571 Quebec ...... 0.038 0.081 0.01 3.884 0 Atlantic provinces . . . −0.015 −0.017 0.016 −0.98 0.327 Female ...... −0.039 −0.089 0.008 −5.229 0 Age ...... 0.041 0.298 0.003 15.99 0 Household income . . . . 0.025 0.192 0.003 9.512 0 Education ...... 0.022 0.08 0.005 4.316 0 Children at home . . . . . 0.008 0.017 0.009 0.883 0.377 Homeowner ...... 0.031 0.068 0.009 3.459 0.001

Note: Observations, 2,739; residual standard error, 0.194; adjusted R2, 0.182. using software rather than the paper form to file their taxes themselves, but their increased knowledge was not associated with differences in their propensity to have someone else prepare their return or not to file at all. The results are similar if we simply distinguish those who handle tax filing themselves and the other respondents. However, the self-reported knowledge of individuals was significantly associated with differences in tax-filing practices. While more self-reported knowledge was not associated with a different propensity to do one’s taxes with software rather than with pen and paper, the difference is much more pronounced when considering the option of asking a friend, a relative, or a professional to produce the respondent’s tax declaration. Respondents considering themselves to be more knowledgeable have much higher odds of doing their taxes themselves. (The ratios of 0.56 and 0.55 indicate that for each additional level on the four-level knowledge scale beyond the first, the odds of having someone else prepare the return compared to preparing it by pen and paper drop by close to half.) Respondents in Alberta and the Atlantic provinces were much more likely to file with pen and paper, and women were much less likely to do so on average. Unsurprisingly, age was associated with higher odds of using pen and paper, and the opposite is true in relation to family income. The fact that individuals base their tax behaviour more on their perceived tax knowledge rather than their knowledge as measured by our instrument should not surprise us. Individuals cannot know the extent of what they don’t know. The fact that more self-reported knowledge leads to lower odds of asking someone to file taxes rather than do it personally seems intuitive. These results must also be framed in the larger discussion about tax compliance. As mentioned earlier, most explorations of the concept of tax literacy focus on its effect on compliance. The differences in compliance between those who prepare their declaration with pen and paper and tax literacy: a canadian perspective n 1003

TABLE 7 Predictors of Self-Reported Tax Knowledge—Ordinary Least Squares Regression

Standard Estimate Standardized error t value Pr(>|t|)

Intercept ...... 2.244 0 0.049 45.83 0 Region British Columbia ...... −0.013 −0.005 0.045 −0.279 0.78 Alberta ...... −0.06 −0.028 0.049 −1.227 0.22 Manitoba and Saskatchewan . . . . . −0.037 −0.012 0.061 −0.614 0.539 Quebec ...... −0.144 −0.087 0.037 −3.85 0 Atlantic provinces . . . . −0.009 −0.003 0.06 −0.158 0.874 Female ...... −0.083 −0.054 0.029 −2.894 0.004 Age ...... 0.09 0.187 0.01 9.242 0 Household income . . . . . 0.054 0.115 0.01 5.273 0 Education ...... 0.102 0.107 0.019 5.31 0 Children at home . . . . . −0.036 −0.021 0.035 −1.031 0.303 Homeowner ...... −0.016 −0.01 0.035 −0.452 0.651

Note: Observations, 2,739; residual standard error, 0.743; adjusted R2, 0.072. those who ask a friend or relative to do it for them are not so obvious. We might infer that those who use professional services have the lowest chances of being non- compliant, but they might also overestimate the burden of tax filing and thus could be better off doing it themselves. The choice among the filing options ismost likely influenced by legal knowledge about taxation, in combination with the complexity of the individual’s situation and available resources. Table 4, for instance, shows that pen-and-paper filing is highly skewed toward low-income respondents and older respondents—a finding that fits well with the resource factor.

CONCLUSION Our study is, we believe, the first attempt at a Canadian analysis of tax literacy. Similar to previous research in Quebec and around the world, it uses quiz questions embedded in a survey to measure overall levels of knowledge. Results show that tax literacy follows patterns that are also observed for general financial literacy: age, education, and family income are all associated with higher scores, as is being a man. There also appear to be regional effects, with Quebecers showing better than average tax literacy and Atlantic Canadians showing lower scores. Given Quebec’s parallel tax system, we think it is wise to mention that the counterpart to tax knowledge is certainly tax complexity. Countries differ significantly in the simplicity of their tax rules and tax-filing procedures. When considering the regional groups we identified as having lower scores on our test, it should be kept in mind that governments bear some responsibility for those results. Tax knowledge as measured by our instrument seems to be highly correlated with that knowledge as evaluated by respondents themselves. This matters for 1004 n canadian tax journal / revue fiscale canadienne (2020) 68:4 0.2 0.19 0.08* 0.03** 0.33 0*** 0.73 0.05* 0.12 0.41 0.03** 0.53 0.5 0.1 0.4 0.61 1.29 0.7 1.29 0.93 0.45 0.57 -value (OR) p Did not file 0.49 0.52 1.3 0.42 0.51 0.25 0.31 0.1 0.25 0.11 0.21 0.41 0.37 β (SE) − 0.63 − 0.68 − 2.27 − 0.92 − 0.5 − 0.35 − 0.07 − 0.79 − 0.56 0.18 0.03** 0.91 0.58 0*** 0*** 0.28 0*** 0.49 0.6 0.99 0.7 0.55 0.96 0.89 0.39 1.78 0.94 1.31 1.08 1.12 1 -value (OR) p Tax professional Tax 0.27 0.27 0.35 0.22 0.28 0.58 0.16 0.06 0.27 0.06 0.08 0.11 0.11 0.21 0 0.19 β (SE) − 0.36 − 0.6 − 0.04 − 0.12 − 0.93 − 0.06 0.13 0.03** 0.28 0.44 0.01** 0*** 0*** 0*** 0.08* 0.11 0.82 0.63 0.5 0.64 0.83 0.43 2.39 0.76 1.52 0.8 0.69 0.95 -value (OR) p Reference category: Pen and paper 0.3 0.32 0.41 0.24 0.33 0.87 0.19 0.07 0.42 0.07 0.12 0.23 0.22 Friend or relative β (SE) − 0.46 − 0.69 − 0.44 − 0.19 − 0.84 − 0.27 − 0.22 − 0.37 − 0.05 0.23 0.09* 0.32 0*** 0*** 0.05* 0*** 0*** 0.33 0.27 0.21 0.73 0.64 0.7 0.41 0.35 1.38 0.83 1.34 1.11 0.79 0.78 -value (OR) (Table 8 is concluded on the next page.) (Table p Tax software Tax 0.26 0.27 0.35 0.22 0.28 0.06 0.11 0.21 0.19 0.32 0.29 0.11 0.17 0.06 β (SE) − 0.31 − 0.19 − 0.45 − 0.35 − 0.9 − 1.06 − 0.23 − 0.24 to Individual Characteristics Behaviours According Tax-Filing . . . . for Different Propensities . British Columbia Alberta . . Manitoba and Saskatchewan Quebec Atlantic provinces . Region TABLE 8 TABLE Female . Age Family income Education . Children at home . Homeowner tax literacy: a canadian perspective n 1005 0.29 0*** 0*** 0.42 0.43 13.33 < 0.01; p -value (OR) p Did not file 0.8 0.2 2.59 0.61 β (SE) − 0.86 − 0.85 < 0.05; *** p < 0.1; ** p 0.72 0*** 0*** 1.17 0.55 13.81 -value (OR) p Tax professional Tax 0.15 0.43 0.11 2.63 0.39 β (SE) − 0.6 0.32 0*** 0*** 0.62 0.56 10.24 -value (OR) p Reference category: Pen and paper 0.48 0.13 2.33 0.43 Friend or relative β (SE) − 0.48 − 0.57 0*** 0.84 0*** 3.74 3.46 0.98 β , model coefficient; SE, standard error; OR, odds ratio; * -value (OR) p Tax software Tax 1.24 0.43 0.11 1.32 0.4 β (SE) − 0.02 . . Concluded . Tax literacy score Tax Akaike information criterion of the model, 6,577.09. TABLE 8 TABLE Self-reported tax knowledge Constant Note: Reference categories for factor variables in parentheses; 1006 n canadian tax journal / revue fiscale canadienne (2020) 68:4 literacy issues because individuals need to trust their knowledge level in order to make decisions based on it. We provide evidence that self-reported knowledge has an impact on the choices that individuals make regarding the preparation of their tax returns. However, tax literacy, as measured by our survey instrument, does not seem to be a driver of that decision. These results offer a foundation for further research. The importance of knowledge and literacy is recognized in various models of tax compliance as a factor contributing to the perceived legitimacy of tax authorities and thus favourable to compliance.26 Nonetheless, few instruments have been built to measure those attrib- utes. We provide such an instrument for the Canadian context and confirm that it exhibits some construct stability over time by comparing it with a previous similar survey (see the appendix to this article). Researchers in financial literacy often use the same three main quiz questions, and this allows for better comparison over time and in different settings. Those questions do not cover every aspect of individual financial planning, but they provide a useful general-purpose tool for the further study of financial literacy. We are hoping to contribute to the field of tax literacy in a similar manner, even though research in this area is still in the early stages of development. Our quiz does not cover every aspect of taxation, but we contend that it covers some of the most important ones and that it is simple enough to be adapted to the tax systems of other countries. We hope that it can be reused by other researchers trying to construct theories about tax knowledge, attitudes, and behaviours in the future.

APPENDIX CHANGES IN THE TAX LITERACY OF QUEBECERS BETWEEN 2015 AND 2019 Because a third of our respondents were from Quebec, we were able to compare our results with those reported in the earlier study by Genest-Grégoire et al., which focused on a similar population sample and included similar survey items.27 We believe that we may be among the first researchers who have been able to examine changes in tax literacy between two moments in time. Except for the sale of a primary residence, all items presented in our question about the taxation of different types of income were also included in the earlier study. (Genest-Grégoire et al. asked a question about the sale of a primary residence, but as a stand-alone true/false question.) Table A1 shows that Quebec respondents answered correctly in similar propor- tions in the two periods. The only notable differences are for items on RRSP withdrawals and the sale of a primary residence. These gaps cancel out, and the

26 See Erik Kirchler, Erik Hoelzl, and Ingrid Wahl, “Enforced Versus Voluntary Tax Compliance: The ‘Slippery Slope’ Framework” (2008) 29:2 Journal of Economic Psychology 210 - 25 (https:// doi.org/10.1016/j.joep.2007.05.004); and Knut Eriksen and Lars Fallan, “Tax Knowledge and Attitudes Towards Taxation; A Report on a Quasi-Experiment” (1996) 17:3 Journal of Economic Psychology 387 - 402 (https://doi.org/10.1016/0167-4870(96)00015-3). 27 Genest-Grégoire et al., The Knowledge Deficit About Taxes, supra note 6. tax literacy: a canadian perspective n 1007

TABLE A1 Correct Answers for Selected Tax Literacy Questions, Quebec Respondents, 2015 and 2019 (Percent)

2015 2019

RRSP withdrawal* ...... 69 78 TFSA withdrawal ...... 75 74 EI benefits ...... 71 70 Lottery ...... 66 66 Child support payments ...... 34 34 Primary residence* ...... 47 41 Mean ...... 60 61

EI = employment insurance; RRSP = registered retirement savings plan; TFSA = tax-free savings account. Notes: Asterisks indicate a significant ( p < 0.05) difference between categories. Data for 2015 were reported in Antoine Genest-Grégoire, Luc Godbout, and Jean-Herman Guay, Littératie fiscale : Exploration du concept et bulletin de la population québécoise, Working Paper no. 2016/03 (Sherbrooke: Université de Sherbrooke, Chaire de recherche en fiscalité et en finances publiques, March 2016) (https://cffp.recherche.usherbrooke.ca/wp-content/uploads/ 2018/12/cr_2016-03_litteratie_fiscale_expl_du_concept.pdf ). For an English version of the same results, see Antoine Genest-Grégoire, Luc Godbout, and Jean-Herman Guay, The Knowledge Deficit About Taxes: Who It Affects and What To Do About ,It C.D. Howe Institute Commentary no. 484 (Toronto: C.D. Howe Institute, July 2017). The data for 2019 were collected in the present study. difference in global score for the comparable question is not statistically different (even if the primary residence item is removed from the calculation). This indicates that tax literacy is generally constant over time, or at least that the questions we used measure the same thing. canadian tax journal / revue fiscale canadienne (2020) 68:4, 1009 - 29 https://doi.org/10.32721/ctj.2020.68.4.pf.tillotson

Policy Forum: Crisis, Cleanup, and the Prospect of Long-Term Fiscal Change

Shirley Tillotson*

PRÉCIS Dans cet article, Shirley Tillotson compare les politiques adoptées par le gouvernement canadien pour répondre aux exigences fiscales de la Seconde Guerre mondiale avec une mesure d’aide au revenu largement accessible qui a été instaurée en 2020 en réponse à la pandémie de COVID-19. Elle examine également les éléments de l’administration fiscale d’après-guerre afin d’établir des parallèles qui peuvent nous aider à réfléchir à notre situation actuelle et à ce que nous envisageons de faire dans un avenir proche. Soulignant les grandes difficultés et les inquiétudes qui ont pesé sur les décisionnaires au cours de la période plus ancienne, l’auteure souligne que nous devons faire preuve de modération dans nos attentes, et elle recommande de prêter attention tant aux contingences du moment (comme l’expérience de vie des principaux décisionnaires) qu’aux dynamiques structurelles à plus long terme en jeu (comme les changements dans la composition du marché du travail).

ABSTRACT In this article, Shirley Tillotson compares policies adopted by the Canadian government to address the fiscal demands of the Second World War with a widely available income- assistance measure introduced in 2020 in response to the COVID-19 pandemic. She also discusses elements of post-war tax administration with an eye to parallels that can help us to think about our current situation and plans for the near future. Emphasizing the great difficulties and worries that burdened decision makers in the earlier period, Tillotson underscores the need for modesty in our expectations and urges attention to both contingencies of the moment (such as the life experience of key decision makers) and the longer-term structural dynamics at play (such as changes in the composition of the labour market). KEYWORDS: PERSONAL INCOME TAXES n CANADA EMERGENCY RESPONSE BENEFIT n HISTORY n FISCAL PLANNING n TAX ADMINISTRATION n TRANSFERS

* Professor emeritus, Department of History, Dalhousie University, Halifax, and Inglis professor at the University of King’s College (e-mail: [email protected]).

1009 1010 n canadian tax journal / revue fiscale canadienne (2020) 68:4

CONTENTS Introduction 1010 Why Compare the Second World War PIT and the CERB? 1010 The Context of the Changes 1014 The Short-Term Response to the Changes 1020 The New Normal: Longer-Term Consequences 1024

INTRODUCTION The COVID-19 crisis is not a war. But we know that, like a war, the pandemic-related events of 2020 have massively disrupted lives, markets, and institutions. From such disruptions often come tax changes. So there may be policy-relevant perspectives that we can usefully draw on from the fiscal history of the Second World War and post-war reconstruction in Canada. In this article, I sketch some features of that history and discuss aspects worth thinking about in relation to our present fiscal situation and our near future. Within the scope of “fiscal,” I mainly focus on the personal income tax (PIT) measures introduced in 1942 and on the Canada emergency response benefit CER( B) introduced in 2020.

WHY COMPARE THE SECOND WORLD WAR PIT AND THE CERB? Given that there are fundamental differences between the world in 2020 and the world in the 1940s and 1950s, readers should not infer that “what worked then will work now.”1 There are no exact historical analogues. But five similarities in the two conjunctures that I am comparing make the comparison something to think with. First, although I am comparing a tax measure and a spending measure, both were emergency measures designed to rapidly achieve a crucial collective purpose. War finance before the 1942 budget had been careful and deliberate, drawing on the best economic thinking of the day. Initially, the parliamentary Opposition and other parties expressed some criticism during the House of Commons debate on the 1939 war budget. After this slow start, however, by 1942, even critics agreed that the war situation meant that revenue measures had to be passed quickly.2 Similarly, in our own time, by late March 2020, it had become clear to most decision makers that disease control was going to require drastic temporary constraints on citizens’

1 See David Hackett Fisher, Historians’ Fallacies: Toward a Logic of Historical Thought (New York: Harper Perennial, 1970), at 243-59, for a thoroughgoing caution against certain misuses of historical analogy. 2 Canada, House of Commons, Debates, September 9, 1939, at 55; July 15, 1942, at 4250; and July 16, 1942, at 4324. policy forum: crisis, cleanup, and long-term fiscal change n 1011 mobility and that compliance with those constraints would mean that many would lose their source of income. Given the paltry amount of many Canadians’ savings and the considerable amount of their personal debt, governments could not simply expect people to cope on their own. The policy choices were either to tolerate non- compliance by businesses that could still operate profitably in spite of the lockdown or to replace lost incomes with a tax-funded benefit for the millions who complied. At stake was public order, at several levels. As Sir Robert Borden wrote in 1935, destitute workers “will not consent to starve.”3 Second, the sheer scale of the PIT and the CERB mean that their impacts have been widely felt. Both the war income tax and the COVID-19 income support directly affected millions of Canadians. Between 1939 and 1942, the number of personal income taxpayers jumped almost 10-fold, rising from the mid-200,000s in the late 1930s to 2.1 million in the 1943 tax year, and reaching 2.6 million in 1948.4 The biggest growth came in 1942, when the basic personal exemptions were dropped to unprecedently low levels. This measure came on top of the earlier (1940) creation of a separate tax on wage income, the national defence tax (NDT).5 In 1942, these two kinds of income tax were patched together, with the result that in 1943, income taxpayers numbered 26.4 percent of the population aged 15-69.6 In 2020, in the four weeks between mid-June and mid-July, 18.4 percent of the same cohort (those aged 15-69) had received either CERB, employment insurance benefits, or the Canada emergency student benefit.7 An impact on this scale gave macroeconomic weight to the policies. The wartime income tax was an anti-inflation measure, while theCERB was an anti-deflation device. And both affected labour force participation. The CERB eased the desperate compulsion among the unemployed to take on work that would jeopardize public health, and the PIT initially discouraged some from work

3 Sir Robert Borden to Edgar N. Rhodes, 30 January 1935, Nova Scotia archives, Edgar Nelson Rhodes fonds, MG 2, vol. 1185, file 12, document no. 71753. 4 J. Harvey Perry, Taxes, Tariffs, & Subsidies: A History of Canadian Fiscal Development, vol. 2 (Toronto: University of Toronto Press, 1955), at 698. An emphasis on the 10-fold increase during the period can be found in Frank H. Brown, “Ten Months of Income Tax Administration” (2013) Commerce Journal at 14. 5 Shirley Tillotson, Give and Take: The Citizen-Taxpayer and the Rise of Canadian Democracy (Vancouver: UBC Press, 2017), at 177-78. For a description of the 1942 budget that includes the full range of tax measures in addition to the PIT, see Colin Campbell, “J.L. Ilsley and the Transformation of the Canadian Tax System: 1939-1943” (2013) 61:3 Canadian Tax Journal 633-70, at 659-63. 6 Calculated from the Dominion Bureau of Statistics, Canada Year Book 1945 (Ottawa: King’s Printer, 1945), at 130. 7 Statistics Canada, “Labour Force Survey, July 2020,” Daily, August 7, 2020 (www150.statcan .gc.ca/n1/daily-quotidien/200807/dq200807a-eng.htm). 1012 n canadian tax journal / revue fiscale canadienne (2020) 68:4 that helped the war effort.8 As large-scale programs, both landed heavily in people’s lives, sometimes in surprising ways.9 Third, the 1942 tax changes and the COVID-19 income assistance programs were bolted onto existing law and administrative practice, and in both cases, the programs to which they were attached had a history, a politics, and a set of problems ripe for reform. During the war years, both in short-term responses to criticism and in the cleanup after the crisis, politicians and administrators had to grapple with the pre-war politics, the new experiences of wartime, and unpredictable changes in conditions. I try here to dispel some of the golden mists that surround much of popular memory, and even some scholarship, on the political and social history of the Second World War.10 However impressive the accomplishments of wartime leaders, ours is a complicated country, and it is hard to wrangle us into a genuinely unani- mous pursuit of a common purpose. Then, as now, leaders aiming to motivate citizen collaboration in collective behaviour resorted to existentially inflected rhetoric, asking Canadians to see our behavioural choices as matters of life and death.11 To be sure, that rhetoric was sometimes warranted. Canadians, even non-combatants, had life-changing new experiences during the war. (Indeed, when Grace Fowler, a farm- hand and war plant worker, recalled her first time paying income tax in 1939, she said that the war meant that she had finally started earning a good wage, and she had been “paying her dues” ever since.)12 But during and after the war, ideas about effective fiscal policy and tax administration drew on beliefs and expectations shaped in the 1930s and before.

8 The impact on low-income Canadians’ work incentives is illuminated by a discussion of the interaction of CERB and income assistance in Gillian Petit and Lindsay M. Tedds, “The Effect of Differences in Treatment of the Canada Emergency Response Benefit Across Provincial and Territorial Income Assistance Programs” (2020) 46, Supplement 1 Canadian Public Policy S29-43. 9 As with the pandemic itself, the greater the number of people affected by a fiscal policy, the greater the number of cases of rare effects; thus, the scale of the event paradoxically empowers people who have a minority experience, because those minorities are made larger by a large- scale event. See Ed Yong, “Immunology Is Where Intuition Goes To Die,” Atlantic, August 5, 2020 (www.theatlantic.com/health/archive/2020/08/covid-19-immunity-is-the-pandemics -central-mystery/614956). 10 For Canadian history, one of the best general works to take this kind of approach is Jeffrey A. Keshen, Saints, Sinners, and Soldiers: Canada’s Second World War (Vancouver: UBC Press, 2004). 11 See, for example, Ilsley to Miss J. Barrett, 8 July 1942, Library and Archives Canada (herein referred to as “LAC”), Department of Finance records, RG 19, vol. 452, file 111-14M. Ilsley concludes that the tax burden is light compared to “the toll that a victorious Germany would levy on us.” 12 Older, Stronger, Wiser, directed by Claire Prieto (Montreal: National Film Board of Canada, 1989), 28 minutes (www.nfb.ca/film/older_stronger_wiser). The reference to paying income tax is at 11:10-11:20 minutes. policy forum: crisis, cleanup, and long-term fiscal change n 1013

Fourth, the urgency of the situation inevitably meant then, and means now, that decision makers will not hesitate to colour outside the lines in order to get short- term results, even though fast action produces its own problems. As C.D. Howe, the wartime minister of munitions and supply, supposedly said, “Here is the job. Get going. We will settle details later.”13 Of course, one person’s “details” may be another person’s essential mechanisms of fairness, quality control, and economy, as Bothwell and Kilbourn show in their history of Howe’s wartime management.14 Fifth, and finally, there is a sense that a large crisis may have a long-term impact and that planning how to shape that impact is part of a proper crisis response. As I write this, we are beginning, like the reconstruction committees of those war years, to take stock of raised aspirations and chronic concerns, and to consider what changes might be feasible and useful after the crisis. We are thinking about overlap- ping issues in the care economy (children, elders, public health, and health care generally), the service sector (retail, leisure, the arts, and entertainment), and public finance (tax, debt, and credit), and, more generally, issues relating to economic and social development (prosperity, stability, environmental sustainability, and a just society).15 The post-war measures and their origins that I discuss below suggest that, in choosing a focus (or several) for long-term change, we can best spend our energies on policies that effectively address either (1) problems widely acknowledged to be pressing even before the crisis or (2) problems that a narrower but influential group holds to be pressing. Without one of these kinds of effective backing, none of the responses to the crisis becomes a post-crisis reality. Even with such support, a policy idea will take priority (or lose it) because of circumstances and events that may not yet be predictable from our current moment. Over the coming decade, there will phases of policy, politics, and administration that are simply not visible during the early stages of the emergency in North

13 Frank Flaherty, “The New Income Tax Chief” ( January 1947) Canadian Business, at 62, quoted in Ralph L.C. Armorer and Ron Lemieux, Canada’s Taxfalers: History of the Administration at Revenue Canada—Taxation 1916-1981, vol. 1 (Ottawa: 1982), at 172. 14 Robert Bothwell and William Kilbourn, C.D. Howe: A Biography (Toronto: McClelland and Stewart, 1979), at 144-54. 15 Without attempting to be comprehensive but merely to illustrate, see, for example, Jennifer Robson, “Radical Incrementalism and Trust in the Citizen: Income Security in Canada in the Time of COVID-19” (2020) 46, Supplement 1 Canadian Public Policy S1-18; Joel Blit, “Automation and Reallocation: Will COVID-19 Usher in the Future of Work?” (2020) 46, Supplement 2 Canadian Public Policy S192-202; “Rebuild Canada,” Public Policy Forum (https://ppforum.ca/project/rebuild-canada); “Covid-19: The C.D. Howe Institute’s Complete Policy Analysis,” C.D. Howe Institute, October 29, 2020 (www.cdhowe.org/public-policy -research/covid-19-cd-howe-institute%E2%80%99s-complete-policy-analysis); The Recovery Project (www.recoveryproject.ca); and Mike Moffat and John McNally, “Making a Green Recovery Inclusive for All Canadians,” Smart Prosperity Institute, September 11, 2020 (https://institute.smartprosperity.ca/InclusiveGreenRecovery). 1014 n canadian tax journal / revue fiscale canadienne (2020) 68:4

America. The impact of the steps taken so far may be massively outweighed by future events. Our sense of this moment’s importance may come to seem overblown. But if 2020 and 1942 are indeed analogous, the big emergency fiscal measures launched this year will, like the 1942 tax measures, leave politicians, civil servants, scholars, and citizens with a lot to clean up and to try out in the months, years, and even decades to come. What is achieved in that post-crisis work will not simply be given by fate and human nature, but will depend on who reads well both the contingencies of our moment and the longer-term structural dynamics at play. The analysis that follows sketches this mix of factors and is divided into three parts, arranged chrono- logically: the context of the 1942 changes, the short-term response to them, and their longer-term consequences. After each part, I suggest some implications for our present situation. Readers will, I hope, also see others.

THE CONTEXT OF THE CHANGES Canada had a plan for war financing in 1939, and it was a good one. The plan was based on one imperative: Avoid the mistakes of the First World War. That meant a more pay-as-you-go war, drawing more heavily on current tax revenue and propor- tionately less on debt. The legacy of the Great War’s financing had been deep and sometimes violent class conflict arising from repeated economic crises, and in Canada, intractable and interminable tensions in fiscal federalism. Inequality of both regions and classes had to be avoided. The ministers of finance and revenue (Ilsley and Gibson, respectively), the governor of the—still new—Bank of Canada (Towers), and the powerful deputy minister of finance (Clark) all agreed. Start easy with taxation while war production and enlistment took up the economic slack, and then try to keep borrowing in check or at least balanced by taxation as needs changed. In 1939, these leading policy makers assumed, as no doubt most Canad- ians did, that the costs of the war would be limited and manageable.16 But, of course, events change plans. Two and a half years after that confident budget of September 1939, Canada’s people and politicians had suffered some hard blows, both at home and in battle. Canadians and the Canadian forces needed their government to supply basic needs and manage what threatened to be chaos.17 In response, the federal budget ballooned—expenditures of $1.3 billion were anticipated in fiscal 1940-41, up from $5.5 million in fiscal 1938-39, the last peacetime budget.18 In the spring of 1940, Prime Minister King worried that a budget that size might trigger a run on the banks. Later that fall, he told Cabinet colleagues that he feared Canada would

16 Campbell, supra note 5, at 636-38. 17 The foundational account is C.P. Stacey, Arms, Men, and Governments: The War Policies of Canada 1939-1945 (Ottawa: Queen’s Printer, 1970), at 37-54. Keshen, supra note 10, provides multiple perspectives on the strains that accelerated in 1941-43 and the pressures on government to respond. 18 Campbell, supra note 5, at 640 and 648. policy forum: crisis, cleanup, and long-term fiscal change n 1015 collapse financially if its spending continued at the current pace.19 His uncertainty was reasonable for a man who had seen the 1929 stock market crash metamorphose into a seven-year depression. King’s caution in 1940 was not a sign of poor foresight. He had little other than his earlier experience from which to judge. His government had an unnervingly thin body of actual economic data to work with.20 In January 1942, Governor Towers admitted that there was really no research-based consensus from which his staff could provide Finance with a statement on the ratio of national income to war expenditure.21 Equally challenging was the work that Finance faced in 1941 and 1942 to negotiate the wartime tax agreements, announced in 1941 and, after much haggling, wrestled into final form a year later. Ilsley admitted he had “never attempted anything half so difficult in [his] life” as these agreements with the provinces, which allowed the federal government to temporarily take over their income tax powers.22 In 1940, caution had been a sensible judgment, given how uncertain were the facts at hand and how constraining the constitutional situation. As war costs mounted, so did the political costs of war finance. By 1942, cautious steps were less feasible. While Finance was having to make guesses and take worrying risks, the Depart- ment of National Revenue lacked the resources to deliver on a new mass income tax system. To illustrate, in February 1939 the department added to its small staff of lawyers a 25-year-old lawyer, fresh from law school, Heward Stikeman. Stikeman was a bit surprised to be asked, early on, to write an Excise Profits Tax Act (EPTA) for war revenue. The bill that he drafted was passed, along with a suite of other war legislation, during the hectic week of September 8-13, 1939. But it was so badly designed that, eight months later, among the 1940 budget bills, a wholly revamped version replaced it. Stikeman was undoubtedly bright (as evidenced by his subsequent career), but throwing the job of drafting the EPT bill at someone who was still being trained in tax law was perhaps a sign that the department did not have much strength on its legal bench.23 Stikeman was not alone in being expected to rise to the

19 Robert A. Wardhaugh, Behind the Scenes: The Life and Work of William Clifford Clark (Toronto: University of Toronto Press, 2010), at 173 and 184. 20 David W. Slater and Robert B. Bryce, War, Finance, and Reconstruction: The Role of Canada’s Department of Finance, 1939-1946 (Ottawa: Department of Finance, 1995), at 32; Wardhaugh, supra note 19, at 157, 184, 216, and 218-19; and Duncan McDowall, Sum of the Satisfactions: Canada in the Age of National Accounting (Montreal and Kingston, ON: McGill-Queen’s University Press, 2008), at 53-64. 21 There were competing estimates that had been produced by the Bank of Nova Scotia, the Royal Commission on Dominion Provincial Relations, and the Dominion Bureau of Statistics: Wardhaugh, supra note 19, at 218-19. 22 Transcript of meeting, December 18, 1941, LAC, RG 19, vol. 2702, at 8, cited in Campbell, supra note 5, at 658. 23 Armorer and Lemieux, supra note 13, at 140-41; Campbell, supra note 5, at 644-45. 1016 n canadian tax journal / revue fiscale canadienne (2020) 68:4 emergency and cope with extraordinary demands. Top-level clerks were pressed into doing assessors’ work.24 The department hired an accountant to train staff and then, ruefully, saw the trainer quit for a private-sector job at almost double the salary.25 Trying to retain professional staff and training hordes of clerical staff was hard work. Equally so was the task of persuading the Treasury Board to a support a hir- ing surge. Deputy Minister of National Revenue C.F. Elliott pointed out that the department was administering two new taxes, the EPT and the NDT (distinct in rates and collection methods from the PIT), and was handling several labour-intensive changes to the Income War Tax Act (IWTA) itself. Elliott got more staff, but never enough to allow the department’s national network of offices to catch up, even on basic assessments. In 1944, Elliott told Ilsley that, if the public knew how swamped they were, there would be—Elliott’s deepest worry—a loss of confidence.26 But by then Ilsley already knew from his mailbox how hard a blow the mass income tax had delivered to public confidence in his policy choices, and how much of the weight of that blow came from citizens’ experiences dealing with the nation’s tax offices. The tax measures that ushered in a mass base for the PIT were delivered on June 23, 1942. They had been finalized only one week earlier, after an unusually short period of preparation.27 In mid-July 1942, Prime Minister King wrote in his diary that “Ilsley has gone much too far in his budget— . . . his taxes are unnecessarily heavy. They are putting a burden on the people greater than they can bear.”28 I have written in detail elsewhere about the multiple complaints that lower-income Canadians delivered to Ilsley between his 1942 and 1943 budgets.29 In summary, the new measures threatened financial security and even subsistence in working-class and lower-middle-class households. On security, the questions raised in the letters to Ilsley were about how to keep up insurance premiums or mortgage payments. On subsistence, the questions revealed how many different kinds of kin one wage supported, how close to the bone retirees with small investment portfolios lived, and

24 Canada, Senate, Special Committee on Income War Tax Act and Excess Profits Tax Act, Proceedings of the Special Committee no. 3, 20th Parl., 1st sess., November 20, 1945, at 63. 25 Canada, Senate, Special Committee on Income War Tax Act and Excess Profits Tax Act, Proceedings of the Special Committee no. 1, 20th Parl., 1st sess., November 14, 1945, at 22-23. 26 Armorer and Lemieux, supra note 13, at 132, 919, and 156; Special Committee on Income War Tax Act and Excess Profits Tax Act, supra note 25, at 18-19. 27 Wardhaugh, supra note 19, at 222-23. 28 Memorandum, Grant Dexter to G.V. Ferguson, July 19, 1942, 2, Queen’s University Archives, Dexter fonds, collection 2142, box 3, file 23; King diaries, 16 July 1942. 29 See Tillotson, supra note 5, at 179-88; Shirley Tillotson, “The Family as Tax Dodge: Partnership, Individuality, and Gender in the Personal Income Tax Act, 1942 to 1970” (2010) 90:3 Canadian Historical Review 391-426, at 392-94, 403-5, and 408-12; and Shirley Tillotson, “Relations of Extraction: Taxation and Women’s Citizenship in the Maritimes, 1914-1955” (2010) 39:1 Acadiensis 27-57, at 50-55. policy forum: crisis, cleanup, and long-term fiscal change n 1017

As this cartoon shows, there was editorial support for the 1942 tax measures, but the finance minister also found his mailbox stuffed with accounts of the threats to security and even subsistence faced by the new lower-income payers of personal income tax. Source: Halifax Herald, June 24, 1942, at 6. how deeply medical expenses (whether private insurance premiums or emergency expenses) cut. Small amounts of income tax equalled the cost of several days’ food supply. Like the CERB, the 1942 PIT made a big difference in household budgets, in ways that were not wholly expected by those who wrote the statute. Dealing with such widespread worries gave rise to other problems—problems that were about trust in the tax administration. Finance had attempted to ease the impact of the increased PIT by making some of it refundable, with interest, at the end of the war. Ilsley promised that government certificates for the amounts saved would be issued to employees. But these would come only after the taxpayer had filed a return and paid the tax, and the tax authority had assessed the return, made a refund, or required further payment. Employees feared that the employer might lose tax remittance records (in a fire, for example) and thus in effect misplace their savings.30 Also, overpayments on the employer-deducted NDT had become an issue: member of Parliament G.H. Castleden described a form so complex and delays so long that men who were owed small NDT refunds (equivalent to about $150 in 2020

30 A.M. Farr to Ilsley, June 30, 1944, LAC, RG 19, vol. 452, file 111-14E. 1018 n canadian tax journal / revue fiscale canadienne (2020) 68:4 terms) had ceased to expect them.31 The kind of record keeping that was required for accountability exacted a cost in fairness, a tradeoff that we might also have to consider. Given current rates of non-filing among low-income people, if we try to use the accountability mechanisms in the tax system, the results could be quite odd.32 Meanwhile, one class of lower-income taxpayer was discovering what tax people will recognize as a notch problem: lower-income taxpayers with no dependants were paying a 100 percent tax rate on the first $73 over earnings of $660, or on the first $162 over $1,200. For some, that meant unpaid overtime, and they started refusing those extra hours, a worrisome problem for war production goals.33 Finally, as Elliott’s post-war successor freely admitted, evasion went largely unchecked during the war. The department simply did not have the enforcement capacity. Rumours flew about dodges, large and small.34 Ilsley told one rancher that, just because other ranchers weren’t deducting tax from their farmhands’ pay, that didn’t mean he shouldn’t. In time, Ilsley wrote, the income tax office would find out and “correct” such non-compliance.35 The rancher could reasonably have been skepti- cal, as indeed some observers are today about how effectively taxes are collected. I have emphasized here the trouble, conflict, and worry that surrounded the rapidly devised tax measures of 1942 to make the point that, no matter how wise a general plan may be, it will be battered by the specific context of its implementation. Responses from citizens and various interest groups quickly showed that the 1942 tax measures in their initial form did not sufficiently take into account the conditions of those they taxed. To be sure, policy makers had given some thought to the impact on household budgets. Making medical expenses deductible up to 5 percent of income was evidence of this. But citizen response in the weeks and months after the 1942 budget gave Finance a more complete understanding of the pinch points. Deploring the competition between saving and paying new taxes, the notch problem, the unfairness of deductions from irregular earnings, and more, lower-income taxpayers explained that the new measures were cutting into the household budget cushion on which people depended for emergency needs.36 King knew enough to be worried about the electoral costs of these experiences. But the choice to tax smaller incomes was the easy and familiar one. During the

31 Canada, House of Commons, Debates, July 17, 1942, at 4329. 32 Anna Cameron, Lindsay M. Tedds, Jennifer Robson, and Saul Schwartz, Tax Policy Trends: The Merits of Automatic Income Tax Assessments for Low-Income Canadians (Calgary: University of Calgary, School of Public Policy, February 2020) (www.policyschool.ca/wp-content/uploads/ 2020/02/Tax-policy-Trends-Feb-2020.pdf ). 33 Tillotson, supra note 5, at 179-81, 190-91, and 202-4. 34 Ibid., at 187-88. 35 C.H. Pitts, secretary, Ashcroft Ranchers’ Association to T.J. O’Neill, MP, 11 May 1943; J.L. Ilsley (drafted by J. Harvey Perry), to C.H. Pitts, 25 May 1943, LAC, RG 19, file 111-14-754. 36 Canada, House of Commons, Debates, June 23, 1942, at 3580. policy forum: crisis, cleanup, and long-term fiscal change n 1019 fiscal crisis of the Depression, the PIT base had been broadened in the same way, though to a lesser extent, by reducing the amount of exempt income.37 Even in the wartime crisis, the government preferred to extend the old approach rather than adopt a new one. In retrospect, it is easy to see the seeds of the new Keynesian thinking in this period of war financing. But, perhaps inevitably, the experts in Finance were thinking about the present with tools of the past. Faced with a mounting crisis, they achieved policy-making speed by avoiding too much novelty. Thus, the strategy they chose (to broaden the tax base) was a replay of an earlier crisis response. New tools are risky and sometimes not very good: consider how limited were the newly emerging national income statistics that policy makers needed to help them “see” the economy. They relied considerably on best guesses and took some chances. Old intuitions, like those of Prime Minister King, were not a bad guide to grasping the social impact of the 1942 taxes. But new measures adopted for old reasons sometimes usher in unexpectedly fundamental changes. In defending the mass income tax, Finance began to articulate a new role for the federal government in social provision.38 The battering of the 1939 plan by hard circumstances is an example of the processes by which old thinking can produce unexpected and lasting change. Today, the CERB is justified partly in familiar terms—using income assistance to support consumer demand. But the CERB’s extraordinary scale prompts both hopes of a lasting expansion in income support and worries about a fiscal crisis (as did war debt and social security in the 1940s).39 Whether significant challenges to monetary and fiscal orthodoxies can—or should—prevail I leave to those who are more expert than I in economics. But political history suggests that, whatever innovations may be proposed, there will need to be a concept that blends old and new to help sell significant change, as “social security” did. Emerging in the 1930s, social security was an interdisciplinary and politically evocative concept.40 Its technical meaning

37 Two provincial governments—Manitoba’s and British Columbia’s—went after smaller incomes even more aggressively during the Depression and prompted vigorous mass protest as a result. Tillotson, supra note 5, at 94-102. In the course of the preparation of the 1942 budget, Finance economist R.B. Bryce gave serious thought to taxing capital. But the method that Bryce proposed was complex, and perhaps beyond the capacity of National Revenue to administer at that point. For Bryce’s proposal, see Memorandum on the Financing and the Conscription of Capital, April 1, 1942, LAC, RG 19, vol. 3978. 38 Tillotson, supra note 5, at 199-201. 39 Ibid., at 210-11 and 238-40. 40 Raymond B. Blake, From Rights to Needs: A History of Family Allowances in Canada, 1929-92 (Vancouver: UBC Press, 2009), at 68-76. For the origins and resonances of “social security” in politics and social work, beyond the economic dimensions of income support, see Suzanne Morton, Wisdom, Justice, and Charity: Canadian Social Welfare Through the Life of Jane B. Wisdom, 1884-1975 (Toronto: University of Toronto Press, 2014), at 87-88, 181-83, and 207-8; and Nancy Christie, Engendering the State: Family, Work, and Welfare in Canada (Toronto: University of Toronto Press, 2000), at 266-70. 1020 n canadian tax journal / revue fiscale canadienne (2020) 68:4 lay in social insurance. But its evocation in relation to the 1941 Atlantic charter41 drew to it the larger resonances of freedom from fear and want, achieved by whatever means. A year later, Britain’s influential Beveridge report linked social security, justice, and international security.42 The meaning of social security was further enlarged by the role of unemployment insurance as a macroeconomic tool that blended prosperity for all and care for the vulnerable. The fuzzier the meaning of social security became, the easier it was to say, as did the Throne speech in 1943, that it served “the general interest.”43 Today’s equivalently evocative term might be “resilience.” Like “social security,” resilience is a familiar concept that travels across disciplines—psychology, ethics, international relations, and economics. As a metaphor, it reaches many publics and can mobilize multiple government ministries. Its admittedly obnoxious malleability may enable it to couple otherwise divergent political currents, reaching toward a “general interest.”44 There will be arguments about what kinds of policy will generate resilience, addressing problems in medical care, in supply chains, in business plans, and in household labour choices. But the notion of resilience might help to achieve broad consent to lasting change.

THE SHORT-TERM RESPONSE TO THE CHANGES In both 1942 and 2020, big steps have been followed closely by fixes—measures to repair unintended effects and answer criticisms. In 1942, some of these came during the course of the second reading of the IWTA amendment bill; others came eight

41 Blake, supra note 40, at 82-83. 42 William H. Beveridge, Social Insurance and Allied Services (London: King’s Printer, 1942). 43 The 1943 throne speech is quoted in Leonard Marsh, Report on Social Security for Canada, A New Edition with an Introduction by Allan Moscovitch (Montreal and Kingston, ON: McGill- Queen’s University Press, 2018), at xvi. In this report, the term’s slipperiness was already evident. Marsh referred to social security as a method of welfare provision grounded in social insurance, but acknowledged that security from hazards required social services that were not capable of being provided even through social insurance against risks to income. And he identified policies to support employment as the “first positive measure in providing social security,” even though he also described a job as a source of “individual security.” Marsh and several of the authorities that he relied upon referred both to the fears and uncertainties of post-war demobilization and to the memories of the Depression as occasions of insecurity with bad social consequences. In this context, “social” was both a resource (contributions and taxes contributed by the many to be allocated rationally by public administration) and an object of worry, involving social risks, such as anomie and family strife, along with economic risks. See Marsh, ibid., at 10, 14-17, 24, 26, 65-66, 132, 167, and 170-71. For the coupling of social and international security, see the Beveridge report, supra note 42, at 172. 44 The organizing power of metaphors that migrate among different disciplines and spheres of activity is discussed in Daniel T. Rodgers, Age of Fracture (Cambridge, MA: Harvard University Press, 2011), at 9-13. I explain the persuasive effect of such metaphors in Shirley Tillotson, Contributing Citizens: Modern Charitable Fundraising and the Making of the Welfare State 1920-66 (Vancouver: UBC Press, 2008), at 19-20. policy forum: crisis, cleanup, and long-term fiscal change n 1021 months later, in the 1943 budget. The concerns of at least some voters weighed heavily. That weight came from two kinds of leverage. One kind—always visible to decision makers—was the intervention of well-organized groups, even small ones, which defended against a dramatic impact on their interests. The other kind was non-compliance by the unorganized (citizens who were not represented by a formal association), expressed through the withholding of necessary labour—a form of non-compliance that breaks no laws but gums up the machinery of any large project. In the pandemic, as in the war years, even an unorganized “no” from many citizens can be influential. It was the latter kind of pressure that prompted Ilsley to introduce in July 1942 one of the first measures to provide relief from thePIT that he had just proposed in June. The relief was designed to keep married women at their paid jobs. There was no organization of wage-earning women to carry this cause, but in June and early July, married women and their husbands had written to Ilsley to protest the impact of the 1942 tax measures on their household budgets. With a lighter tax burden, they had been able to pay for home help to care for their children and manage housework. Now, the household accounts no longer balanced. Their motives for working had included patriotism, along with catching up on household debt and maybe building savings. Patriotism was not enough by itself. Under the new tax measures, wives’ wage earning left the family with little or nothing extra after they paid the costs of their employment. Ilsley’s response was to allow married men to keep their “married” exemption and so reduce the family’s tax contribution. He regretted having to make even this concession, fearing that it would “prove a considerable cost to the treasury.”45 But it was required in order “to prevent this exodus from employment.”46 The more familiar kind of leverage, that of a well-organized small group deeply affected by a tax measure, was apparent in Finance’s rapid response to the concerns of men who were paying alimony to their ex-wives. Given the context of 1940s divorce law and norms, these would generally have been relatively prosperous men. Just a week after Ilsley’s budget speech, there appeared in the Toronto press a news story and an advertisement about a lawyer, a partner in a “well-established and well- respected law firm,” who was organizing an Association for the Relief of Alimony Payers. Inspired by developments in the United States, this lawyer who, “for obvious reasons, [did] not wish his name divulged,” urged that alimony payers have deducted from their taxable income the whole of the amount that they paid to their ex-wives, and to have those payments taxed in the ex-wives’ hands (at a lower rate for those with a lower income).47 Concerned again with the cost to the treasury, Ilsley and National Revenue Minister Gibson quickly came up with a “split-the- difference” solution. Men who paid alimony would still be taxed on the whole of their

45 Canada, House of Commons, Debates, July 15, 1942, at 4253. 46 Ibid. 47 “Alimony Payers Urge Ex-Wife Pay Income Tax,” Globe and Mail, July 3, 1942. 1022 n canadian tax journal / revue fiscale canadienne (2020) 68:4 income, but their ex-wives would be taxed on the payments they received; the amount of tax they paid could then be taken as a credit by their ex-husbands. The matter was “under consideration” on July 17, 1942, and a resolution to amend the bill was accepted in the House on July 22, just a month after the budget speech.48 It seems that alimony payers had heft in the halls of power. Small groups that have a big stake in tax questions often do. When a large group also becomes a well-organized one, able to mobilize for non-compliance, it can have a massive effect. This combination, in the form of the wartime labour movement, drove Ilsley to reluctantly introduce further changes in his 1943 budget. In 1940-41, organized labour had spoken out in forums such as the National War Labour Board about the heavy taxation of working people’s small incomes. In 1941-42, there was a surge of organizing efforts and strikes for bargaining rights among the increasingly successful industrial unions. A political alliance was emerging between these unions and the Canadian Commonwealth Federation (CCF), fuelling political wins for that party. A poll by the newly arrived Gallup organization suggested that there were more wins to come. Prime Minister King and Finance Minister Ilsley tried to gauge how much and what kind of tax relief would assuage wage earners’ real worries and at the same time reinforce the message that this was no time to slacken the war financing effort. King thought that concessions would keep the electorate from turning to the left. In the 1943 budget, Ilsley delivered a series of small modifications—and a partial remedy to the notch problem—that returned some flexibility to the household budgets of wage-earning families. And National Revenue organized administrative accommodations, such as allowing travelling construction workers to claim living expenses.49 This kind of ad hoc measure reflected the heightened influence of workers whose labour was now in high demand. In addition, that influence showed up when, in the 1943 budget, Ilsley foreshadowed 1944’s headline social security measure, the introduction of family allowances. The labour movement’s long organizational struggles during the 1930s and the early 1940s had made workers’ unions and their alliance with the CCF a force whose weight was more than mere moral suasion. When labour markets changed again after the war and Cold War politics changed electoral politics, the labour movement lost most of the particular leverage that it had been able to deploy in wartime.50

48 Canada, House of Commons, Debates, July 17, 1942, at 4361; and July 22, 1942, at 4532. 49 Tillotson, supra note 5, at 246-7 and 404, note 42. 50 There is considerable debate on the extent to which the labour movement was well served or not by the new regime of industrial relations law that grew out of the foundations laid during the war. Some (for example, Panitch and Swartz, Heron, McInnis) emphasize that the legal regime constrained labour tactics and supported management rights to the detriment of workers. Others (Morton, Tillotson) see greater value in the relative institutional stability that positioned labour more effectively for political action. Leo Panitch and David Swartz, From Consent to Coercion (Toronto: Garamond Press, 1985); Craig Heron, “Male Wage-Earners and the State in Canada,” in Michael Earle, ed., Workers and the State in Twentieth Century Nova policy forum: crisis, cleanup, and long-term fiscal change n 1023

The immediate fixes introduced in response to the unexpected consequences of the 1942 tax policy remind us how hard it is to anticipate the effect of a policy of such scope and so direct a pocketbook touch. Equally, those fixes illustrate how some kinds of power can instigate a rapid adjustment in government policy. Of particular importance in analyzing our present situation is the non-compliance of the unorganized. It underpins ad hoc mass protest. When social control matters—in some kinds of war and also during a pandemic—a more diffuse form of power may carry unusual weight.51 Getting people to stay at home or wear masks is exactly the sort of governance project that cannot depend mostly on coercive power. That kind of governance requires mass compliance—social control—and as a result, it empowers the unorganized. The need for social control can also change a latent community of interest into a force that is newly salient for policy. New structural conditions can give significance to the dissent or consent of people who previously might not have seen themselves as sharing a common interest. In our present moment, that kind of structural change may fuel real changes in the care economy. Hundreds of thousands of Can- adians—in a recent estimate, roughly 40 percent of Canada’s 1.2 million working families52—have come to depend on paid child care to support their household’s labour income. That child-care services support women’s labour force participation is well established.53 In the 1940s, women’s value as a reserve labour force—one whose care work had not been properly priced in the tax measures—prompted a quick fix when married women said “no” to effectively uncompensated war work.

Scotia (Fredericton, NB: Acadiensis, 1989), 252-53; Peter S. McInnis, Harnessing Labour Confrontation: Shaping the Postwar Settlement in Canada, 1943-1950 (Toronto: University of Toronto Press, 2002); Desmond Morton, Working People: An Illustrated History of the Canadian Labour Movement, 5th ed. (Montreal and Kingston, ON: McGill-Queen’s University Press, 2007), at 184-85, 198, 201, and 227; and Shirley Tillotson, “ ‘When Our Membership Awakens’: Welfare Work and Canadian Union Activism, 1950-1965” (1997) 40 Labour/Le Travail 137-70. In policy terms, the foundational work is David Kwavnick, Organized Labour and Pressure Politics (Montreal and Kingston, ON: McGill-Queen’s University Press, 1972). Kwavnick analyzes closely the policy areas in which labour continued to be an effective policy force in the late 1950s and early 1960s. 51 Consumer activism and the organized consumer movement deploy the power of non- compliance, and especially so in the political economy of the United States and Canada, in which state management of consumption is shot through with politics. For a survey of the American literature on consumer politics, see Lawrence B. Glickman, “Consumer Activism, Consumer Regimes, and the Consumer Movement: Rethinking the History of Consumer Politics in the United States,” in Frank Trentmann, ed., Oxford Handbook of the History of Consumption (Oxford: Oxford University Press, 2012), 399-417. 52 Alexandre Laurin and Kevin Milligan, Tax Options for Childcare That Encourage Work, Flexibility, Choice, Fairness and Quality, C.D. Howe Institute Commentary no. 481 (Toronto: C.D. Howe Institute, May 2017), at 4-5. 53 Rebecca Wallace and Elizabeth Goodyear-Grant, “News Coverage of Child Care During COVID-19: Where Are Women and Gender?” (2020) Politics & Gender 1-8 (https://doi.org/ 10.1017/S1743923X20000598). 1024 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Today, care work, once done mainly by women and mainly outside the measured economy, has a very much clearer market cost, because women’s labour is no longer mostly unpaid. Policy makers may have to respond to the force of many quiet and not-so-quiet refusals and their impact on economic life. There are solutions waiting to be implemented.54

THE NEW NORMAL: LONGER-TERM CONSEQUENCES In 1945, the outlines of the tax transfer system that would become Canada’s welfare state were emerging. Amid much change and uncertainty, there was a sense of urgency around tax. By December 1946, there was a new finance minister (Abbott) and at National Revenue, a new minister (McCann) and deputy minister (Brown). Brown, a banker and wartime “dollar-a-year man,” had served for 10 turbulent months that left him “desperately tired” and in need of a “long rest,” as he said when he walked out in October 1947.55 His resignation letter to Prime Minister King was brutally frank. National Revenue needed major repair. Since 1945, Can- adians whose incomes were easy to tax had been aggressively (even “ruthlessly”) pursued for taxes not paid during the war. There were large numbers of well-off people (how many, no one really knew) who evaded tax by keeping inadequate business accounts or deliberately deceptive ones. Attempts to catch up on the enormous backlog of assessment and audit had made for “arbitrariness” and “prejudice.” Staff were badly paid, and bad pay had made some assessors easily corruptible. The statute was a mess, jumbling together “gross” and “net” in its treatment of income. A “tendency to resistance and evasion” had flourished. “Legitimate” revenues were being lost. So—where to start? Making big changes to produce a new normal means heavy lifting, both in policy and in administration. In the post-war years, this was a slow, piecemeal process. Minister Ilsley and Deputy Minister Clark hoped that pride in the war effort and the endorsement of voters would rally united support for fiscal reform, against some premiers’ opposition to a more centralized regime.56 But support for unity was hard to find. Even in wartime, Canada had been troubled by divisions of many sorts—struggles between capital and labour, Quebec’s nationalism, hostility to aliens,

54 Access to current research and policy work may be found at the website of Child Care Canada (www.childcarecanada.org). 55 Telegram, Frank H. Brown to Prime Minister King, October 29, 1947, quoted in Armorer and Lemieux, supra note 13, at 222. Brown first attempted to resign in February 1947, only two months after starting work at National Revenue. Letter, Frank H. Brown to Prime Minister King, February 1947, LAC, MG 26-J1, vol. 420, at 381962-383018, quoted in Armorer and Lemieux, supra note 13, at 200-3. 56 Wardhaugh, supra note 19, at 291 and 310. In retrospect, Clark claimed that bolder attitudes based in pride in the war effort had supported economic development after the war: W.C. Clark, “Canada’s Postwar Finance” (1953) 43:2 American Economic Review 1-18. policy forum: crisis, cleanup, and long-term fiscal change n 1025

Canadians awaited post-war tax reform with more than the usual anxiety. Meeting their expectations exhausted one deputy minister and called heavily on the energies of new ministers in Finance and National Revenue. Source: Halifax Herald, April 28, 1947, at 6. and more. These conflicts had all been part of Canadian life before the war and would continue to shape the stakes in fiscal and social policy. Facing these realities, National Revenue and Finance made a series of decisions that helped to ease tensions and made the wartime mass taxpayers into contributors to the cost of the welfare state (and, in the early 1950s, the Korean War). In these decisions, we can see an attempt to ease the difficulties of compliance for the new income taxpayer. In our post-pandemic world, similar—or greater—carefulness will surely be required. The spirit of the new day was evident in Brown’s early policy memorandums to National Revenue’s network of tax administrators. For example, in March 1947, he wrote that the tax debts of armed forces personnel who died on active service should be cancelled, and people who owed small amounts of tax on small incomes should not have instalments on their tax debt taken out of their pay packets, except as a last resort. Brown reminded tax administrators that “many small taxpayers did not pay because they could not, and that it would work great personal hardship to attach their wages.”57 If tax debtors had compulsory savings refunds coming to them, the tax

57 Armorer and Lemieux, supra note 13, at 175-76. 1026 n canadian tax journal / revue fiscale canadienne (2020) 68:4 office should settle the debt from that source. In the same spirit, in 1951, the department would instruct payroll administrators to begin deducting tax instalments from paycheques at a rate that would, in most wage earners’ situations, ensure a refund rather than require a cash payment at tax-filing time. And a new income tax return form was introduced in 1948 that would suit the 87 percent of Canadians who had little or no investment income.58 Such measures eased the transition from war tax to normal tax. Less well managed was the transition from one of the wartime measures that Elliott had granted “in a rather grand manner”59 during the war, the deductibility of travel expenses for construction workers. After the war, this deduction (along with the deductibility of expenses more generally) remained a point of chronic tension. An attempt by National Revenue in 1956 to reverse the wartime measure provoked an uproar from which the Progressive Conservatives made electoral hay. On winning government in 1957, the Conservatives amended the Income Tax Act to make the temporary measure a permanent one.60 While revisions to the statute enacted in 1948 were extremely important for businesses and for wealthier taxpayers, making their tax planning and tax appeals more feasible, the smaller measures that tailored tax administration to the millions of “new” taxpayers recognized that the steps taken hastily during the war could be preserved and put to new use only if adjustments were made in National Revenue’s procedures and mindset. Frictions with taxpayers were unfortunately exacerbated in the case of farmers, a chronically hard-to-tax constituency. In the late 1940s, farmers found themselves singled out as tax dodgers and became the object of debate, help, and scrutiny. During the war, they had not been pressed to comply with the tax law, but on Brown’s watch, that changed. Tax administrators knew them to be careless keepers of accounts, if indeed they kept accounts at all. Ontario farmers might have remem- bered fondly the days, under the pre-1936 Ontario Assessment Act when farm income was off limits to the taxman.61 Questions were asked in the House, journal- ists remarked on the low tax yields from prosperous farm districts, and the Canadian Federation of Agriculture spoke in defence of the sector’s honour. At least one Liberal went down to defeat in a farming district in the 1949 federal election.62 Making new taxpayers for a new state was no trivial pursuit. In the case of farmers, it was full of both political and administrative trouble.

58 Ibid., at 190. 59 The characterization of Elliott’s style is Kenneth Eaton’s, in Memorandum to the Minister re: Canadian Construction Brief, Taxation of Allowances, July 18, 1957, LAC, RG 19, vol. 4247, file 4044-03-02. 60 Tillotson, supra note 5, at 224, 229-32, and 246-48. 61 Section 4.18 of the Assessment Act, RSO, 1927, c. 238, exempted from taxation “the income of a farmer derived from his farm.” 62 Armorer and Lemieux, supra note 13, at 176-81; Tillotson, supra note 5, at 213. policy forum: crisis, cleanup, and long-term fiscal change n 1027

Another profound post-war conflict around income taxation, made resonant by memories of conscription, was led by the government of Quebec. Clothed in the garb of constitutional correctness, Quebec’s position on fiscal federalism also defended elements of plain economic interest, whether on behalf of the prosperous farmers of the St. Lawrence River valley, the international investors in the mining sector, or the business community of English Montreal.63 The latter community’s tax concerns are illustrated by the case of Montrealer Frank Sura. Sura advanced the theory that half of his very substantial income for the years 1947-1954 was properly taxed in the hands of his wife. They had been married in 1929 under Quebec’s community- of-property regime. The case was decided against him in 1961, but there had been a moment of legal validation in 1957.64 Of equal duration, and fiercer for its foundation in Catholic social doctrine, was the economist F.A. Angers’s 10-year “squabble” with the tax authorities. Angers claimed that the pre-war reduction of taxable income by exempt amounts for each of his children was more consistent with the constitution and his freedom of choice than was the compulsory registration of his children for family allowances.65 Sura and Angers chose to make a noise; others engaged, more or less quietly, in surplus stripping and other methods of reducing the impact of high marginal rates.66 Feelings of unity had not materialized as Clark had hoped. I conclude that there was a twofold legacy of the wartime call to solidarity that had led to the progressive-rate mass income tax. On the one hand, some were now convinced that government could use its expanded fiscal power to make a better country for all. On the other hand, there were now worries (certainly, but not solely, among those facing an 84 percent tax rate on some of their income) that the tax burdens (with municipal and provincial taxes added) were too heavy and that the new tax regime was a threat to provincial autonomy.67 From 1945 to 1975, these hopes and worries animated prolonged political struggles over taxation and social provision. Most of the proposals for social security that had been outlined during the war were delivered, but not all at once and not

63 An approving account of Premier Duplessis’s defence of Quebec taxpayers, especially corporations, is found in Conrad Black, Duplessis (Toronto: McClelland and Stewart, 1977), chapter 14 and, with respect to corporations, at 428-31. 64 See Tillotson, “The Family as Tax Dodge,” supra note 29, at 420-22. 65 Jean-Philippe Carlos, “ ‘Ma chicane avec l’impôt’: une critique traditionaliste de l’État- providence canadien, 1945-57” (2020) 101:1 Canadian Historical Review 76-100. 66 See H. Heward Stikeman and Robert Couzin, “Surplus Stripping” (1995) 43:5 Canadian Tax Journal 1844-60, at 1850-51; and Richard W. Pound, Stikeman Elliott: The First Fifty Years (Montreal and Kingston, ON: McGill-Queen’s University Press, 2002), at 38-39. 67 See Alvin Finkel, Social Policy and Practice in Canada: A History (Waterloo, ON: Wilfrid Laurier Press, 2006), chapter 6; Greg Donaghy, Grit: The Life and Politics of Paul Martin, Sr. (Vancouver: UBC Press, 2015), at 85-86; and Tillotson, supra note 5, at 214, 240-50, and 263-73. 1028 n canadian tax journal / revue fiscale canadienne (2020) 68:4 completely. Around each of these measures—pensions, income assistance for the uninsured unemployed, unemployment benefits for married women, subsidies for prescription drugs, support for people with disabilities, hospital insurance, and medical insurance—there were competing economic interests and social narratives.68 To be sure, the mass income tax and a steeply graduated scale of rates (plus rising rates in the federal sales tax) made the public coffers invitingly full. But insurance firms, marginal employers struggling with structural economic change, doctors as business owners, and others ideologically opposed to what they called socialism fought against the spending of the enlarged revenues on expanded social security. Those fights were all the more intense because the post-war years were not simply “happy days” of affluence, but also the age of anxiety. Recall the foreign exchange crisis of 1947, the demands of the Korean War in 1950-1953, and the distressing recession in 1958-1962 (with the rumbling of the Cuban Revolution troubling the hemisphere). Remember that a 50-year-old in 1952 had been a 30-year-old in 1932.69 In this context, building a new income tax system, one connected at several points to major social transfers (the old age security program and the family allowances) and intimately linked to the long struggle over federalism, was not going to be easy. By the end of the 1950s, a host of new grievances and irritations had arisen. The battle over both income taxation and social spending, simmering through the 1950s and early 1960s, would explode in 1968 to 1975 as both medicare and capital gains taxation changed the nation’s fiscal landscape. The intoxicating dreams of post-war reconstruction were replaced, in the daylight of normal life, with the more sobering work of negotiation and compromise.

68 See James Struthers, “Building a Culture of Retirement: Class, Politics and Pensions in Post-World War II Ontario” (1997) 8:1 Journal of the Canadian Historical Association 259-82; James Struthers, “Shadows from the Thirties: The Federal Government and Unemployment Assistance, 1941-1956,” in Jacqueline S. Ismael, ed., The Canadian Welfare State: Evolution and Transition (Edmonton: University of Alberta Press, 1987), 3-32; Dustin Galer, Working Towards Equity: Disability Rights Activism and Employment in Late Twentieth-Century Canada (Toronto: University of Toronto Press, 2018), chapters 2 and 3; Ann Porter, “Women and Income Security in the Post-War Period: The Case of Unemployment Insurance, 1945-1962” (1993) 31 Labour/Le Travail 111-44; P.E. Bryden, “The Liberal Party and the Achievement of National Medicare,” in Gregory P. Marchildon, ed., Making Medicare: New Perspectives on the History of Medicare in Canada (Toronto: University of Toronto Press, 2012), 71-88; Gregory P. Marchildon and Nicole C. O’Byrne, “From Bennettcare to Medicare: The Morphing of Medical Care Insurance in British Columbia,” ibid., 207-28; and C. David Naylor, Private Practice, Public Payment: Canadian Medicine and the Politics of Health Insurance, 1911-1966 (Montreal and Kingston, ON: McGill-Queen’s University Press, 1986). For further discussion of issues around the medical expenses deduction, see Tillotson, supra note 5, at 245-48. 69 The tax economist Kenneth Eaton referred specifically to the impact of the Depression on Deputy Minister of Finance Clark’s policy orientation in A. Kenneth Eaton, Essays in Taxation, Canadian Tax Paper no. 44 (Toronto: Canadian Tax Foundation, 1966), at 134, quoted in Wardhaugh, supra note 19, at 264. policy forum: crisis, cleanup, and long-term fiscal change n 1029

It was said during the Second World War, and it is said now, that we face an existential crisis threatening our way of life, and that our response must express our solidarity—“We’re all in this together.” But then, as now, the pre-crisis conflicts were not much mitigated by the crisis. In some ways, they were, and are, exacerbated. After the crisis, they come roaring back. So let us not escape into technocratic or moral dreamland. Just as plans based on old ideas change with the pressure of events, the hopeful projects of a new normal demand extraordinary efforts in organizing for influence, negotiating institutional change, and designing new means of administration. In closing, I want to emphasize that the detailed work of easing compliance burdens is at least as important as are big ideas in transitioning to a major new program. The small troubles that poor administration inflicts on millions of citizens harm the perception of government writ large. One risk of this kind of harm might lie in our near future. Ordinarily strict enforcement of small tax liabilities or minor overpayments that result from the CERB deployment would almost certainly have a perverse result: instead of convincing Canadians that government was being scrupulously careful to avoid waste, it would likely put in question the ethical judgment that directs state agencies. Decision makers may well form policies that reorganize our fiscal institutions in broad ways to support economic resilience, usher in prosperity, and remedy structural injustice. But they must not lose sight of the nuts and bolts of management. Otherwise, people without much power will get trampled. And that experience has won and will win supporters for populist anti-statism.70 Readers of this historical sketch will perhaps have recognized more parallels between the past and the present than those I have explicitly drawn out. In a fuller account, the resonances would multiply. Canada has its chronic policy questions, as does modern politics more generally. Continuities may be a comfort to anyone frustrated with policy making and politics in the present. It can be reassuring to know that we struggle where others, similarly flawed, also did their best. There was no race of giants in the past, always guided by reason, sailing effortlessly toward easy consensus. We will not know for some time yet whether we have moved into a post- COVID-19 crisis world. The international impact of the virus, which I have not even touched on in this piece, may have a more powerful influence on our policy future than any of the choices that I have drawn attention to here. This is an uncertain time, and I cannot honestly offer reassurances. The post-war world, prosperous though it was in North America, was not a golden age, and we are unlikely to be entering one now. We must simply do what we can, as circumstances permit, to make things better.

70 For the importance of income tax grievances in building support for a conservative populism in 1950s and 1960s Quebec, see Maurice Pinard, The Rise of a Third Party: A Study in Crisis Politics (Englewood Cliffs, NJ: Prentice-Hall, 1971), at 112-13. canadian tax journal / revue fiscale canadienne (2020) 68:4, 1031 - 33 https://doi.org/10.32721/ctj.2020.68.4.pf.editors

Policy Forum: Editors’ Introduction— The GST/HST Responsibilities of Non-Resident E-Commerce Firms

Update As this journal was about to go to press, the federal government announced proposals for tax changes that are relevant to the discussion in this introduction and the articles following. In the Fall Economic Statement delivered by the minister of finance on November 30, 2020, the government proposed that “non-resident vendors supplying digital products or services (including traditional services) to consumers in Canada be required to register for the GST/ HST [goods and services tax/harmonized sales tax] and to collect and remit the tax on their taxable supplies to Canadian consumers.” Further, to take into account the fact that these vendors may also make supplies indirectly through digital platforms, the government proposed that operators of such platforms “be generally required to register for the GST/HST and to collect and remit the tax on the supplies” facilitated by these platforms. Details of the proposals can be found in the Fall Economic Statement (Annex 4—Tax Measures: Supplementary Information), November 30, 2020, available on the Department of Finance website.

When the goods and services tax (GST) was introduced in Canada in 1991, consumers were generally buying goods and services from suppliers with a physical presence in Canada. There was minimal concern at that time that foreign firms without bricks- and-mortar operations in Canada would become a significant part of the market. Thus, Canadian firms were not thinking that they could be put at a competitive disadvantage because of the non-collection of GST on foreign firms’ sales. Yet, in that same year—1991—a development occurred that laid the groundwork for exactly this result: the principal financial sponsor of the Internet, the National Science Foundation in the United States, changed its acceptable-use policy to allow commercial traffic.1 Since that time, the vast expansion of e-commerce has led to the creation of enterprises such as Netflix, Spotify, Uber, Airbnb, and Stubhub, which, it is generally believed, are not remitting GST or, where applicable, harmonized sales tax (HST) because they do not consider themselves to be carrying on business in Canada. It is estimated that potential GST/HST collections from such firms’ sales in 2016 ranged between $86 million and $107 million.2

1 National Science Foundation, “The Internet” (www.nsf.gov/od/lpa/nsf50/nsfoutreach/htm/ n50_z2/pages_z3/28_pg.htm#answer3). 2 David Bradbury, Allison Christians, Elinore Richardson, and Rosalie Wyonch, “Digital Taxation,” in Report of Proceedings of the Seventieth Tax Conference, 2018 Conference Report (Toronto: Canadian Tax Foundation, 2019), 25:1 - 26, at 25:20.

1031 1032 n canadian tax journal / revue fiscale canadienne (2020) 68:4

A significant concern with any attempt to impose tax obligations on non-resident firms is whether the targeted taxpayers will choose to comply. However, many jurisdictions have found that this is not a problem, perhaps because firms wish to protect their reputations.3 In Canada, Quebec has led the way in requiring non-resident businesses to remit Quebec sales tax on transactions in that province, and some action has been taken by British Columbia and Saskatchewan.4 Thus, the key issue appears to be whether there is the political will to impose such obligations, particularly since the foreign firms involved can be expected to pass on at least part of the tax increase to Canadian consumers. The policy of the current federal government is that it will not be making changes to the GST/HST that would result in Canadians paying more tax on their purchases of digital goods and services. The political cost to a government of imposing a tax increase may be expected to be higher if doing so requires a change in the law as opposed to a change in the tax authority’s assessing policies. Thus, it is important to know whether the current law requires that non-resident e-commerce firms register forGST /HST, collect GST/ HST from their Canadian customers, and remit the tax to the government. In the first of three contributions to this Policy Forum, Nicholas Shatalow makes the argument that these obligations may already exist in the current law. Spe- cifically, he argues that a firm that makes taxable supplies in the course of a business carried on in Canada is subject to the GST/HST requirements even if the firm has no physical presence in Canada. Thus, the key question is whether an e-commerce firm may be considered to be carrying on business in Canada by virtue of the nature of its operations here. Although there are no tax cases specifically on point, Shatalow finds that a non-tax decision of the British Columbia Court of Appeal, Equustek Solutions Inc. v. Google Inc.,5 and the associated lower-court decision,6 make this likely. In that case, which concerned an intellectual property dispute, Google was found to be digitally carrying on business in British Columbia because it gathered information about its BC users through their interactions with the company and used that information to enter into advertising contracts with Canadian-resident firms targeting specific advertisements to those users. Shatalow acknowledges that there is a great deal of uncertainty about his conclusion on the carrying-on-business question. Thus, he recommends that Parliament should remove this uncertainty by amending the Excise Tax Act7 to state that non- resident firms with Canadian sales that exceed a certain threshold should be subject to GST/HST obligations. Further, since some small suppliers may sell their goods or

3 Danny Cisterna and Jan Pedder, “Indirect Tax and E-Commerce,” in 2019 Corporate Tax Conference (Toronto: Canadian Tax Foundation, 2019), 3:1 - 29, at 3:27. 4 Shahrukh Khowaja, “Taking Stock of ‘Netflix Taxes’: British Columbia and Beyond” (2020) 10:3 Canadian Tax Focus 8. 5 2015 BCCA 265. 6 Equustek Solutions Inc. v. Jack, 2014 BCSC 1063. 7 RSC 1985, c. E- 15, as amended. policy forum: editors’ introduction n 1033 services through a market facilitator such as Amazon, those obligations should also apply to all sales made through a facilitator that itself exceeds the sales threshold. In response to Shatalow’s article, Zvi Halpern-Shavim contends that the context of the Equustek case is too far removed from the GST/HST context to assist in resolving the carrying-on-business issue. The reasons for his conclusion are drawn from the fact that the Excise Tax Act is a tax statute, whereas Equustek arose in the non-tax setting of conflict of laws. Halpern-Shavim applies a textual, contextual, and purposive analysis of the phrase “carrying on business” and its permutations, and finds that this points away from applying to a non-resident with no physical presence in Canada. He also notes that the provinces that have decided to impose sales tax collection obligations on non-residents have felt the need to create new legal terminology for this purpose. Finally, Halpern-Shavim argues that it is reasonable, on the basis of the tiebreaker rule, that the interpretive dispute should be decided in favour of the taxpayer. In a separate contribution to the understanding of this general area of law, Malcolm Gammie provides a brief review of the approach used by the United King- dom to address issues associated with the recovery of tax on transactions entered into by UK persons with suppliers based in other countries. This perspective is helpful given that the carrying-on-business test in Canada is borrowed from UK law, and UK jurisprudence continues to be cited when this issue arises in Canadian cases. Gammie first discusses the policy principles underlying the UK approach, noting that no state can tolerate a situation in which its residents can avoid the taxes normally imposed on transactions within its jurisdiction by instead transacting with non-residents. He then reviews the common-law jurisdictional rules and surveys their application to business profits tax (income tax) and value-added VA tax( T). Gammie concludes that different taxes require different solutions. For income tax, it is reasonable to look for a sufficient UK presence by the foreign party; however, for VAT, special rules are needed to ensure that tax cannot be avoided through remote supplies. By requiring that VAT be paid in the jurisdiction in which the consumer is using the supply of goods or services, the rules support a key objective of the UK VAT system (and the EU system within which it continues to operate), namely, the creation of a level playing field for domestic and foreign suppliers. Alan Macnaughton Daniel Sandler Editors canadian tax journal / revue fiscale canadienne (2020) 68:4, 1035 - 52 https://doi.org/10.32721/ctj.2020.68.4.pf.shatalow

Policy Forum: The GST/HST Obligations of Non-Resident E-Commerce Firms— Jurisprudence and Policy

Nicholas Shatalow*

Editors’ note As this journal was about to go to press, the federal government announced proposals for tax changes that are relevant to the discussion in this article. In the Fall Economic Statement delivered by the minister of finance on November 30, 2020, the government proposed that “non-resident vendors supplying digital products or services (including traditional services) to consumers in Canada be required to register for the GST/HST [goods and services tax/ harmonized sales tax] and to collect and remit the tax on their taxable supplies to Canadian consumers.” Further, to take into account the fact that these vendors may also make supplies indirectly through digital platforms, the government proposed that operators of such platforms “be generally required to register for the GST/HST and to collect and remit the tax on the supplies” facilitated by these platforms. Details of the proposals can be found in the Fall Economic Statement (Annex 4—Tax Measures: Supplementary Information), November 30, 2020, available on the Department of Finance website.

PRÉCIS Le commerce électronique présente un large éventail de défis pour les décisionnaires politiques. Dans le contexte fiscal, il pourrait occulter la relation entre une entreprise et la juridiction fiscale appropriée. Cet article examine les obligations en matière de taxe sur les produits et services (TPS)/taxe de vente harmonisée (TVH) des entreprises de commerce électronique non résidentes qui n’ont pas de présence physique au Canada — en particulier, si ces entreprises peuvent être tenues de s’inscrire au régime de la TPS/TVH, ainsi que de percevoir et verser ces taxes, pour leurs activités sur le marché canadien. L’évolution récente de la common law (dans un autre contexte, cependant, que la fiscalité) indique que les entreprises de commerce électronique non résidentes n’ayant pas de présence physique au Canada peuvent être considérées comme exploitant une entreprise au Canada. Puisque l’exploitation d’une entreprise au Canada

* Master of laws candidate (business law), Faculty of Law, University of Toronto (e-mail: [email protected]). I would like to thank Simon Thang for introducing me to this topic and Sara Truuvert for providing feedback during the editing process.

1035 1036 n canadian tax journal / revue fiscale canadienne (2020) 68:4 est un élément clé pour déterminer les obligations d’une personne en matière de TPS/ TVH, il est probable que les entreprises de commerce électronique non résidentes qui remplissent cette condition seront assujetties à de telles obligations pour les fournitures taxables effectuées sous forme numérique sur le marché canadien. Cela dit, il existe actuellement une grande incertitude quant aux facteurs à utiliser pour déterminer si les activités d’une entreprise de commerce électronique non résidente sont suffisantes pour atteindre le seuil d’exploitation d’une entreprise. L’auteur de cet article recommande au Parlement de remédier à cette incertitude en modifiant la Loi sur la taxe d’accise afin que les entreprises de commerce électronique non résidentes soient réputées exploiter une entreprise au Canada si leur chiffre d’affaires dépasse des seuils de minimis donnés.

ABSTRACT E-commerce presents a broad range of challenges for policy makers. In the tax context, it has the potential to obscure the relationship between a firm and the appropriate taxing jurisdiction. This article explores the goods and services tax (GST)/harmonized sales tax (HST) obligations of non-resident e-commerce firms with no physical presence in Canada—specifically, whether those firms may be required to register for, collect, and remit GST/HST in respect of their activities in the Canadian market. Recent developments in the common law (albeit in a non-tax context) indicate that non-resident e-commerce firms with no physical presence in Canada may be found to be carrying on business in Canada. Since carrying on business in Canada is a key element in determining a person’s GST/HST obligations, it is likely that non-resident e-commerce firms that are found to meet the requirement will be subject to such obligations in respect of taxable supplies made digitally in the Canadian market. That said, there is at present a great deal of uncertainty about the specific factors to be used in determining whether the activities of a non-resident e-commerce firm are sufficient to meet the carrying-on-business threshold. The author of this article recommends that Parliament should address this uncertainty by amending the Excise Tax Act to deem non-resident e-commerce firms to be carrying on business in Canada if they exceed specific de minimis sales thresholds. KEYWORDS: GOODS AND SERVICES TAX n HARMONIZED SALES TAX n ELECTRONIC COMMERCE n NON-RESIDENTS n CARRYING ON BUSINESS n DIGITAL

CONTENTS Introduction 1037 Statutory and Legal Context 1037 The Statutory Scheme 1037 The Carrying-On-Business Concept 1039 Common-Law Factors 1040 Equustek 1040 Overview 1040 Carrying On Business Digitally 1041 Gaps in the Analysis 1043 Applying Equustek 1044 Summary 1045 Van Breda 1045 Overview 1045 policy forum: gst/hst obligations of non-resident e-commerce firms n 1037

“Actual, Not Only Virtual, Presence” 1046 Summary 1048 Recommendations 1048 A “Carrying On Business Digitally” Deeming Rule 1049 De Minimis Thresholds 1049 Market Facilitator Rules 1051 Conclusion 1051

INTRODUCTION The modern economy presents many challenges for Canada’s tax system that could not have been anticipated when the system was conceived. In the goods and services tax (GST)/harmonized sales tax (HST) context, for example, the obligations of non-resident e-commerce firms operating in the Canadian market are at present unclear. The GST/HST obligations of non-residents generally depend on whether the non-residents in question carry on business in Canada. In recent years, Canadian courts have explored the carrying-on-business concept in respect of digital operations in ways that likely allow for non-resident e-commerce firms to be found to be present in Canada for GST/HST purposes. This article offers a summary of Canada’s carrying-on-business jurisprudence as it relates to the GST/HST obligations of non- resident e-commerce firms and a proposal for how Parliament could amend the Excise Tax Act1 to provide greater certainty on this matter. Before proceeding, it is helpful to clarify the focus of this article. The article seeks to address the question of whether non-resident e-commerce firms that have no physical presence in Canada and conduct their activities in the Canadian market substantially online are required to register for, collect, and remit GST/HST in respect of their supplies made in Canada. The article does not address any other potential tax obligations, such as income tax or provincial sales tax, to which these firms may also be subject. Additionally, it does not consider the administrative policy of the Canada Revenue Agency (CRA), which may use different factors than those established in the common law to assess whether a firm is carrying on business in Canada.

STATUTORY AND LEGAL CONTEXT The Statutory Scheme For taxable supplies “made in Canada,”2 the ETA requires “every person who makes a taxable supply” to collect the GST/HST payable on that supply.3 This obligation is

1 RSC 1985, c. E- 15, as amended (herein referred to as “the ETA”). 2 ETA section 142. 3 ETA subsection 221(1). This collection obligation includes HST payable under section 220.08(1) in respect of supplies of intangibles or services made in a participating province. 1038 n canadian tax journal / revue fiscale canadienne (2020) 68:4 subject to a non-resident override rule,4 which states that a supply made by a non- resident will be deemed to be made outside Canada unless the non-resident carries on business in Canada.5 The GST/HST obligations of non-resident suppliers therefore hinge on two issues: first, whether the non-resident makes taxable supplies in Canada; and second, whether the non-resident carries on business in Canada. E-commerce complicates both of these issues. Although the second issue is the primary focus of this article, the first issue merits brief consideration here. TheETA has different place-of-supply rules for tangible personal property, intangible personal property, and services.6 These rules are potentially problematic in the e-commerce context because e-commerce “blurs the lines between categories.”7 It may not be immediately obvious whether some e-commerce supplies constitute supplies of intangible personal property or supplies of services. Supplies of intangible personal property are deemed to be made in Canada if the property in question “may be used in whole or in part in Canada.”8 Supplies of services are deemed to be made in Canada if the service in question is (or is to be) “performed in whole or in part in Canada.”9 The difference between the availability of property and the performance of a service may therefore be highly consequential to a supply’s ultimate GST/ HST treatment. The actual potency of this issue is likely to vary on a case-by-case basis, but it may nonetheless be relevant for determining the GST/HST obligations of some non-resident e-commerce firms. It is also important to emphasize that the burden for non-residents under this scheme is not the imposition of tax, but the cost of administrative compliance.10 Non-residents that are required to register for the GST/HST, or that voluntarily decide to do so, must generally give and maintain security with the minister of national revenue if they do not have a permanent establishment in Canada.11 To the extent that there is uncertainty about a non-resident’s obligation to register for the GST/HST, the non-resident in question will have to assess its activities in the Canadian market and decide whether to register or not—at the risk of scrutiny from the CRA.

4 Steven D’Arcy, “Carrying On Business in Canada,” in 2005 CPA Canada Commodity Tax Symposium (Toronto: Chartered Professional Accountants of Canada, 2005), paper 24, at paragraph 10. 5 ETA section 143. 6 ETA subsection 142(1). 7 Mike Nienhuis, “Tax Implications for Non-Residents Conducting E-Commerce in Canada” (2011) 9:1 Canadian Journal of Law and Technology 235 - 75, at 257. 8 ETA subparagraph 142(1)(c)(i). 9 ETA paragraph 142(1)(g). 10 Nienhuis, supra note 7, at 237. 11 ETA subsection 240(6). policy forum: gst/hst obligations of non-resident e-commerce firms n 1039

The Carrying-On-Business Concept “Carrying on business” is a private-law concept that has been embedded in Canada’s tax legislation—notably, in the context of this article, section 143 of the ETA. Whether a non-resident carries on business in Canada is a question of fact.12 Early jurisprudence established a distinction between carrying on business with a jurisdiction and carrying on business in a jurisdiction. The latter requires a firm to have “something more” to connect it to the jurisdiction than the former.13 In the tax context, the carrying-on-business concept is used as a tool for determining whether a non-resident’s activities in the Canadian market are sufficiently involved that the non-resident should be subject to the Canadian tax system. In explaining the carrying-on-business concept, it is helpful to consider its origins in late 19th-century British case law. In Werle v. Colquhoun, the Lord Esher explained how courts determine whether a firm carries on business in England:

It is a question of fact which is divided into two. Is there a trade carried on, and, if so, is that trade carried on in England? It is a question of fact in each case. Well, if it is a question of fact in each case it will be impossible to make an exhaustive exposition of the facts which will constitute a trade. The question in each case must be, do the facts which are shown to exist in this particular case amount to a carrying on of a trade and a carrying on of a trade in England?14

The Lord Esher’s words still hold true today in the Canadian context.15 The carrying- on-business concept simply prompts courts to examine and articulate, on the basis of the particular facts of the case, whether a firm has sufficient commercial activity in a jurisdiction that it can be said that the firm does business there. Importantly, the carrying-on-business analysis does not ask courts to employ a hard-and-fast check- list of indicia for jurisdictional presence. This distinction is relevant for this article, since the lack of such indicia means that the application of the carrying-on-business concept to novel fact situations, such as e-commerce, cannot be rejected on its face. That said, courts have identified specific factors that are used to inform the analysis of whether a firm carries on business in Canada.16

12 D’Arcy, supra note 4, at paragraph 8. 13 Constantine A. Kyres, “Carrying On Business in Canada” (1995) 43:5 Canadian Tax Journal 1629 - 71, at 1632. 14 (1888), 2 TC 402, at 408 (CA) (emphasis added). 15 Kyres, supra note 13, at 1632. 16 Ibid., at 1633. 1040 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Common-Law Factors Courts have established a number of factors to be considered when determining whether a firm carries on business in Canada. Often, no single factor is determinative.17 The recognized factors include the following:18

n the place where the commercial contracts are made; n the place where the profits arise; n the place of delivery; n the place of payment; n the place where purchases are made; n the place of production; n the place where transactions are solicited; n the place where goods are located; n the place where bank accounts are located; and n the place where the business is listed in a directory.

In theory, several of these factors, such as the place of profit generation, the place of delivery, and the place of purchase, could be interpreted to capture e-commerce activity. However, until recently, Canadian jurisprudence had not applied the carrying-on-business concept to e-commerce fact situations. Although the Supreme Court of Canada has yet to address whether the carrying-on-business concept (and the factors that embody it) applies to e-commerce fact situations, lower courts have started to develop such jurisprudence. The next section will examine that jurisprudence with reference to the trial decision in Equustek Solutions Inc. v. Jack19 and Equustek Solutions Inc. v. Google Inc., the appeal decision.20 I will then consider that jurisprudence in the context of the decision of the Supreme Court of Canada in Club Resorts Ltd. v. Van Breda,21 which sets out that court’s only comments to date about the viability of a Canadian doctrine of carrying on business digitally.

EQUUSTEK Overview In Equustek, the Supreme Court of British Columbia found that the non-resident e-commerce firm Google carried on business in British Columbia by virtue of its interactive online presence in the province.22 Before I discuss Equustek in detail, it

17 Ibid., at 1642. 18 Ibid., at 1634 - 46. The place of contract and the place of profit are the two dominant factors in Canada’s carrying-on-business jurisprudence. See D’Arcy, supra note 4, at paragraph 25. 19 2014 BCSC 1063. 20 2015 BCCA 265. 21 2012 SCC 17. 22 Equustek, supra note 19, at paragraphs 47 - 48. policy forum: gst/hst obligations of non-resident e-commerce firms n 1041 is prudent to note that the case involved issues arising in the conflict-of-laws context, not the tax context. In conflict-of-laws jurisprudence, the carrying-on-business concept is used as a presumptive connecting factor to link a defendant to a jurisdiction for private-law actions.23 Importantly, this conflict-of-laws analysis does not alter the carrying-on-business concept, which is a general private-law concept, or the process of determining whether a firm carries on business in Canada, which is a finding of fact. Consequently, it is appropriate to transplant the reasoning in Equustek to the tax context, because taxation law is accessory to private law, and because the determination of whether the non-resident in question is carrying on business is the same in both contexts. Equustek was one of several related cases that culminated in the decision of the Supreme Court of Canada in Google Inc. v. Equustek Solutions Inc.24 The Supreme Court acknowledged in that decision that Google carried on business in British Columbia, but it did not undertake its own analysis of the issue because Google did not contest the finding of the BC Court of Appeal on this aspect of the case.25 The BC Court of Appeal had mainly deferred to the analysis of the BC Supreme Court on this matter.26 Accordingly, my discussion of the Equustek case will refer primarily to the BC Supreme Court’s decision, along with occasional reference to the decision of the BC Court of Appeal. Equustek concerned an intellectual property dispute to which Google was initially not a party.27 The plaintiff (“Equustek”) sought a court order preventing Google from indexing websites that Equustek alleged were being used by the defendants to unlawfully share its property. The case centred on whether BC courts were an appropriate venue for hearing the application and for enforcing such an order against Google.28 For the purposes of this article, the trial and appeal decisions in Equustek are important because they stand for the principle that non-residents with no physical presence in Canada can be found to carry on business in Canada through their digital activities.29

Carrying On Business Digitally In determining that Google carried on business in British Columbia, the BC Supreme Court characterized Google’s business presence in the province as “active” and

23 Sophie Stoyan, “Just a Click Away? Jurisdiction and Virtually Carrying On Business in Canada” (2017) 13:3 Journal of Private International Law 602 - 32, at 604 - 5 (https://doi.org/10.1080/ 17441048.2017.1387681). See also Van Breda, supra note 21, at paragraph 75. 24 2017 SCC 34. 25 Equustek, supra note 20, at paragraph 52. 26 Ibid., at paragraph 52. 27 Equustek, supra note 19, at paragraphs 26 - 27 and 161. 28 Ibid., at paragraph 105. 29 Ibid., at paragraph 51. Stoyan, supra note 23, at 616. 1042 n canadian tax journal / revue fiscale canadienne (2020) 68:4

“interactive” rather than “passive.”30 It appears that the court used these terms with a view to incorporating the US sliding-scale test31 for determining online jurisdiction into Canadian law.32 However, the court did not define these terms in its judgment. The court found that Google performed two services in British Colum- bia: advertising services and search engine services.33 According to the court, the two services were intrinsically linked because Google collected data from BC users of its search engine and then used those data to create targeted advertisements.34 In performing its advertising services, Google also entered into advertising contracts with BC residents.35 The court did not formally tether its carrying-on-business analysis to any of the common-law factors mentioned above. However, its analysis arguably evokes profit generation as a factor in this determination. Although the court did not explicitly consider Google’s profit in its analysis, profit generation is implicit in the court’s finding. In particular, the court’s emphasis on the interaction between Google and its BC users resembles a profit-generation analysis because that interaction was a key component of Google’s profit-making apparatus. Google’s activities in theBC market, facilitated through online interaction with its BC users, arguably indicate that the company’s profit-making apparatus was at least partly located in British Col- umbia because persons in that province had substantial interactive access to that apparatus.36 Phrased differently, Google’s profits in respect of its BC activities were intrinsically linked to British Columbia because they could not have arisen without the input37 of its BC users. In summary, it appears that the court may have used the

30 Equustek, supra note 19, at paragraphs 35, 45, 47 - 48, and 51. See also Stoyan, supra note 23, at 614 - 15. 31 The test in question is from Zippo Mfg. Co. v. Zippo Dot Com, Inc., 952 F Supp. 119 (WD Pa. 1997), a case that, according to the BC Supreme Court, has been “widely considered” by Canadian courts (Equustek, supra note 19, at paragraph 42). Zippo established a three-part sliding scale for assessing digital presence. The strongest presence on the scale is an “active” presence, indicating that the firm is doing business in the jurisdiction. A “passive” presence, at the other end of the scale, indicates that the firm is not carrying on digital business activities in the jurisdiction but is merely making information available online. An “interactive” presence, in the middle of the scale, is assessed on a case-by-case basis. See also Stoyan, supra note 23, at 614 - 15. 32 Equustek, supra note 19, at paragraphs 42 - 46. See also Stoyan, supra note 23, at 615. It remains unclear whether the American sliding-scale test for online jurisdiction is accepted in Canadian law. 33 Equustek, supra note 19, at paragraph 50. 34 Ibid., at paragraphs 54, 60, and 63. 35 Ibid., at paragraph 51. 36 D’Arcy, supra note 4, at paragraph 38. A profit-making apparatus can be located in more than one jurisdiction. 37 The input in this case is the users’ data collected to inform Google’s advertising, from which Google derived profits. policy forum: gst/hst obligations of non-resident e-commerce firms n 1043 concept of interactivity as a way of describing the nexus that Google and its profit- making apparatus have to Canada, though it did not state this explicitly in its judgment. The BC Court of Appeal did not comment on the lower court’s interactivity analysis, thus making its status at common law uncertain.38 Instead, the BC Court of Appeal found that Google carried on business in British Columbia by virtue of its data collection activities, which the court characterized as an “active process of obtaining data that resides in the Province or is the property of individuals in British Columbia.”39 That said, the analysis in the appeal decision may still be read through the lens of profit generation. It can be argued that the BC Court of Appeal found Google to be carrying on business in British Columbia because Google’s profit- making apparatus materially involved obtaining property that was located in, or was owned by individuals physically located in, British Columbia. In the conflict-of-laws context, the discrepancies between the trial and the appeal decisions in Equustek may be consequential, since they could affect the standing of the US sliding-scale test for determining online jurisdiction at common law. In the tax context, however, these discrepancies are less important, since the non-resident override rule is concerned with whether, and not how, non-residents carry on business in Canada. Accordingly, the fact that two courts have found a non-resident e-commerce firm to be carrying on business in Canada without having any physical presence in Canada is noteworthy for potential GST/HST registrants.

Gaps in the Analysis It is difficult to extract definitive takeaways from Equustek because the two courts ostensibly employed different (though not necessarily contradictory) reasoning and left several important concepts undefined. For example, the BC Supreme Court’s interactivity analysis contains a great deal of ambiguity. On a high level, it is unclear whether interactivity is meant to be a new, independent carrying-on-business factor or is merely a new way of describing existing factors. As mentioned above, the court’s interactivity analysis may simply be a novel way of articulating profit generation as a carrying-on-business factor, but this proposition is open to question. Similar concerns arise from the BC Court of Appeal’s emphasis on data collection, since “it is unclear whether ‘data collection’ is meant to bolster the interactivity analysis or if it is a stand-alone consideration for whether a defendant is virtually carrying on business.”40 The decision of the lower court is not helpful in this regard since that court did not discuss data collection in similar detail. Argu- ably, the BC Court of Appeal found data collection to be an important factor because Google collected data from its users in British Columbia for a commercial purpose. Unfortunately, the court did not explicitly refer to the commercial character of

38 Stoyan, supra note 23, at 616 and 619 - 20. 39 Equustek, supra note 20, at paragraph 54. 40 Stoyan, supra note 23, at 620. 1044 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Google’s data collection as a factor in its analysis. However, given the necessarily commercial nature of the carrying-on-business concept, it is possible that the court felt that this point was obvious and not worth mentioning. Moreover, it is unclear which aspects of Google’s business activities—such as data collection, auto-suggestions, or advertising contracts—had the greatest impact on the courts’ finding that Google carried on business in British Columbia. This lack of clarity makes it difficult to predict how the reasoning in Equustek will apply to e-commerce firms whose businesses involve a different balance of activities than Google’s. Since the question of whether a firm carries on business in Canada is a question of fact, a finding by a lower court that a firm is carrying on business in Canada through its digital activities is entitled to a great deal of deference from higher courts.41 Thus, one potential consequence of Equustek is that some non- resident e-commerce firms may be considered to be subject to theGST /HST requirements set out in the ETA even though their digital presence in Canada represents a relatively small share of their total business activities.

Applying Equustek In this section, I will attempt to infer how the reasoning from Equustek could apply beyond the facts at issue in those cases. Although the BC Supreme Court’s discussion of interactivity may, in theory, offer broad grounds to find firms to be present in Canada, that reasoning is arguably harder to apply to different fact situations than is the reasoning of the BC Court of Appeal. That said, Equustek clearly suggests that firms whose activities in the Canadian market are active and interactive may be found to be carrying on business in Canada, especially if those firms also enter into advertising contracts with Canadian residents. Since “active” and “interactive” are broad terms, which were left undefined by the BC Supreme Court, it is difficult to extract more concrete guidance from that court’s decision. The BC Court of Appeal’s discussion of data collection appears to offer grounds for firms to be found to be carrying on business in Canada by virtue of a wide range of digital activities. The court stated that Google’s “active process of obtaining data that resides in the Province or is the property of individuals in British Columbia” constituted carrying on business in British Columbia, in part because such data collection was a key element of Google’s business.42 Importantly, the court made this finding despite the fact that Google’s business in British Columbia represented only a small portion of its total business activity.43 Since the court noted that Google’s data collection activities constituted a separate basis for finding that it carried on business in British Columbia, distinct from its advertising contracts,44 it appears

41 Housen v. Nikolaisen, 2002 SCC 33, at paragraph 10. 42 Equustek, supra note 20, at paragraph 54. 43 Stoyan, supra note 23, at 620. As noted above, the court’s reasoning on this point suggests that the court was undertaking a profit-generation analysis. 44 Equustek, supra note 20, at paragraph 54. policy forum: gst/hst obligations of non-resident e-commerce firms n 1045 that, under the court’s reasoning, firms whose only commercial activities in the Can- adian market are data collection may be found to be carrying on business in Canada. Take, for example, a social media platform. Assume that, like Google, the platform does not charge users a fee to access the platform. The platform collects data from its users. The data are then used in advertising targeted at those users, and the platform earns revenue from the advertisements. If the platform’s Canadian data collection activities are a key part of its business, even if those activities represent a small fraction of its worldwide business activities, the appeal decision in Equustek suggests that the platform would likely be found to be carrying on business in Can- ada. Moreover, if the platform’s activities in the Canadian market can be described as active and interactive, the trial decision provides further support for the proposition that the platform carries on business in Canada. The BC courts’ finding that Google carried on business in Canada is ultimately quite broad, because Google did not receive payments directly from its users. E-commerce firms that engage in similar activities to Google’s (whether through interactivity or data collection) and also receive payments from Canadian users arguably have a greater nexus to Canada on the basis that their profit-making apparatus more strongly involves or is more squarely located in Canada. Those firms should be aware of the decisions inEquustek , particularly if they are targeting consumers in British Columbia.

Summary The impact of Equustek on Canada’s carrying-on-business jurisprudence is at present uncertain. Since the determination of carrying on business is a question of fact (which, in this case, was not analyzed beyond the decision of the BC Court of Appeal),45 it is difficult to know whether or how Canadian courts will employ the reasoning in Equustek to future e-commerce cases. Importantly, Equustek demonstrates that non- resident e-commerce firms can be found to be carrying on business in Canada, at least under BC law, without having any physical presence in the country. Of the many questions raised by this case, perhaps the most important is that of the specific threshold required for a non-resident e-commerce firm to be found to be carrying on business in Canada. That is, if a non-resident firm has a digital presence in Canada, how involved must that firm be in the Canadian market for its business to be considered to be carried on in Canada?

VAN BREDA Overview While Equustek indicates that non-resident e-commerce firms may be found to be carrying on business in Canada by virtue of their digital activities, the decision of the Supreme Court of Canada in Van Breda46 seems to limit this proposition, at least

45 Stoyan, supra note 23, at 620. 46 Van Breda, supra note 21. 1046 n canadian tax journal / revue fiscale canadienne (2020) 68:4 on a superficial reading. Van Breda appears limiting because it provides a brief discussion of digital activities that do not amount to carrying on business in Canada. Although Van Breda actually preceded Equustek, it is important to consider because it contains the Supreme Court’s most direct comments to date about carrying on business in the digital context—notably, that carrying on business requires “actual, not only virtual, presence.”47 On the surface, the court’s comments seem to preclude a viable doctrine of carrying on business digitally. However, closer scrutiny of the decision reveals that, on the contrary, Van Breda opened the door to further development of the jurisprudence on this issue. Procedurally, Van Breda was an appeal of a decision of the Ontario Court of Appeal. The appellant, Club Resorts Ltd. (“Club Resorts”), was a company incorporated in the Cayman Islands. Like Equustek, Van Breda was a conflict-of-laws case, principally concerned with the assumption of jurisdiction for tort claims.48 The actions at issue in Van Breda were advanced by the families of two individuals, one who had suffered serious injuries and another who had died while vacationing at properties in Cuba managed by Club Resorts.49 The Supreme Court ultimately found that Canadian courts had jurisdiction to hear the actions.50 Before considering the court’s comments in detail, it should be noted that they con stituted a short tangent in a long case that was otherwise unrelated to e-commerce. In fact, the court found that Club Resorts was physically present in Canada.51 None- theless, Van Breda is important for this discussion because it reveals the perspective of Canada’s highest court on a novel and uncertain legal issue relevant to the GST/ HST obligations of non-resident e-commerce firms.

“Actual, Not Only Virtual, Presence” At paragraph 87 of the decision in Van Breda, the court wrote, “The notion of carrying on business requires some form of actual, not only virtual, presence in the jurisdiction.”52 As suggested above, this passage may seem to restrict the applicability of the carrying-on-business concept in e-commerce cases. More specifically, the court’s comment may be interpreted as imbuing the carrying-on-business concept at common law with a physical presence requirement. In my view, this interpretation is incorrect for the following reasons.

47 Ibid., at paragraph 87. In a subsequent decision, Chevron Corp. v. Yaiguaje, 2015 SCC 42, the court noted that LeBel J “confined the principles he developed in [paragraph 87 of ] Van Breda to the assumption of jurisdiction in tort actions” (ibid., at paragraph 40). This comment may make it difficult to apply paragraph 87 ofVan Breda outside the conflict-of-laws context. See Chevron, supra, at paragraph 91. 48 Stoyan, supra note 23, at 604. 49 Van Breda, supra note 21, at paragraph 1. 50 Ibid., at paragraph 123. 51 Ibid., at paragraph 122. 52 Ibid., at paragraph 87. policy forum: gst/hst obligations of non-resident e-commerce firms n 1047

First, a plain reading of the text does not support the above interpretation. The wording chosen by the court does not preclude the possibility of carrying on business in Canada through a virtual presence; rather, it indicates that virtual presence alone is not sufficient. That is, a virtual presence may still indicate that a firm is carrying on business in Canada, provided that the firm also has an actual presence. Second, the term “actual presence” does not necessitate “physical presence.” According to the Oxford English Dictionary (OED), “actual” means “[e]xisting in fact,” “real,” “carried out,” or “acted in reality,” and is defined in opposition to the words “potential,” “possible,” and “ideal.”53 In contrast, the OED defines “virtual” (in the context of computing) to mean “not physically present as such but made by software to appear to be so from the point of view of a program or user.”54 In this context, the word “actual” should be read as denoting emphasis, not physicality. This reading is further supported by the fact that the court used the term “physical presence” elsewhere in the judgment,55 yet chose to use “actual . . . presence” in paragraph 87. Accordingly, in light of the OED definitions, it appears likely that the court meant to emphasize the strength and veracity of the non-resident’s presence in a jurisdiction, not the character of that presence. In concrete terms, I take the court’s comments in Van Breda to mean that the mere accessibility of a non-resident’s website (or other passive digital presence) in Canada does not necessarily lead to the conclusion that the non-resident is carrying on business in Canada. Immediately preceding the “actual presence” comment, the court stated that “[a]ctive advertising in the jurisdiction or, for example, the fact that a Web site can be accessed from the jurisdiction would not suffice to establish that the defendant is carrying on business there.”56 Arguably, this statement parallels the longstanding rule from Sudden Valley Inc. v. The Queen that “a mere invitation to treat” does not constitute the carrying on of a business in Canada,57 and is not a meaningful departure from Canada’s existing body of jurisprudence on the issue. As suggested above, on further examination, the court’s comments in Van Breda actually open the door to further case law on the carrying-on-business concept in the context of e-commerce. At paragraph 87 of Van Breda, the court wrote that it was not “asked in this appeal to decide whether and, if so, when e-trade in the jurisdiction would amount to a presence in the jurisdiction.”58 This passage is important for two reasons. First, it implies that e-commerce activities may amount to presence in Canada.59 Second, it establishes a distinction between, on the one hand, “e-trade”

53 OED Online (www.oed.com), the definition of “actual.” In the present context, reference to the ordinary meaning of the word because “actual” does not have a particular legal meaning. 54 OED Online, supra note 53, the definition of “virtual.” 55 Van Breda, supra note 21, at paragraph 122. 56 Ibid., at paragraph 87. 57 76 DTC 6178, at 6180 (FCTD). 58 Van Breda, supra note 21, at paragraph 87. 59 Stoyan, supra note 23, at 605. 1048 n canadian tax journal / revue fiscale canadienne (2020) 68:4

(e-commerce) in a jurisdiction and, on the other hand, “the fact that a Web site can be accessed from the jurisdiction.”60 It is clear that paragraph 87 does not prevent future courts from finding non-residents to have an actual presence in Canada by virtue of their digital activities. That said, the Supreme Court left to future courts the task of adapting the carrying-on-business concept to the digital context.

Summary In summary, Van Breda does not obviously import a physical presence requirement into the carrying-on-business concept at common law. A plain reading of paragraph 87 of the decision demonstrates that carrying on business simply requires a strong presence that extends beyond the mere accessibility of a non-resident’s passive online content in a jurisdiction. Further, in the same paragraph, the court identified a distinction between the treatment of e-commerce and mere website accessibility, implying that non-residents may be able to carry on business in Canada by virtue of their e-commerce activities. While it is not clear whether Van Breda will have a lasting impact in the tax context, it is unlikely that the decision will preclude future courts from finding non-resident e-commerce firms that have no physical presence in Canada to be carrying on business in Canada for GST/HST purposes.

RECOMMENDATIONS Canada clearly has a doctrine (or doctrines) of carrying on business digitally, notwithstanding uncertainty about the application of such a doctrine. In Equustek, the BC Supreme Court and the BC Court of Appeal took an expansive approach to Van Breda, which allowed for the finding that Google carried on business in Canada by virtue of its totally digital presence. Ontario courts have, in contrast, taken a literalist approach, which has resulted in a less broad articulation of the carrying-on-business concept and also less consistent outcomes.61 It is worth noting, however, that in taking this approach the Ontario courts were faced with very different fact situations than the circumstances of the Equustek case. Although the rules for determining whether a firm carries on business digitally in Canada are unclear, and may be applied differently by different courts, it is likely that Canadian courts will continue to find non-resident e-commerce firms to carry on business in Canada by virtue of their digital activities in the Canadian market. But where does this novel, if ambiguous, development in the jurisprudence leave firms? Non-resident e-commerce firms with operations related to Canada must decide on a case-by-case basis whether they will register for the GST/HST. This en- deavour is likely to be a complicated one, given the discrepancies in the existing jurisprudence. Indeed, even diligent tax planning may yield unpredictable results. For consumers, a firm’s decision to register will have an impact on the price of the firm’s supplies to the extent that GST/HST is payable on those supplies. Such a

60 Van Breda, supra note 21, at paragraph 87. See also Stoyan, supra note 23, at 605. 61 Stoyan, supra note 23, at 612 - 13. policy forum: gst/hst obligations of non-resident e-commerce firms n 1049 price increase will in turn affect the relative competitiveness of the firm’s goods or services in the Canadian market. Specifically, firms that do not register will gain a competitive advantage over firms that do. This arrangement produces an incentive for some domestic (or otherwise registrant) firms to restructure their affairs to achieve a price advantage. Such a result is undesirable since it erodes Canada’s GST/ HST base, and it may put domestic and registrant firms at a disadvantage relative to competing non-resident non-registrant firms.

A “Carrying On Business Digitally” Deeming Rule To address the uncertainty arising from the existing jurisprudence, I recommend that Parliament amend the ETA by adding a new rule that would deem firms with a significant digital presence in Canada to carry on business in Canada for GST/HST purposes. Although e-commerce taxation is generally a global problem that requires internationally cooperative solutions,62 the destination-based structure of the GST/ HST means that Canada can unilaterally tax e-commerce in this instance. It has been said that the “recommended indicia [of carrying on business digitally] can be unified under a single principle: whether a defendant is trying to gain an economic benefit from the forum [jurisdiction].”63 A good deeming rule would therefore establish a clear and straightforward threshold to quantify the benefit that non-resident e- commerce firms gain from their activities in the Canadian market.

De Minimis Thresholds In my view, the simplest way to structure the proposed deeming rule is by the use of de minimis standards for digital presence. Fortunately, Canadian policy makers do not have to look far for guidance on how to construct such standards. In the wake of the US Supreme Court decision in South Dakota v. Wayfair, Inc.,64 43 of the 45 US states with sales tax regimes have adopted collection and remittance obligations for remote sellers.65 Although discussion of the Wayfair decision is beyond the scope of this article, given that it hinged on a number of issues unique to the US context, these state sales tax rules are worth considering because they demonstrate how Canada could use de minimis sales standards to “digitize” its GST/HST registration requirements. These standards would admittedly not be a perfect proxy for digital presence; however, as a supplement to the common-law meaning of the carrying-on-business concept, they would provide some much-needed certainty for potential GST/HST registrants.

62 See Jinyan Li, “E-Commerce Tax Policy in Australia, Canada and the United States” (2000) 23:2 University of New South Wales Law Journal 313 - 29, at 315. 63 Stoyan, supra note 23, at 625. 64 585 US 2018. 65 See Jared Walczak and Janelle Cammenga, “State Sales Taxes in the Post-Wayfair Era,” Tax Foundation Fiscal Fact no. 680, December 2019, at 5. 1050 n canadian tax journal / revue fiscale canadienne (2020) 68:4

The sales tax obligations of remote sellers in most US states hinge on whether the seller exceeds a specified sales threshold. The de minimis sales thresholds are quantified by gross sales revenue, total transaction volume, or both, depending on state legislation. The most common threshold in 2019 was gross sales revenue of US $100,000; it applied in 34 states, 24 of which supplemented this figure with a transaction volume threshold.66 Excluding Kansas, which has no threshold, US $100,000 represented the lowest gross sales revenue threshold in the US states. At the other end of the spectrum, four states have introduced a US $500,000 gross sales revenue threshold.67 Three of these states—California, New York, and Texas— rank among the biggest state economies in America.68 In my view, these thresholds represent a reasonable range of options and offer a helpful frame of reference for Canada. In constructing a Canadian de minimis threshold, Parliament should be sensitive to the differences between the GST/HST context and the US state sales tax context. Compared to the United States, Canada has the benefit of fewer sales tax regimes (principally the GST/HST and provincial sales taxes in provinces that have not opted to participate in the HST). This means that GST/HST registration imposes a smaller burden for firms doing business in one or more participating provinces, compared to those subject to US state sales tax equivalents. The fact that many firms doing extensive business across Canada will have a relatively lower administrative burden than firms doing extensive business across the United States may justify the use of relatively low de minimis thresholds for the GST/HST. I recommend that Parliament implement a de minimis threshold for firms making taxable supplies of tangible personal property, intangible personal property, and/or services in Canada. The threshold should be based principally on gross sales revenue and should be set somewhere between the $100,000 and $500,000 thresholds used in the United States. The specific amount set for the threshold should take into account the size of the Canadian economy and the administrative burden for non-resident firms that will be required to register for the GST/HST. A sales volume threshold may also be considered to supplement the gross sales revenue threshold; however, such a method is less sensitive to balancing the administrative costs of compliance with the extent of a firm’s economic activity in Canada.69 The legislated de minimis thresholds should be complemented by additional rules applicable to large market facilitators where supplies are made in Canada through these entities. This proposal is discussed below.

66 Ibid., at 10. 67 Ibid., at 10 - 11. 68 United States, Department of Commerce, Bureau of Economic Analysis, “Gross Domestic Product by State, Fourth Quarter and Annual 2018,” News Release BEA 19 - 19, May 1, 2019, table 3 (www.bea.gov/system/files/2019-04/qgdpstate0519_4.pdf ). 69 Walczak and Cammenga, supra note 65, at 11. policy forum: gst/hst obligations of non-resident e-commerce firms n 1051

Market Facilitator Rules Determining where to set the proposed de minimis thresholds will undoubtedly be a delicate exercise, requiring policy makers to balance the burden of compliance with the desire to achieve greater GST/HST coverage. However, simple supplier- based de minimis thresholds may still fail to capture supplies made by many small firms through a widely used common e-commerce marketplace such as Amazon, eBay, or Etsy. In these cases, individual suppliers may fall below the general de minimis threshold despite net sales facilitated by the e-commerce marketplace far exceeding the threshold. In these instances, it may be preferable from a compliance perspective to have additional market facilitator rules for large e-commerce marketplaces that surpass the general de minimis threshold.70 These rules should shift the burden of registering for, collecting, and remitting GST/HST from the individual supplier to the marketplace in instances where the marketplace surpasses the general de minimis threshold. This approach is desirable because a small number of e-commerce marketplaces are responsible for a significant share of all e-commerce transactions in Canada. For example, the Canadian website of the US-based marketplace Amazon was estimated to have facilitated between 40 percent and 45 percent of all Canadian retail e-commerce sales in 2019.71 Data such as these clearly support the implementation of market facilitator rules.72 To avoid confusion or conflict between market facilitators and suppliers, any market facilitator rules should clearly demarcate when the e-commerce marketplace, rather than the supplier, is required to register for, collect, and remit GST/HST.

CONCLUSION In summary, recent jurisprudence suggests that non-resident e-commerce firms operating in the Canadian market could be obligated to register for, collect, and remit GST/HST on the basis that they carry on business digitally in Canada. There is, admittedly, much uncertainty surrounding the factors that indicate whether such a firm’s activities constitute carrying on business in Canada. This uncertainty threatens the integrity of the GST/HST regime in that non-resident e-commerce firms may enjoy an unfair price advantage while diligent firms may remain unaware of their GST/HST obligations. The Equustek case, despite its deficiencies, suggests a framework for adapting the carrying-on-business concept in Canada to the e-commerce context. Future developments in the common law are desirable, since the judiciary

70 For a brief discussion of this approach, see Pascal Emmanuel Mécéjour, “US States’ Marketplace Facilitator Laws Extend Reach of Tax Collection” (2020) 10:3 Canadian Tax Focus 8 - 9. Such rules have been adopted by several provinces in Canada, as well as many US states. 71 “E-Commerce Revenue Analytics amazon.ca,” ecommerceDB (https://ecommercedb.com/en/ store/amazon.ca). 72 For a discussion of this issue, see David Gamage, Darien Shanske, and Adam Thimmesch, “Taxing E-Commerce in the Post-Wayfair World” (2019) 58 Journal of Law and Policy 71 - 94. 1052 n canadian tax journal / revue fiscale canadienne (2020) 68:4 is arguably better positioned than Parliament to develop an “open-ended flexible test with a multitude of factors.”73 However, in light of the growing presence of non-resident e-commerce firms in the Canadian market, it is not enough to simply wait for the jurisprudence to evolve. To maintain Canada’s tax policy principles and to clarify the obligations of potential GST/HST registrants, Parliament should amend the ETA to deem firms that exceed specified de minimis sales thresholds to be carrying on business in Canada for GST/HST purposes. Although in-depth statutory interpretation may assist in delineating the GST/ HST obligations of non-resident e-commerce firms,74 such interpretation is unlikely to insulate all such firms from liability. Since the determination of carrying on business results from a factual inquiry,75 a finding that a firm carries on business in Canada will be entitled to a strong degree of deference from higher courts. Where such a firm makes taxable supplies in Canada, it is clear that the firm is required to register for, collect, and remit GST/HST.

73 Stoyan, supra note 23, at 624. 74 This point is explored further by Zvi Halpern-Shavim, “Policy Forum: Carrying On About Carrying On Business: A Response to ‘The GST/HST Obligations of Non-Resident E-Commerce Firms’ ” in this feature. 75 For further discussion of this factual inquiry and the development of the carrying-on-business concept, see Malcolm Gammie, “Policy Forum: Much Ado About Doing Not Much: Some Reflections on the Jurisdiction To Tax Business Transactions” in this feature. canadian tax journal / revue fiscale canadienne (2020) 68:4, 1053 - 67 https://doi.org/10.32721/ctj.2020.68.4.pf.halpern-shavim

Policy Forum: Carrying On About Carrying On Business: A Response to “The GST/HST Obligations of Non-Resident E-Commerce Firms”

Zvi Halpern-Shavim*

PRÉCIS Les fournisseurs non résidents de produits et services numériques ne sont pas tenus de percevoir et de remettre la taxe sur les produits et services (TPS)/taxe de vente harmonisée (TVH) s’ils n’ « exploitent pas une entreprise » au Canada. L’expression « exploiter une entreprise » au Canada n’est pas définie dans la législation. Laissant de côté les arguments politiques en faveur, ou contre, l’obligation que les fournisseurs non résidents sans présence physique au Canada s’inscrivent au régime de la TPS/TVH, l’auteur de cet article examine s’il serait judicieux de réinterpréter le cadre juridique actuel pour exiger que ces fournisseurs s’inscrivent au régime de la TPS/TVH sur la base d’une présence purement numérique. Il aborde ces questions en examinant la signification textuelle, contextuelle et intentionnelle de l’expression « exploiter une entreprise » (et ses permutations) dans la Loi sur la taxe d’accise (Canada) (LTA). Il

* Of Blake Cassels & Graydon LLP, Toronto (e-mail: [email protected]).

1053 1054 n canadian tax journal / revue fiscale canadienne (2020) 68:4 termine en précisant que si les changements dans d’autres domaines du droit, tels que la jurisprudence en matière de conflits de juridictions, peuvent élargir le sens en common law de l’expression « exploiter une entreprise » dans une compétence pour inclure le fait d’avoir une présence purement numérique, le cadre d’interprétation particulier des lois fiscales au Canada et l’utilisation de cette expression dans laLTA elle-même amènent à conclure qu’un changement législatif serait le moyen la plus approprié d’imposer des obligations d’enregistrement et de recouvrement à ces fournisseurs non résidents.

ABSTRACT Non-resident suppliers of digital products and services are not required to collect and remit goods and services tax (GST)/harmonized sales tax (HST) if they are not “carrying on business” in Canada. The term “carrying on business” in Canada is not defined in the legislation. Leaving aside policy arguments in favour of, or against, requiring such non-resident suppliers without a physical presence in Canada to register, the author of this article considers whether the current legal framework should be reinterpreted to require such suppliers to become registered for GST/HST on the basis of a purely digital presence. He addresses these issues by considering the textual, contextual, and purposive meaning of the term “carrying on business” (and its permutations) in the Excise Tax Act (Canada) (ETA). He concludes that while developments in other areas of law, such as conflict-of-laws jurisprudence, may expand the common-law meaning of “carrying on business” in a jurisdiction to include having a purely digital presence, the interpretive framework particular for tax statutes in Canada, and the specific use of the term in the ETA itself, lead to the conclusion that a specific change in law would be the more appropriate way to impose registration and collection obligations on such non- resident suppliers. KEYWORDS: GST n ELECTRONIC COMMERCE n NON-RESIDENTS n CARRYING ON BUSINESS n DIGITAL n REGISTRATION

CONTENTS Introduction 1055 Interpreting the ETA as a Tax Statute 1056 A Note Regarding Terminology 1058 The Textual, Contextual, and Purposive Analysis 1059 Textual Analysis 1060 Contextual Analysis 1061 Purposive Analysis 1062 Consistency, Predictability, and Fairness 1063 Quebec 1064 Saskatchewan 1065 Van Breda 1066 Conclusion: Ambiguities Should Be Decided in Favour of the Taxpayer 1067

policy forum: carrying on about carrying on business: a response n 1055

INTRODUCTION More and more frequently, one can find news reports or opinion pieces in Canada about goods and services tax (GST)/harmonized sales tax (HST) that is not being paid on cross-border supplies of services and digital files to Canadian consumers. A helpful summary of the issue by Rosalie Wyonch of the C.D. Howe Institute runs as follows:

[Non-resident] providers of digital products and services, ranging from e-books and online games to streaming services . . . are not obligated to collect and remit sales tax if they are not “carrying on business” in Canada. Instead, the consumers of the service are responsible for determining and paying the associated GST/HST. This creates two major problems. 1 Among general consumers, compliance is virtually nonexistent. This leaves significant amounts of tax revenue uncollected, but enforcement at the individual level would be prohibitively expensive and, likely, incredibly unpopular. 2 Since the sales tax is not being paid, foreign suppliers have a competitive advantage over domestic companies that are required to collect and remit the GST/HST on behalf of consumers. Sales taxes collected by domestic suppliers add 5 percent to 15 percent to the price of their products. Therefore, foreign vendors can extract more revenue, while charging the same final price as a domestic company, by not charging GST/HST.1

One might wonder why consumers are unable to comply with their GST/HST obligations while they do not have the same difficulty with paying income taxes on a self-assessment basis. But in any case, a growing consensus is taking shape that if Canada wants consumers to pay GST/HST on cross-border supplies of digital services and products, it should impose an obligation on the vendors to collect the tax, even if that means requiring non-Canadian entities with no physical operations in Canada to register for GST/HST purposes. Furthermore, there is generally an assumption that such obligations on non- residents with no physical presence in Canada do not currently exist, and that to implement such a change would require amendments to the Excise Tax Act.2 However, it has been argued that perhaps the ETA as currently drafted already imposes such obligations on non-resident vendors, provided that one is willing to revisit first principles. After all, the legislation only requires that the non-resident be carrying on business in Canada in order to be required to register.3 Are we so

1 Rosalie Wyonch, Bits, Bytes, and Taxes: VAT and the Digital Economy in Canada, C.D. Howe Institute Commentary no. 487 (Toronto: C.D. Howe Institute, August 2017), at 2 (www .cdhowe.org/sites/default/files/attachments/research_papers/mixed/Commentary_487.pdf ). 2 RSC 1985, c. E- 15, as amended (herein referred to as “the ETA”). 3 See ETA paragraph 240(1)(c). 1056 n canadian tax journal / revue fiscale canadienne (2020) 68:4 certain that a digital presence in Canada is not considered to be carrying on business? Such an argument was suggested in the context of the sharing economy, in a paper in which my co-author and I made the following comment on the Equustek case:4

Although it originated outside of the tax context (that is, in the context of extraterri- torial reach of a provincial court), the relevance of this kind of reasoning for GST/HST purposes remains to be seen. If such reasoning were applicable to the GST/HST context, a non-resident platform provider could potentially be found to be carrying on business in Canada, in which case it could be required to register for GST/HST purposes and collect GST/HST on its fees.5

The position has been expanded upon in the preceding article in this feature, by reference to the Equustek case and the Supreme Court of Canada’s decision in Club Resorts Ltd. v. Van Breda.6 However, having considered the author’s analysis of the existing case law and upon further reflection, it seems to me that the distinctions between the context of the Equustek case—namely, conflict-of-laws jurisprudence— and the GST/HST context are too significant. The conclusions in Equustek (and other conflict-of-laws cases that refer to a carrying-on-business test) should not have a material bearing on the GST/HST analysis.

INTERPRETING THE ETA AS A TAX STATUTE A person is required to register for GST/HST purposes if it makes a taxable supply in Canada in the course of a commercial activity in Canada, except where “the person is a non-resident person who does not carry on any business in Canada.”7 A corollary rule is that a taxable supply is deemed to be made outside Canada by a non-registered non-resident (and is therefore not subject to GST/HST), unless the supply is made “in the course of a business carried on in Canada.”8 The two rules work together. Where a non-registered non-resident is not carrying on business in Canada, it is deemed not to make any taxable supplies in Canada and is not required to register for GST/HST purposes. Where the non-resident does carry on “any business in Canada,” it is similarly not required to register. But if the non-resident

4 Equustek Solutions Inc. v. Jack, 2014 BCSC 1063; and Equustek Solutions Inc. v. Google Inc., 2015 BCCA 265. 5 Ian Caines and Zvi Halpern-Shavim, “Income and Commodity Tax Considerations for the Sharing Economy,” in Report of Proceedings of the Sixty-Eighth Tax Conference, 2016 Conference Report (Toronto: Canadian Tax Foundation, 2017), 35:1 - 26, at 35:15. 6 2012 SCC 17. See Nicholas Shatalow, “Policy Forum: The GST/HST Obligations of Non-Resident E-Commerce Firms: Jurisprudence and Policy” in this feature. 7 ETA paragraph 240(1)(c). Other exceptions apply for small suppliers and one-off supplies of real property, but these are not relevant to the analysis here. 8 ETA paragraph 143(1)(a). Again, other exceptions apply that are not relevant here. policy forum: carrying on about carrying on business: a response n 1057 makes taxable supplies in the course a business carried on in Canada, such supplies are made in Canada and the non-resident is required to register. The question of whether a person is “carrying on business in Canada,” or whether a business is “carried on in Canada,” for the purposes of mandatory GST/HST registration under the ETA is principally a question of statutory interpretation. The statute that is to be interpreted is the ETA. Accordingly, the meaning given to that phrase in the Equustek case, where the BC courts established that they had jurisdiction over Google Inc. under a BC statute, the Court Jurisdiction and Proceedings Transfer Act (CJPTA),9 appears to be of very limited usefulness. This is for three reasons, all drawing from the fact that the ETA is a tax statute, whereas the CJPTA (and the common-law torts context from which it is derived) is not. First, it is well known that Driedger’s modern principle of statutory interpretation is generally applicable to Canadian legislation. According to that principle, “the words of an Act are to be read in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament.”10 But in the case of tax statutes, Driedger’s principle has been modified to address unique issues that arise in the sphere of taxation. The Tax Court of Canada recently described the hierarchy of interpretive rules as follows:

In tax legislation, if the text of a statute is clear it must be applied. Where, however, the text admits of more than one reasonable interpretation, greater emphasis on and recourse to context, purpose and the scheme of the legislation may be necessary and may reveal or resolve latent ambiguities which may not have initially been apparent. If latent or explicit ambiguities exist, courts must also look at the context and the purpose of the provision to determine the most plausible interpretation. In rare instances, where a textual, contextual and purposive approach to statutory interpretation does not resolve the interpretative issue, there is a residual presumption, to be applied ex- ceptionally, in favour of the taxpayer.11

Second, as noted at the end of the extract above, there is a residual presumption that ambiguities should be resolved in favour of the taxpayer. Underlying this rule is the fundamental principle that if the government intends to impose taxes on its citizens, it must do so with clear legislation. Where the statute is determined to be truly ambiguous, following the analytical process described above, the issue is to be resolved in favour of the taxpayer, not the state. Third, the Supreme Court of Canada’s injunction in Canada Trustco Mortgage Co. v. Canada12 regarding the interpretation of income tax legislation is equally relevant

9 SBC 2003, c. 28. 10 E.A. Driedger, Construction of Statutes, 2d ed. (Toronto: Butterworths, 1983), at 87. 11 Montecristo Jewellers Inc. v. The Queen, 2019 TCC 31, at paragraph 43. 12 2005 SCC 54. 1058 n canadian tax journal / revue fiscale canadienne (2020) 68:4 to the ETA. Notably, “[t]he provisions of the Income Tax Act must be interpreted in order to achieve consistency, predictability and fairness so that taxpayers may manage their affairs intelligently.”13 Similarly, the provisions of the ETA—including, in particular, the provisions under scrutiny here relating to mandatory GST/HST registration—must be interpreted in a manner that is consistent, predictable, and fair. Usually this injunction leads to an emphasis on the textual meaning of a provision.14 But where the textual meaning is not necessarily fixed, as in the phrase “carrying on business in Canada,” courts should still be loath to introduce new meanings that impose new obligations on taxpayers that are inconsistent with the settled expectations of the public. There is no robust body of case law interpreting the meaning of “carrying on business” for GST/HST purposes under the ETA. It has generally been assumed by many, including the Canada Revenue Agency (CRA), that the factors developed by courts in the income tax context would also apply in the GST/HST context, with suitable changes to take into account the differences between those contexts. In any event, by following the process described above for interpreting tax statutes, and bearing in mind the presumption toward consistency and the presumption in favour of the taxpayer in the case of true ambiguity, I argue that it would be inappropriate for the government to require non-resident, cross-border service providers without any physical presence in Canada to register for GST/HST purposes, on the basis of a development in the common law in a non-tax context.

A NOTE REGARDING TERMINOLOGY In Equustek, the BC Supreme Court appears to have ultimately found that Google Inc. was carrying on business in British Columbia because it operated an interactive website (the Google search engine) that collected a wide range of user information from BC residents, and it also entered into advertising contracts with BC residents.15 The BC Court of Appeal summarized the activities of Google Inc. in British Colum- bia as follows:

While Google does not have servers or offices in the province and does not have resident staff here, I agree with the chambers judge’s conclusion that key parts of Google’s business are carried on here. The judge concentrated on the advertising aspects of Google’s business in making her findings. In my view, it can also be said that the gather ing of information through proprietary web crawler software (“Googlebot”) takes place in British Columbia. This active process of obtaining data that resides in the province or is the property of individuals in British Columbia is a key part of Google’s business.16

13 Ibid., at paragraph 12. 14 See ibid., at paragraph 11. 15 Equustek, supra note 4 (BCSC), at paragraphs 48 - 50. 16 Equustek, supra note 4 (BCCA), at paragraph 54. policy forum: carrying on about carrying on business: a response n 1059

In this article, references to non-resident, cross-border digital service providers are intended to capture businesses with similar characteristics.17 Namely, the service providers

n have no servers or offices in Canada, or any employees in Canada; n may obtain information about individuals or businesses in Canada from software that operates on computers situated in Canada (owned by third parties); and n contract with Canadian-resident customers to supply intangible personal property or services over the Internet from outside Canada.

THE TEXTUAL, CONTEXTUAL, AND PURPOSIVE ANALYSIS Although the term “carrying on business” (or any of its permutations) is not defined in the ETA, I argue that the textual, contextual, and purposive analysis of that term points away from a non-resident with no physical presence in Canada. As a backdrop to this analysis, I note that the term was included in the ETA having already been subject to some interpretation by common-law courts in the tax context. In early British jurisprudence, the place where the contract was made was a determinative test for where a person was “carrying on business.”18 However, this began to change following Lord Atkin’s decision in F.L. Smidth & Co. v. F. Green- wood, in which he stated:

The contracts in this case were made abroad. But I am not prepared to hold that this test is decisive. I can imagine cases where a contract of re-sale is made abroad, and yet the manufacture of the goods, some negotiation of the terms, and complete execution of the contract take place here under such circumstances that the trade was in truth exercised here. I think that the question is, where do the operations take place from which the profits in substance arise?19

Smidth marks the beginning of the dominant modern approach in Canadian jurisprudence to determining whether a non-resident was carrying on a business in Canada: the question is, in what jurisdiction did the operations take place from which the profits arise? Subsequent case law has suggested that certain factors are useful in determining whether a non-resident person is carrying on business in Canada. These factors include20

17 Shatalow, supra note 6, refers to such businesses as “non-resident e-commerce firms”; presumably, that term is intended to capture the same set of characteristics. 18 See Erichsen v. Last (1881), 4 TC 422 (CA). 19 (1922), 8 TC 193, at 203 - 4 (HL); aff’g [1921] 3 KB 583 (CA). 20 See GST/HST Policy Statement P- 051R2, “Carrying On Business in Canada.” 1060 n canadian tax journal / revue fiscale canadienne (2020) 68:4

n the place of delivery of goods, n the place where services are rendered, n the place of payment for the goods or services, n the place where purchases are made in connection with goods or services, n the place of manufacture or production, n the place from which transactions are solicited, n the location of an inventory of goods, n the location of a bank account relating to the business, n the place where the non-resident’s name and business are listed in a directory, and n the location of a branch office.

Of course, it is always possible that additional factors could be added by subsequent cases. It is therefore important to consider the interpretation of the phrase more closely.

Textual Analysis The textual meaning of the phrases “a business carried on in Canada” and “carry on any business in Canada” is ambiguous. The CRA has noted that the phrase consists of two requirements: (1) there must be a “business,” and the business must be “carried on [in Canada].”21 With regard to the “business” prong of the test, the CRA notes that the ETA has a unique definition of “business,” which leads to an expanded meaning of carrying on business for GST/HST purposes:

[U]nlike the definition in theIncome Tax Act, the definition of “business” in the ETA[ ] also includes any activity engaged in on a regular or continuous basis that involves the supply of property by way of lease, licence or similar arrangement. Therefore, where it is determined that a non-resident person is carrying on business in Canada for GST/ HST purposes, this does not necessarily mean that the non-resident is considered to be carrying on business in Canada for income tax purposes.22

With regard to the “carried on in Canada” prong, the CRA provides the set of factors gleaned from the income tax jurisprudence listed above, without any particular explanation for what it means to carry on a business, except that “[i]n general, a non-resident person must have a significant presence in Canada to be considered to be carrying on business in Canada.”23 The question of whether a “digital” presence in Canada would qualify cannot be determined from the text. Therefore, the textual analysis is ambiguous.

21 Ibid. 22 Ibid. 23 Ibid. policy forum: carrying on about carrying on business: a response n 1061

Contextual Analysis The common-law context from which the phrase “carrying on business” in a jurisdiction was developed, before the GST legislation was brought into force, has been described above. The provisions that require a determination of whether a person is carrying on business in Canada, or a business is carried on in Canada, have also been discussed above. It may be helpful to contextualize the statutory rules further in the overall scheme of the ETA. In this regard, the following two points appear useful. First, not everyone is, or should be, eligible to register for GST/HST purposes. The discussion above has been limited to mandatory registration for non-residents, but the ETA also governs voluntary registrations. A GST/HST registration can be beneficial for a business because only registered entities can claim input tax credits. Without the benefit of registration, any GST/HST payable by a non-resident is a cost.24 To qualify for voluntary registration, a non-resident digital supplier must have a sufficient connection to Canada. In particular, the non-resident may register if it “is engaged in a commercial activity in Canada.”25 Alternatively, the non-resident may register if in the ordinary course of carrying on business outside Canada, it has entered into an agreement for the supply of services to be performed in Canada or a supply of intangible personal property to be used in Canada.26 Second, the ETA uses broader terminology than “carrying on business” when it wants to capture any activity performed in Canada. As noted above, the voluntary registration rules permit registration when a person “is engaged in a commercial activity in Canada,” and “commercial activity” is defined to include a “business.”27 A lesser level of activity in Canada is described to impose tax on a particular kind of “qualifying taxpayer” (a financial institution) where that person, at any time in a specified year, “carries on, engages in or conducts an activity in Canada.”28 It appears that, under the voluntary registration rules, a non-resident that carries on business outside Canada and makes supplies of intangible property or services to Canadians may register voluntarily. It would be odd to interpret the mandatory registration rule of “carrying on business in Canada” to include the same set of circumstances where the non-resident makes supplies of intangible property or services to Canadians. Furthermore, when considering whether a person is “carrying on business in Canada,” one must also harmonize the meaning of that term with other

24 In fact, when Quebec imposed its non-resident Quebec sales tax (QST) registration regime, it created a specific class of registration that only permitted tax collection, without any input tax refund eligibility. 25 ETA paragraph 240(3)(a). 26 ETA paragraph 240(3)(b). 27 ETA subsection 123(1), “commercial activity.” 28 ETA subparagraph 217.1(1)(b)(iii). 1062 n canadian tax journal / revue fiscale canadienne (2020) 68:4 terms in the ETA that describe some business activity in Canada and that do not create GST/HST registration obligations: namely, being “engaged in” a commercial activity and “carrying on, engaging in, or conducting an activity” in Canada.

Purposive Analysis As Steven D’Arcy (now a judge at the Tax Court of Canada) argued in 2005, one of the key design features of the ETA was to exclude most non-residents from the administrative requirements to register for, collect, and remit GST, and fileGST returns:

It was felt that placing onerous administrative requirements on non-residents would discourage them from selling goods and services to Canadian businesses and from purchasing goods and services from Canadian businesses. In addition, it was felt that, from a tax collection viewpoint, the government should minimize the instances where a non-resident collects GST for subsequent remittance to the federal government. The designers of the GST Act [that is, part IX of the ETA] believed that they accomplished the above by including in the legislation a carrying on business test for determining when a non-resident is subject to the administrative requirements of the GST Act. If a non-resident is not carrying on business in Canada then it is not required to collect GST on supplies it makes, register for GST purposes or file GST returns. In other words, a non-resident who is not carrying on business in Canada is not subject to GST administrative requirements. . . . The application of the GST (including the administrative requirements) to non- residents involves a delicate balancing act. In the first instance the government must ensure that GST is not an impediment for non-residents who wish to do business with Canada. This must be balanced with the need to protect tax revenues and the competitiveness of Canadian businesses.29

This balance between competitiveness and protecting tax revenues is critical. Additionally, there is a reasonable concern that GST/HST that is collected by non- residents may not be remitted. Any development in the meaning of “carrying on business” must take into account both sides of the balance. In Equustek, at the BC Supreme Court, Google made a submission that interpreting the CJPTA to grant British Columbia jurisdiction would amount to an impediment to non-residents doing business in Canada. Google argued that the court’s analysis would give every state in the world jurisdiction over its search services. The court responded:

That may be so. But if so, it flows as a natural consequence of Google doing business on a global scale, not from a flaw in the territorial competence analysis. As Janet Walker writes in Castel & Walker: Canadian Conflict of Laws, loose-leaf, 6 ed (Markham,

29 Steven D’Arcy, “Carrying On Business in Canada,” in 2005 CPA Canada Commodity Tax Symposium (Toronto: Chartered Professional Accountants of Canada, 2005), paper 24, at paragraphs 2 - 3 and 5. policy forum: carrying on about carrying on business: a response n 1063

Ontario: LexisNexis, 2005), ch 11 at 27, a legal person such as a corporation can be subject to multiple jurisdictions whether because it is resident there through registration, or because it is carrying on business in that jurisdiction. Further, the territorial competence analysis would not give every state unlimited jurisdiction over Google; jurisdiction will be confined to issues closely associated with the forum in accordance with private international law.30

In other words, British Columbia’s jurisdictional rights strike the appropriate balance for non-residents between the different jurisdictions, and the conclusion that BC courts have limited jurisdiction over non-residents is appropriately confined to the issues closely associated with the jurisdiction. The GST/HST registration requirement is different in certain respects from the question of territorial competence in that it imposes obligations on non-residents that are disproportionate to the connection to Canada, and thus arguably creates an imbalance. First, registration changes the status of the entire legal entity, not just the portion that does business with Canada. If a division of a multinational corporation becomes registered for GST/HST purposes, every other business division is required to charge GST/HST to its Canadian customers, whether it has a connection to Canada or not. Furthermore, Canadian customers may no longer be able to zero- rate their supplies to the multinational corporation, since several zero-rating rules rely on the non-resident being non-registered. Finally, the non-resident is not only subject to Canadian tax legislation; it is also subject to the administrative requirements of filing tax returns, issuing compliant invoices, and related obligations. The failure to meet such requirements could lead to monetary losses for the non-resident. To the extent that the ETA intended to maintain the appropriate balance for non- resident suppliers that do not “carry on business in Canada,” to keep them outside the GST/HST filing and reporting obligations, and to maintain their ability to stay outside the Canadian tax net when doing business in Canada from a foreign jurisdiction (with respect to both sales in Canada and zero-rated sales from Canada), any attempt to expand the meaning of that term arguably frustrates the purpose of the legislation. Furthermore, to the extent that the framers of the ETA wanted to keep the GST/HST in Canada owing to concerns over recoverability, such concerns must also be taken into account.

CONSISTENCY, PREDICTABILITY, AND FAIRNESS As noted above, the ETA should be interpreted “in order to achieve consistency, predictability and fairness so that taxpayers may manage their affairs intelligently.”31 The CRA has the power to register a person without that person’s consent,32 and so any new interpretation of “carrying on business in Canada” that could impose new

30 Equustek, supra note 4 (BCSC), at paragraph 64. 31 See supra note 13 and the accompanying text. 32 ETA subsection 241(1.5). 1064 n canadian tax journal / revue fiscale canadienne (2020) 68:4 obligations on a non-resident person should be viewed with suspicion. This is not an argument against the incremental development of the common law in general, but it does mean that developments in the common law should not necessarily translate into reinterpretations of tax statutes where taxpayers place significant reliance on consistent and predictable interpretations of the law. At the very least, the CRA should not change its interpretation of the law and its assessing practices without first providing some published guidance to the community regarding its intentions. At present, the CRA does not appear to consider a person to be carrying on business in Canada if the person’s only presence in Canada is on the Internet and the only supplies that are made in Canada are supplies of intangible personal property. The CRA’s published guidance aligns with this view. For instance, the CRA published the following example to illustrate a scenario where a non-resident digital supplier is not carrying on business in Canada:

A non-resident corporation supplies downloadable audio files by way of sale. The non- resident has a Web site hosted on its own server located at its main office in the United States, and advertises its Web site on the Internet. The advertisements are directed to the Canadian market. The Web site and server are fully interactive: the Canadian customer may view product listings of music and other advertising, place orders (including payment for audio files selected), and download a copy of the purchased audio files without any contact with the non-resident’s personnel. The place of contract is in Canada. The customer pays by credit card and an independent ISP [Internet service provider] located in Canada processes payments for the non-resident. Once the audio files are received by the customer, they may be used in Canada. All customer service and after-sales support is provided by means of telephone or e-mail communication by the non-resident’s personnel located in its main office in the United States. In this example the non-resident is not considered to be carrying on business in Canada. Of the list of factors to be considered, only the following would indicate some business activity of the non-resident in Canada: advertising is directed to potential customers in Canada; the place of contract is in Canada; product purchases are made in Canada; and payment is processed in Canada. However, in general these factors would not, by themselves, indicate that the business was being carried on in Canada.33

Federal taxes are not the only tax regimes in Canada that rely on a carrying-on- business test. As discussed below, two provinces that have decided to impose sales tax collection obligations on non-residents (Quebec and Saskatchewan) have not relied on their pre-existing tests.

Quebec The Quebec sales tax (QST) is substantially harmonized with the federal GST/HST, and it also generally relies on a “carrying on business in Quebec” test to determine

33 GST Technical Information Bulletin B- 090, “GST/HST and Electronic Commerce,” July 2002. The CRA has made similar comments in more recent rulings and interpretation letters. policy forum: carrying on about carrying on business: a response n 1065 whether a person is required to register for QST purposes. On June 12, 2018, the National Assembly of Quebec passed an amendment to An Act Respecting the Québec Sales Tax.34 Prior to the amendment, non-resident suppliers of taxable movable property or services to consumers in Quebec were not required to register with Revenu Québec for the purposes of collecting and remitting QST applicable to their taxable supplies unless they had a permanent establishment, or carried on business, in Quebec. The amended QST legislation requires non-residents of Quebec that supply taxable movable property or services to “specified Quebec consumers” to register with Revenu Québec under the new “specified registration system,” and to collect and remit QST applicable to their taxable supplies, even though they do not have a permanent establishment, or carry on business, in Quebec. Certain operators of “specified digital platforms” are also required to register under the specified registration system where they do not carry on business in Quebec but the platform enables unregistered suppliers outside Quebec to make taxable supplies to consumers in Quebec. The specified registration system is distinct from normal QST registration, and non-resident registrants under the specified registration system are not eligible for input tax refunds. The registration number is also different from the normal QST number, which has 10 digits and the suffixTQ 0001; the specified registration number begins with the prefix NR“ .” The fact that the Quebec government did not try to reinterpret its laws to allow Revenu Québec to register non-residents with only a digital presence in the province, and instead opted to enact new legislation with an entirely new registration system, indicates that the government did not interpret its statute as being sufficiently broad to impose obligations on such non-residents without explicit legislation.

Saskatchewan In 2017, 2018, and 2020, Saskatchewan made successive changes to the definition of “vendor” in the province’s sales tax legislation, in order to impose additional obligations on out-of-province vendors that do not carry on business in Saskatchewan. A “vendor” is defined in section 3(1) of the Provincial Sales Tax Act as follows:

[A]ny person who, within the province and in the course of his business or in the course of continuous or successive acts: (i) sells or leases tangible personal property to a consumer or user at a retail sale in the province for purposes of consumption or use, and not for resale; (ii) sells or leases taxable services to a user at a retail sale in the province for purposes of use and not for resale; or (iii) sells tangible personal property to a consumer or user to be used by the consumer or user for the purpose of promotional distribution.35

34 CQLR c. T- 0.1. 35 Provincial Sales Tax Act, RSS 1978, c. P- 34.1, section 3(1) (emphasis added). 1066 n canadian tax journal / revue fiscale canadienne (2020) 68:4

In three separate amendments, Saskatchewan expanded the definition of “vendor” to include vendors that do not otherwise “carry on business in the province” but that make sales to Saskatchewan consumers.

For the purpose of the definition of “vendor” and subject to the regulations, a retail sale in the province includes a retail sale of tangible personal property by a person who does not otherwise carry on business in the province, if the property is acquired for use or consumption in or relating to Saskatchewan.36 For the purpose of the definition of “vendor” and subject to the regulations, a retail sale in the province includes a retail sale of tangible personal property or of a taxable service by a person who does not otherwise carry on business in Saskatchewan, if the tangible personal property or the taxable service is acquired for use or consumption in or relating to Saskatchewan.37 For the purposes of the definition of “vendor” and subject to the regulations: (a) a marketplace facilitator is a vendor for the purposes of this Act, whether or not the marketplace facilitator carries on business in Saskatchewan, and (b) the operator of an online accommodation platform is a vendor for the purposes of this Act, whether or not the operator of that online accommodation platform carries on business in Saskatchewan.38

In each case, Saskatchewan explicitly refers to a vendor that is not carrying on business in the province in order to impose registration obligations on non-residents. While Saskatchewan might have relied on an expanded notion of “carrying on business” in the province to impose obligations on such suppliers, such an approach was not followed. Rather, Saskatchewan opted to take for granted that such businesses are not, or may not be, carrying on business in Saskatchewan, and to impose registration obligations regardless.

VAN BREDA It has been suggested that the Supreme Court of Canada’s decision in Van Breda provides support for the position that non-resident, cross-border digital suppliers may be resident in Canada. The case does perhaps leave open that possibility,39 but the question of whether a non-resident, cross-border digital supplier was carrying on business in Canada was not at issue in that case. Rather, the case elaborated upon the “real and substantial connection” test that underlay conflict-of-laws statutes such as the BC CJPTA. Although statutes similar to the BC CJPTA have been promul- gated in several provinces, Ontario’s conflict-of-laws rules remained governed by the procedural rules under the Rules of Civil Procedure and the common law.40

36 The Provincial Sales Tax Amendment Act, 2017, SS 2017, c. 24, section 3(2) (emphasis added). 37 The Provincial Sales Tax Amendment Act, 2018, SS 2018, c. 30, section 3(2) (emphasis added). 38 The Provincial Sales Tax Amendment Act, 2020, SS 2020, c. 34, section 4 (emphasis added). 39 In one sentence of comments made in obiter: Van Breda, supra note 6, at paragraph 87. 40 Ibid., at paragraphs 40 - 43. policy forum: carrying on about carrying on business: a response n 1067

Accordingly, the case develops a common-law framework for determining when a court may take jurisdiction of a tort case, as follows:

To recap, in a case concerning a tort, the following factors are presumptive connecting factors that, prima facie, entitle a court to assume jurisdiction over a dispute: (a) the defendant is domiciled or resident in the province; (b) the defendant carries on business in the province; (c) the tort was committed in the province; and (d) a contract connected with the dispute was made in the province.41

The case does not actually represent a development in the common law in respect of the meaning of “carrying on business” in Canada. But at any rate, following the line of argument above, a development in the common law in an area unconnected to tax or GST/HST, where the context and purpose underlying the use of the term were not carefully weighed and considered, should not automatically modify the meaning of a term used in the ETA.

CONCLUSION: AMBIGUITIES SHOULD BE DECIDED IN FAVOUR OF THE TAXPAYER Finally, as noted above, if the interpretation of a provision in a tax statute is truly ambiguous, the disputed interpretation should be resolved in favour of the taxpayer. As noted above, the context and purpose of the carrying-on-business test were intended to maintain the distinction between non-residents that are required to register for GST/HST and those that may register voluntarily or may not register at all. The underlying common-law test for carrying on business is the jurisdiction where the operations from which the profits arise are located. We have already assumed that the non-resident digital suppliers have no physical computer servers, employees, or premises in Canada. To the extent that there is ambiguity as to whether the concept of “operations” could include operations that occur “on the Internet” (that is, by sending or receiving signals to or from computers in Canada that are owned by someone other than the non-resident supplier), it seems reasonable on the basis of the tiebreaker rule that the interpretive dispute should be settled in favour of the taxpayer. Amendments to the legislation could be introduced to impose a legal obligation on non-resident digital suppliers to collect GST/HST, and the federal government may already be considering its legislative options following the recent changes by Quebec and Saskatchewan. But without such changes, for the reasons described above, it seems inappropriate for the tax authorities or the courts to interpret the ETA as already imposing such obligations on non-resident, cross-border digital suppliers.

41 Ibid., at paragraph 90. canadian tax journal / revue fiscale canadienne (2020) 68:4, 1069 - 82 https://doi.org/10.32721/ctj.2020.68.4.pf.gammie

Policy Forum: Much Ado About Doing Not Much: Some Reflections on the Jurisdiction To Tax Business Transactions

Malcolm Gammie*

PRÉCIS La plupart des États aspirent à imposer, d’une manière ou d’une autre, la production, les produits ou les bénéfices des activités commerciales menées sur leur territoire. La portée de l’imposition est toutefois limitée lorsque l’entreprise a son siège social à l’étranger et n’entretient de relations commerciales qu’avec des personnes se trouvant dans l’État en question. Un point de départ dans de telles situations consiste à examiner si l’État peut revendiquer sa juridiction à l’égard d’une entreprise étrangère pour faire valoir ses droits. La question suivante est de savoir si l’objet de la dette fiscale, soit la base fiscale, relève de la juridiction de l’État. À cet égard, des considérations différentes s’appliquent pour l’imposition d’une taxe à la consommation, telle que la taxe sur la valeur ajoutée (TVA), par rapport à un impôt sur les bénéfices des entreprises. Cet article examine en premier lieu la base juridique de la common law adoptée par les tribunaux anglais avant de considérer le critère créé par les tribunaux britanniques au XIXe et au début du XXe siècle pour déterminer si les bénéfices d’une entreprise ayant son siège social à l’étranger relèvent néanmoins de la juridiction fiscale du Royaume-Uni. Il compare ensuite l’approche britannique à l’imposition des bénéfices tirés de l’activité commerciale britannique par une entreprise étrangère avec l’approche adoptée pour la TVA telle qu’elle est appliquée, notamment, à la fourniture à distance de services numériques depuis l’étranger. Les règles du Royaume-Uni découlent actuellement de directives européennes, et il reste à voir si elles changeront au fil du temps après le Brexit.

ABSTRACT Most states aim to tax in one way or another the outputs, products, or profits of business activity conducted within their jurisdiction. The scope to tax is limited, however, when the business is based abroad and trades only with persons in the state in question. A starting point in such situations is to consider whether the state can claim jurisdiction over a foreign business to enforce its claims. The next question is whether the subject matter of the tax charge—the tax base—is amenable to the state’s jurisdiction. In this respect,

* Past chair of the Tax Law Review Committee, Institute for Fiscal Studies, London, United Kingdom (e-mail: [email protected]).

1069 1070 n canadian tax journal / revue fiscale canadienne (2020) 68:4 different considerations apply in imposing a consumption tax, such as value-added tax (VAT), as compared with a business profits tax. This article looks first at the common-law jurisdictional basis adopted by the English courts before considering the test developed by the UK courts in the 19th and early 20th centuries to determine whether the profits of a business based abroad nevertheless fall within the United Kingdom’s taxing jurisdiction. The UK approach to taxing the profits of UK business activity by a foreign business is then contrasted with the approach adopted for VAT as applied, in particular, to the remote supply of digital services from abroad. The United Kingdom’s rules currently derive from EU directives, and it remains open whether they will diverge over time following Brexit. KEYWORDS: JURISDICTION n NON-RESIDENTS n ELECTRONIC COMMERCE n UNITED KINGDOM n JURISPRUDENCE n VALUE-ADDED TAX

CONTENTS Introduction 1070 The Common-Law Jurisdictional Rules 1071 The Charge To Tax Business Profits 1074 Value-Added Tax 1078 Conclusion 1082

INTRODUCTION Fiscal sovereignty, the power to impose tax, is among the more important aspects of a state’s power over its citizens and any others who come within its borders. Gener- ally speaking, a state has an exclusive jurisdiction to tax within its territory, and its scope to levy tax extraterritorially is limited, both by international law and conventions and by practical considerations of the state’s ability to enforce its taxation claims through action in its own courts or those of other states. This article provides a brief review of the approach adopted by the United Kingdom to the issues raised in recovering tax on transactions entered into by UK persons with businesses based abroad. UK legislation is ordinarily construed, unless there is provision to the contrary, as applying only to persons or things over which Parliament has jurisdiction.1 Colquhoun v. Brooks2 is early authority for the basic proposition that the Income Tax Acts impose a territorial limit, such that either the source of the taxable income should be found within the United Kingdom or the person whose income is to be taxed must be resident there.

1 See, for example, Colquhoun v. Heddon (1890), 2 TC 621, at 626, per Lord Esher. In that case, a tax deduction for a premium paid to a US insurance company was denied because the Income Tax Acts extended only to UK insurance companies. Within the European Union, such a provision would now be contrary to EU law; see, for example, Safir v. Skattemyndigheten i Dalarnas Län, Case C- 118/96. 2 (1889), 2 TC 490. policy forum: much ado about not doing much n 1071

Business transactions entered into with UK persons are an obvious “source” in respect of which the United Kingdom might wish to claim tax. In principle, therefore, businesses based outside the United Kingdom that wish to conduct their business with UK residents can do so only if they are prepared to subject themselves to whatever taxes the United Kingdom chooses to impose on their transactions. From the United Kingdom’s perspective, its taxation choices may deter those abroad from transacting with its residents, and may encourage countermeasures abroad that may have an impact on UK businesses that wish to export their goods and services. Nevertheless, no state can necessarily tolerate a situation in which persons based abroad can conduct their business with the state’s residents from abroad—outside the state’s jurisdiction—and at the same time avoid the taxes that the state imposes on equivalent business conducted within its jurisdiction. Even within the European Union, where the freedom of movement of goods and services, of businesses and employment, and of capital has made such inroads into the traditional international tax regimes of its member states, there has been no significant undermining of the idea that within a state there should be a level playing field for all who choose to conduct their business with residents of that state. The rapid development of e-commerce, however, has substantially increased the opportunities for businesses abroad to transact with a state’s residents “remotely,” leaving the state to contemplate both its choice of taxes and, perhaps more importantly, its practical ability to levy and collect tax. With regard to collection, it may be possible to recover tax indirectly, as in the case of a withholding tax. Alternatively, the state may impose the tax directly on its residents (recognizing that they may in fact bear the tax supposedly charged on foreign businesses in any event). To the extent, however, that its target remains the foreign business that is choosing to transact with its residents, a state must adopt other methods. This is far from being a new problem for the 21st century. A suitable starting point for considering an appropriate response to the greater ability to conduct business transactions remotely may therefore be the rules entitling domestic courts to exercise their jurisdiction over persons based abroad, assumed here to be corporations incorporated outside the United Kingdom. If the domestic courts have no natural jurisdiction, special measures will be needed.

THE COMMON-LAW JURISDICTIONAL RULES The general principle of English law is that a court has jurisdiction to entertain a claim in personam (for example, the recovery of a debt) if the defendant is served with process in England or abroad in circumstances authorized by, and in the manner prescribed by, statute or statutory order. Jurisdiction in relation to many civil or commercial claims is a matter of EU law derived from the Brussels and Lugano conventions.3 For present purposes, however, it is the common-law rules that we need

3 The Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters, signed at Brussels on September 27, 1968 (“the Brussels convention”) and the Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial 1072 n canadian tax journal / revue fiscale canadienne (2020) 68:4 to consider in order to provide the appropriate background to the development of the United Kingdom’s taxation rules. The foundation of the English court’s jurisdiction in personam was that the person in question should be “present” in England, even if only temporarily so. The common-law rules concerning the presence of a corporation were subject to detailed consideration by the Court of Appeal in Adams v. Cape Industries Plc.4 The court noted that the residence or presence of a corporation can be a difficult concept but that a corporation that owned a place of business from which it carried on its business would be present for the purposes of in personam jurisdiction. Furthermore, a corporation that had no place of business might instead carry on its business through an agent. In that case, the question would be whether the commercial acts of the agent should be regarded as being undertaken by the agent in the course of the agent’s business or represented the conduct of the corporation’s business. From its review of the common-law authorities, the court derived three propositions relating to the “presence” of a foreign trading corporation for jurisdictional purposes:

1. The English courts would be likely to treat a foreign trading corporation as present within the jurisdiction only if either a. it had established and maintained at its own expense (whether as owner or lessee) a fixed place of business of its own in the other country, and for more than a minimal period of time it carried on its own business at or from such premises through its servants or agents (a “branch office” case); or b. a representative of the foreign corporation had for more than a minimal period of time been carrying on the foreign corporation’s business in the other country at or from some fixed place of business. 2. In either of these two cases, presence would be established only if it could fairly be said that the foreign corporation’s business (whether or not together with the representative’s own business) had been transacted at or from the fixed place of business. The determination of whether the representative had been carrying on the foreign corporation’s business or had been doing no more than carry on his own business would necessitate an investigation of the functions that it had been performing and all aspects of the relationship between the representative and the foreign corporation. 3. In particular, the following questions were likely to be relevant on such investigation: a. whether or not the fixed place of business from which the representative operates was originally acquired for the purpose of enabling the representative to act on behalf of the foreign corporation;

Matters, signed at Lugano on September 18, 1989 (“the Lugano convention”), as enacted in the Civil Jurisdiction and Judgments Act 1982, as amended by the Civil Jurisdiction and Judgments Act 1991. 4 [1990] Ch 433, at 523 - 31. policy forum: much ado about not doing much n 1073

b. whether the foreign corporation had directly reimbursed the representative for i. the cost of accommodation at the fixed place of business and/or ii. the cost of staff; c. what other contributions, if any, the foreign corporation had made to the financing of the business carried on by the representative; d. whether the representative was remunerated by reference to transactions —for example, by commission—or by fixed regular payments or in some other way; e. what degree of control the foreign corporation exercised over the running of the business conducted by the representative; f. whether the representative reserved i. part of its accommodation and/or ii. part of its staff for conducting business related to the foreign corporation; g. whether the representative displayed the foreign corporation’s name at its premises or on its stationery, and if so, whether the representative did so in such a way as to indicate that it was a representative of the foreign corporation; h. what business, if any, the representative transacted as principal exclusively on its own behalf; and i. whether the representative made contracts with customers or other third parties in the name of the foreign corporation, or otherwise in such a manner as to bind the foreign corporation, and if so, whether the representative required specific authority in advance before binding the foreign corporation to contractual obligations.

As the court emphasized, its list of questions was not exhaustive, and the answer to any of them was not necessarily conclusive. Indeed, every case was likely to involve “a nice examination of all the facts, and inferences must be drawn from a number of facts adjusted together and contrasted.”5 As a general principle, however, the court agreed that

[a] corporation resides in a country if it carries on business there at a fixed place of business, and, in the case of an agency, the principal test to be applied in determining whether the corporation is carrying on business at the agency is to ascertain whether the agent has authority to enter into contracts on behalf of the corporation without submitting them to the corporation for approval.6

5 Ibid., at 531, citing La Bourgogne, [1899] P 1, at 18, per Collins LJ. 6 Adams v. Cape Industries Plc, supra note 4, at 531, citing F. & K. Jabbour v. Custodian of Israeli Absentee Property, [1954] 1 WLR 139, at 146, per Pearson J. 1074 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Accordingly, the fact that a representative, whether with or without prior approval, never makes contracts in the name of the foreign corporation, or otherwise so as to bind it, is, a fortiori, a powerful factor in common law pointing away from the presence of the foreign corporation.

THE CHARGE TO TAX BUSINESS PROFITS The present-day expansion of tax treaties has tended to place the focus on the need for a permanent establishment (PE) to establish source-state entitlement to tax business profits. Indeed, the ability to manipulate business trading methods to avoid the creation of a PE was a reason for the introduction of the United Kingdom’s diverted profits tax and its concept of an “avoidedPE ,” as well as the adoption by the Organ- isation for Economic Co-operation and Development (OECD) of modifications to the definition of “permanent establishment” in article 5 of theOECD model treaty.7 Since the inception of the UK income tax, however, a non-resident person’s liability for tax on business profits has depended on whether a trade is being conducted “with” or “within” the United Kingdom. In other words, the domestic criterion on which the United Kingdom bases its jurisdiction to tax business profits depends on a foreign trader carrying on its trade within the United Kingdom. Given the common- law jurisdictional background against which the English courts were contemplating these matters, the approach that they adopted in answering this tax question is perhaps unsurprising. Thus, in Maclaine v. Eccott, Viscount Cave LC concluded as follows:

The question whether a trade is exercised in the United Kingdom is a question of fact, and it is undesirable to attempt to lay down any exhaustive test of what constitutes such an exercise of trade; but I think it must now be taken as established that in the case of a merchant’s business, the primary object of which is to sell goods at a profit, the trade is (speaking generally) exercised or carried on . . . at the place where the contracts are made. No doubt reference has sometimes been made to the place where payment is made for the goods sold or to the place where the goods are delivered, and it may be that in certain circumstances these are material considerations; but the most important, and indeed the crucial,[8] question is, where are the contracts of sale made?9

7 Organisation for Economic Co-operation and Development, Model Tax Convention on Income and on Capital: Full Version 2017 (Paris: OECD, 2019). The modifications to the PE definition were among a number of changes to the treaty resulting from the OECD’s base erosion and profit shifting (BEPS) project. 8 In Wilcock v. Pinto & Co, [1925] 1 KB 30, at 45, Scrutton LJ said that the courts had attached “preponderating importance” to where contracts are made, while going on to note that the combination of the making of contracts, delivery of goods, and payment in the United Kingdom had always resulted in the finding of a trade being conducted there. Lord Radcliffe in Firestone Tyre & Rubber Co. Ltd. v. Llewellin (HM Inspector of Taxes) (1957), 37 TC 111 (HL), did not “find great assistance in the use of [such] descriptive adjectives,” but nevertheless apparently acknowledged that “great importance” would attach to the place of sale, which other circumstances must outweigh. See ibid., at 142. 9 Maclaine v. Eccott, [1926] AC 424, at 432. policy forum: much ado about not doing much n 1075

Maclaine v. Eccott concerned a Javanese firm (Maclaine, Watson & Co) that dealt in the products of the Dutch East Indies and for which an associated London partnership, Maclaine & Co, acted as its general agent, selling goods in London directly or through brokers to other non-resident persons. Ordinarily, goods sold abroad between two non-resident persons would not be within the scope of UK income tax. In this case, however, the Javanese firm was selling its goods in the UK market through a regular UK agent. The fact that the purchasers were abroad—in some cases in Canada—and the Javanese firm delivered its goods directly to the purchasers abroad was irrelevant. It was conducting its trade in the United Kingdom through a regular UK agent based there.10 In the later case of F.L. Smidth & Co. v. F. Greenwood, Atkin LJ suggested that the relevant question as to whether trade was being conducted with or within the United Kingdom was, “[W]here do the operations take place from which the profits in substance arise?”11 His dictum, however, should be read in context. He was responding to an argument that if the contract was made and the goods were delivered outside the United Kingdom, that was conclusive to show that the trade was not exercised in the United Kingdom. The taxpayer was a Danish firm based in Copenhagen that manufactured and dealt in cement-making equipment and similar machinery, which it exported worldwide. It had an office in London with an employee who was a qualified engineer (“the engineer”). The engineer received inquiries about the Danish firm’s machinery, discussed the requirements of prospective purchasers, inspected sites for proposed installation, and took soil samples that he sent to Copenhagen for testing. He had general oversight of the erection of the machinery. The Danish firm did not advertise its equipment in the United Kingdom or maintain price lists or catalogues there. Contracts were negotiated between the UK customer and the Danish firm in Copenhagen. Once terms had been agreed, the contract was signed in the United Kingdom by the customer and forwarded (either directly or via the London office) to Copenhagen for signature. Contracts were governed by English law, and the equipment was delivered f.o.b. at Copenhagen. Payment was sometimes made directly to Copenhagen, but also in London. Atkin LJ rejected the idea that the place of contract is decisive. The only true test is that the question of where a trade is carried on is a question of fact, and in the case of a merchanting trade, where the sales contracts are regularly or habitually entered into is significant, albeit not decisive.12 The other elements to which Atkin LJ drew

10 “Foreign-to-foreign” sales were also at the heart of the argument in Firestone Tyre, supra note 8. On the basis of the contractual arrangements made, the US Firestone company (“Akron”) was found to be carrying on its trade in the United Kingdom through its UK manufacturing subsidiary (“Brentford”), by getting Brentford to fulfill orders by supplying tires to Akron’s foreign distributors. 11 (1922), 8 TC 193, at 204 (HL). 12 Ibid., at 203. Atkin LJ’s judgment at the Court of Appeal was not referred to in the House of Lords, where the Law Lords instead merely referred to Grainger & Son v. Gough, [1896] AC 1076 n canadian tax journal / revue fiscale canadienne (2020) 68:4 attention related either to the making of the sales contract or its performance. His focus remained on the sales contract, from which the profit in substance arises. On the facts, he concluded, in answer to his question, “To my mind there is no evidence in the present case of any other place than Denmark.”13 He continued by noting

[n]o doubt operations of importance take place here, orders are solicited, and the successful adapting of the goods bought for the purposes of the buyer’s business is supervised here. But in the words of Lord Watson in the case cited [Grainger & Son v. Gough, [1896] AC 325], at page 340:—“there may, in my opinion, be transactions by or on behalf of a foreign merchant in this country so intimately connected with his business abroad that without them it could not be successfully carried on, which are nevertheless insufficient to constitute an exercise of his trade here within the meaning of Schedule D,” and he instances the case of the purchase of goods here for the purpose of sale abroad [Sulley v. Attorney-General (1860), 2 TC 149].14

Thus, “operations of importance” conducted in the United Kingdom do not automatically amount to the exercise of trade in the United Kingdom even though they are “intimately connected” with the non-resident person’s trade in the form of the conclusion and performance of sales contracts abroad. A distinguishing feature is that, notwithstanding their importance to the business as a whole, many operations are often representative of the costs of doing business. The sales contracts, in contrast, represent the profit-generating activity. It is the profits that are sought to be taxed by demonstrating that the trade is being carried on in the United Kingdom. Lord Radcliffe was the only Law Lord in Firestone Tyre to allude expressly to Atkin LJ’s dictum.15 Lord Radcliffe emphasized that the answer to the statutory question of whether a taxpayer is exercising its trade within the United Kingdom is a matter of factual analysis, and the place in which contracts of sale are made in a merchanting business is not a substitute for factual inquiry and analysis.16 In particular, he said:

That would be to rewrite the words of the taxing Act, and could only be justified if there was nothing more in trading than the act of sale itself. There is, of course, much

325, and endorsed the judgment of Lord Sterndale MR at the Court of Appeal. Lord Sterndale (like Atkin LJ) had not accepted the taxpayer’s submission that if the contract was made and the goods were delivered outside the United Kingdom, that was conclusive to show that the trade was not exercised there. 13 F.L. Smidth & Co. v. F. Greenwood, supra note 11, at 204. 14 Ibid. 15 Lord Hanworth MR in Nielson, Andersen & Company v. Collins, [1927] 1 KB 780, alluded to Atkin LJ’s dictum in conjunction with Grainger & Son v. Gough, supra note 12, and Werle v. Colquhoun (1888), 2 TC 402 (CA), but the dictum was not referred to by the House of Lords in the Neilson case. 16 This much, of course, had been established in F.L. Smidth & Co. v. F. Greenwood, supra note 11, and Maclaine v. Eccott, supra note 9. policy forum: much ado about not doing much n 1077

more.[17] But, if “crucial”[18] does not mean as much as this, it cannot mean more than that the law requires that great importance should be attached to the circumstance of the place of sale. It follows, then, that the place of sale will not be the determining factor if there are other circumstances present that outweigh its importance or unless there are no other circumstances that can.19

The sales contracts and their performance therefore remain the focus of both Atkin LJ and Lord Radcliffe, as the operations from which the profits in substance arise. Lord Watson expressed the point as follows in Grainger & Son v. Gough:

I agree with the opinion expressed in [Erichsen v. Last] (8 Q.B.D., at 420) that whenever a foreigner, either by himself or through a representative in this country, “habitually” does and contracts to do a thing capable of producing profit, and for the “purposes of producing profit, he carries on a trade or business,” and that the profits or gains arising from these transactions in the United Kingdom are liable to income tax.20

Within a business there may be many operations (some essential and some just desirable) that facilitate the making of sales contracts or their performance, but that are neither capable of producing profits nor entered into for the purpose of doing so, in the absence of sales. In essence, the sales contracts represent the profit- generating activity, the results of which are the subject matter of the charge to tax, whereas the broader business framework is part of the cost of doing business. Arguably, Atkin LJ was saying nothing particularly radical or new. As Lord Her- schell had noted in Grainger & Son v. Gough,

[s]omething more [than soliciting and selling to UK customers] must be necessary in order to constitute the exercise of a trade within this country. How does a wine merchant exercise his trade? I take it, by making or buying wine and selling it again with a view to profit. If all that a merchant does in any particular country is to solicit orders, I do not think he can reasonably be said to exercise or carry on his trade in that country. What is done there is only ancillary to the exercise of his trade in the country where he buys or makes, stores, and sells his goods.21

Within that statement is the direction to consider the nature of the trade in question and have regard to its essential profit-making elements—where the profits in substance arise—including those elements to which Atkin LJ referred: the manufacture of the goods to be sold and the negotiation and execution of the sales contracts.

17 As F.L. Smidth & Co. v. F. Greenwood, supra note 11, illustrates, but where both the Court of Appeal (including Atkin LJ) and the House of Lords concluded that the “more” was not enough. 18 The word used by Lord Cave in Maclaine v. Eccott; see supra note 9 and the accompanying text. 19 Firestone Tyre, supra note 8, at 142. 20 Grainger & Son v. Gough, supra note 12, at 340. 21 Ibid., at 336. 1078 n canadian tax journal / revue fiscale canadienne (2020) 68:4

VALUE-ADDED TAX The jurisdictional basis of a consumption tax differs from that of a profits tax. Under a consumption tax, only transactions taking place within the state’s territory are ordinarily subject to the tax, and foreign transactions are disregarded even if carried out by the state’s nationals or residents.22 Conversely, domestic transactions are taxed regardless of whether they are effected by nationals or foreigners. Notwithstanding the United Kingdom’s cessation of membership in the Euro- pean Union, the UK value-added tax (VAT) rules currently remain those derived from EU directives and must be considered in that context. The United Kingdom does not yet have complete freedom to adopt rules of its own making. In the present context, what follows can only be a brief and incomplete outline of the complex rules. The basic position is that a supply of goods or services is within the scope of the UK VAT if it is made by a taxable person in the United Kingdom in the course or furtherance of its business. Therefore, the critical question is whether the supply is made (or is treated as being made) in the United Kingdom, rather than whether the supplier is based in the United Kingdom at the time that the supply is made.23 The two questions are not completely unrelated, because a UK-based supplier is likely to be making UK supplies; but it is a UK supply, not a UK establishment, that is important.24 A trader with a UK establishment that makes UK taxable supplies above the current registration threshold must register for VAT. A UK establishment exists if essential management decisions are made, and the business’s central administration is carried out, in the United Kingdom or if the business has a permanent physical presence, with the human and technical resources to make or receive taxable supplies, in the United Kingdom. An overseas trader that has no UK establishment may nevertheless still be obliged to register and account for VAT on its taxable supplies in the United Kingdom irrespective of their value. The obligation arises if

n the person makes UK taxable supplies, or there are reasonable grounds for believing that the person will make UK taxable supplies within 30 days beginning at the particular time; n those supplies (or any of them) are or will be made in the course or furtherance of a business carried on by the person;

22 Thus, in contrast to Maclaine v. Eccott, supra note 9, there would be no value-added tax (VAT) liability on the sale of goods originating in Java and delivered to Canada, even if the sale was arranged through a regular UK agent. 23 The place-of-supply rules, especially those for international services, are detailed and complex, and vary according to whether the supply is to a business customer or a final consumer. This article focuses on the current rules relating to the remote supply of digital services. 24 A UK-established person that makes only foreign supplies may still be entitled to register voluntarily to benefit from the recovery of input VAT. policy forum: much ado about not doing much n 1079

n the person has no business establishment, or other fixed establishment, in the United Kingdom in relation to any business carried on by the person; and n the person is not VAT registered.

In general terms, the obligation to register in the United Kingdom does not arise on the exporting of goods to the United Kingdom because that will not usually amount to a taxable UK supply.25 Similarly, if a UK recipient of services accounts for the UK VAT under the reverse charge procedure (discussed below), the overseas supplier will not need to register in relation to those services. Remote supplies of services to UK consumers, however, can create an obligation to register.26 Failure to register attracts penalties. HM Revenue & Customs (HMRC) can require overseas traders to appoint a UK tax representative, who can then be made jointly and severally liable for any UK VAT due. In relation to a non-resident trader that sells goods in the United Kingdom via an online marketplace, HMRC has had powers since 2016 that are designed to ensure compliance with the registration requirement. In the first instance, HMRC will normally contact any business that it considers to have failed to register, requesting that the business adhere to the registration requirement. If the overseas seller fails to do so, HMRC can put the operator of the online marketplace through which the seller is trading on notice that it could be held jointly and severally liable for any UK VAT due. An “online marketplace” is a website, or any other means by which information is made available over the Internet, through which persons other than the operator are able to offer goods for sale. The “operator” is the person that controls access to, and the content of, the online marketplace. Online marketplaces are required to display the UK VAT registration number of the non-UK business on their website and check the validity of the numbers that are provided. In addition, online sales of goods in the United Kingdom by third-country traders where the order is fulfilled by another person (a fulfillment house) have been tackled by requiring fulfillment houses (since 2017) to register with HMRC, (since 2018) to be approved, and (since 2019) to undertake due diligence checks and keep records of third-country traders. Any business that does trade without approval risks the imposition of a large financial penalty and possible criminal sanctions. Whether a supply is one of goods or of services is important because different rules apply to the place and time of supply in each case, and the treatment of international supplies is different. In the case of goods, their physical existence and the need for delivery usually suffice to identify the place of supply and therefore the EU

25 The obligation to pay VAT on UK imports of most goods from non-EU countries normally falls on the importer. See below for a discussion of online sales of goods using a fulfillment house or online marketplace. Different rules apply to cross-border acquisitions of goods within the European Union. 26 Subject to being registered in another EU member state under the EU “mini one-stop shop” (MOSS) scheme for broadcasting, telecommunications, and electronically supplied services (BTE services) (discussed below). 1080 n canadian tax journal / revue fiscale canadienne (2020) 68:4 state (including the United Kingdom) that has the jurisdiction to charge and recover VAT. To the extent that goods never enter or leave the United Kingdom, they will be outside the scope of the UK VAT. Services, however, can present more significant problems, in particular when they can be provided remotely. The EU place-of-supply rules for international services have evolved over a number of years, with particular changes in 2010, 2011, 2013, and 2015. The usual rule for cross-border services since 2010 (as for goods) distinguishes business-to-business (B2B) supplies from business-to-customer (B2C) supplies. In the case of cross-border services (as for goods), the VAT liability for most B2B supplies depends on where the business customer is based rather than where the supplier is based. In the case of services, a business customer may then have to engage a reverse charge procedure on receipt of the service, under which the supply of the service will be included in the customer’s business supplies on which it must account for VAT. If the business is fully taxable, the customer will obtain credit for the VAT against its subsequent supplies. Otherwise, the customer will have to account for and absorb the VAT as it would on an equivalent domestic supply. The existence of the reverse charge procedure does not, however, prevent an overseas supplier from registering for VAT in the United Kingdom. If it does register, the supplier must invoice UK VAT in the normal way, and the business recipient is not then required to account for VAT under the reverse charge procedure. By contrast, for B2C supplies, it is the location of the supplier that usually determines the VAT liability, so that VAT will be charged and accounted for by the supplier under the applicable domestic rules. If the place of supply is outside the United Kingdom, the services will ordinarily be outside the scope of the UK VAT. However, where the place of supply is within the United Kingdom, the services will be subject to normal UK VAT rules, and the supply will be standard-rated, zero-rated, or exempt (or subject to a reduced rate of VAT) as if the services were supplies between two UK parties. Since 2015, however, all digital services have been subject to VAT in the jurisdiction of the customer, regardless of the customer’s status (that is, whether business or consumer) or the location of the supplier (EU or non-EU).27 The only exceptions are for services that are effectively used and enjoyed outside the European Union. Further changes to the EU e-commerce rules were proposed in directives issued in 2017 and 2019, to be implemented in 2019 and 2021, respectively.28 Digital services

27 For a description of the evolution of the EU rules for digital services over a longer time frame, see Marie Lamensch, European Value Added Tax in the Digital Era: A Critical Analysis and Proposals for Reform, IBFD Doctoral Series vol. 36 (Amsterdam: International Bureau of Fiscal Documentation, 2015), chapter 2. 28 For a summary of the proposals, see Mariya Senyk, The Origin and Destination Principles as Alternative Approaches Towards VAT Allocation: Analysis in the WTO, the OECD and the EU Legal Frameworks, IBFD Doctoral Series vol. 53 (Amsterdam: International Bureau of Fiscal Documentation, 2020), appendix 2. See also European Commission, Explanatory Notes on VAT E-Commerce Rules, September 2020 (https://ec.europa.eu/taxation_customs/sites/ taxation/files/vatecommerceexplanatory_notes_30092020.pdf ). policy forum: much ado about not doing much n 1081 for these purposes are broadcasting, telecommunications, and electronically supplied services (BTE services), which include the following:29

n “the supply of audio and audio-visual content for simultaneous listening or viewing by the general public on the basis of a programme schedule by a person that has editorial responsibility”; n “live streaming through the internet if broadcast at the same time as transmission by radio or television”; n “the transmission of signals of any nature by wire, optical, electromagnetic or other system” (such as fixed and mobile telephone services; switching of voice, data, and video; voice over Internet protocol [VoIP]; voice mail; call waiting, call forwarding, caller identification, three-way calling, and other call management services; paging services; and access to the Internet); n supplies of music, films, and games, programs on demand, and online magazines; n website supply or web-hosting services; n “distance maintenance of programmes and equipment” and “supplies of software and software updates”; and n “advertising space on a website.”

“Electronically supplied” covers e-services that are automatically delivered over the Internet, or an electronic network, where there is minimal or no human intervention. However, using the Internet or some electronic means of communication just to communicate or facilitate trading does not necessarily mean that a business is supplying e-services. Where BTE services are supplied to business customers, the reverse charge procedure may still apply. For B2C supplies of such services, however, the supplier is required to register and charge VAT. Unlike the distance selling rules for goods, a supplier of BTE services creates an obligation to register and charge VAT based on the location of the customer. Thus, a single sale to a consumer in the United King- dom involves the need to account for UK VAT in relation to that sale, either by registering for VAT or by using a special scheme known as “the mini one-stop shop” (MOSS). This scheme allows registration in a single EU member state where the VAT will be collected, thereby avoiding the need to register in every member state in which such services are supplied. In a recent simplification, the place-of-supply rule for BTE services has been modified to provide a de minimis exception where the total value of such services is less than €10,000. In that case, the place of supply will be the member state in which the supplier is established. This allows small businesses to avoid the need to register for VAT or use the MOSS system.

29 See HM Revenue & Customs, “Guidance—VAT Rules for Supplies of Digital Services to Consumers in the EU,” December 19, 2014 (updated November 19, 2018) (www.gov.uk/ guidance/the-vat-rules-if-you-supply-digital-services-to-private-consumers#euvatrules). 1082 n canadian tax journal / revue fiscale canadienne (2020) 68:4

CONCLUSION As this brief review illustrates, different taxes require different solutions to the issues of cross-border business activity. The real issue is not whether transactions entered into between a foreign corporation and its UK-resident customers represent conduct of the foreign company’s business; plainly, they do. The essential question is whether the manner in which those transactions are effected render the foreign company sufficiently subject to the United Kingdom’s jurisdiction to justify the imposition of the tax in question, by establishing a sufficientUK presence to enable the courts to assume jurisdiction over the foreign corporation or to enable the courts to conclude that the foreign corporation is conducting its business within the United Kingdom rather than just with UK customers. The relevant tests for jurisdiction and for income tax liability may not be identical, but they both illustrate the need for something more than the provision of supplies to UK customers. That does not, however, hold true for a consumption tax such as VAT. Inevit- ably, special rules are needed to ensure that foreign suppliers cannot avoid any liability for VAT by making remote supplies to UK customers. An objective of the EU VAT system is to create a level playing field and avoid distortion of competition between suppliers of the same or similar goods and services. This objective is mainly achieved by ensuring that VAT is paid in the member state in which, under EU rules, the consumer is regarded as using the supply of goods or services. In theory, it may be thought that a consumption tax should be capable of being collected directly from consumers in the case of cross-border supplies; in practice, as in many other areas of the tax system, collection may be difficult to police satisfactorily, neces- sitating alternative measures targeted at the better organized and smaller number of suppliers. canadian tax journal / revue fiscale canadienne (2020) 68:4, 1083 - 1122 https://doi.org/10.32721/ctj.2020.68.4.fon

Finances of the Nation PROVINCIAL DEBT SUSTAINABILITY IN CANADA: DEMOGRAPHICS, FEDERAL TRANSFERS, AND COVID-19 Trevor Tombe*

For almost 60 years, the Canadian Tax Foundation published an annual monograph, Finances of the Nation, and its predecessor, The National Finances. In a change of format, the 2014 Canadian Tax Journal introduced a new “Finances of the Nation” feature, which presents annual surveys of provincial and territorial budgets and topical articles on taxation and public expenditures in Canada. The underlying data for the Finances of the Nation monographs and for the articles in this journal will be published online in the near future. In this article, Trevor Tombe examines the sustainability of Canada’s public debt in the face of steadily rising provincial debt, a severe economic shock from COVID-19, and mounting health-care costs associated with an aging population. He finds that while the federal debt is solidly sustainable, despite a large increase owing to COVID-19, the debt burden of most provincial governments is not. He discusses some of the policy options available to improve fiscal outlooks, focusing in particular on reform of federal transfers. KEYWORDS: PUBLIC DEBT n FEDERAL-PROVINCIAL n DEMOGRAPHY n TRANSFER PAYMENTS n FISCAL POLICY

CONTENTS Introduction 1084 A Primer on Public Debt Dynamics 1089 Simple Debt Sustainability Arithmetic 1090 A General Framework for Debt Sustainability Analysis 1094 The Effect of a Temporary Fiscal Shock 1096 Projecting Government Finances over the Long Run 1098 The Projection Model 1098 Projecting Future Health-Care Expenditures 1100 The Long-Run Fiscal Sustainability of Canada’s Provinces 1103 What Affects Provincial Fiscal Gaps? 1106 Policy Options To Improve Fiscal Outlooks 1110

* Of the Department of Economics, University of Calgary (e-mail: [email protected]). I gratefully thank Ken McKenzie, Michael Smart, and numerous participants at a Finances of the Nation seminar for their valuable feedback. I am also grateful to the University of Calgary’s Advanced Research Computing resources that were used in this analysis.

1083 1084 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Federal Transfers and Provincial Financial Sustainability 1112 Index CHT Growth to Demographics 1117 Supplement the CHT To Cover All Demographic Costs 1117 The Effect of COVID-19 and Other Macroeconomic Developments 1119 Conclusion 1120

INTRODUCTION Rising debt in the past, unprecedented fiscal and economic disruptions in the present, and aging populations in the future all raise questions around the long-term viability of Canada’s public debt. Federally, emergency spending measures introduced in 2020 in response to the COVID-19 pandemic will add more to the federal debt than has occurred in any other single fiscal year since the Second World War. Given the extent and magnitude of the crisis, the government’s response was appropriate. But this short-term fiscal shock is dwarfed by a slower-moving and significantly larger challenge from aging populations. Provincial governments face mounting health-care costs, including but by no means limited to the added burden of managing the pandemic, and all governments face potentially slower rates of economic growth. This article outlines a simple but powerful approach to quantifying long-run fiscal challenges in Canada and pays particularly close attention to provincial governments. Building on readily available data from Statistics Canada, I develop a rich model of future public finances and explore a wide variety of scenarios. I find that provincial government finances are strained, while federal finances remain strong, despite recent deficits. Options abound for most governments to improve their capacity to meet present and future challenges. The analysis highlights the importance of federal-provincial transfers for provincial debt sustainability and proposes several reforms to help mitigate fiscal pressures. Examining in detail the future trajectory of Canada’s public debt is important, especially now. Following the largest economic and fiscal shock in generations, overall debt levels are set to approach 110 percent of gross domestic product (GDP) in 2020—a historically high level exceeded only during the Great Depression and the Second World War. To put this in context, I display Canada’s overall public debt levels since 1870 in figure 1(a). Only once before in Canada’s post-war experience has debt exceeded 100 percent of GDP, and this moment in the mid-1990s ushered in a period of substantial fiscal consolidation. Going into the COVID-19 crisis, though, Canada’s two orders of government face very different fiscal situations. Separating federal and provincial debt in figure 1(b), we see that Canada’s provincial governments have been continuously and systematically increasing their debt levels over the past 60 years. Overall, provincial debt has roughly tripled as a share of GDP since 1960, rising by nearly 40 percentage points. Meanwhile, the federal debt rises and falls with fiscal developments but is lower today than in 1960, and even the COVID-19 shock is unlikely to bring debt to levels seen in the 1990s. Not only do provincial governments account for a larger share of the public debt today, but the coming years will see substantial fiscal pressures from an aging population that will only add to this burden. The share of Canada’s population aged 65 and over may finances of the nation n 1085 increase from 18 percent today to nearly 24 percent by 2040, and the share aged 75 and over may double from 7 percent today to 14 percent over the same time.1 With provinces responsible for health-care delivery, incremental costs from this aging will be disproportionately borne by them. Recent debt increases and the coming demographic challenge mean that the future of government debt sustainability in Canada will be determined by the fiscal health of the provinces. Before providing details, it is worth appreciating intuitively what “sustainable” fiscal policy means. Today’s debt is potentially a burden on the future, and additional borrowing to cover budget deficits adds to that burden. Public finances are sustainable if the future burden is manageable under current fiscal policy. In a growing economy, where the ability to carry and service debt is rising over time, the burden of debt is captured by interest costs as a share of total income. So continuously increasing the stock of debt faster than the pace of economic growth cannot continue indefinitely. At some point, abrupt changes in fiscal policy—either increasing revenues or decreasing program spending—would be required to avoid default. A stable debt-to-GDP ratio is therefore a common and useful metric of sustainable public finances. Interest rates and growth rates are also important determinants of sustainability. If interest rates exceed growth rates, future revenues must exceed program spending by enough to repay current debt. If interest rates are less than growth rates, governments can sustainably borrow to cover program spending above revenues—but assuming a reasonable time horizon, the extent of such borrowing is limited. To clarify these issues, I provide a detailed framework that makes it possible to project future government revenues and expenditures, and to quantify any gap between current fiscal policy and an alternative sustainable policy. If current fiscal policy is not sustainable, increases in revenues or decreases in program spending are required. The magnitude of the required fiscal adjustment is commonly known as “the fiscal gap.”2 Although there are many complexities and uncertainties to consider, in a world where interest rates are roughly equal to economic growth

1 Author’s calculation from Statistics Canada tables 17 - 10 - 0005 - 01 (formerly CANSIM table 051 - 0001), “Population Estimates on July 1st, by Age and Sex”; and 17 - 10 - 0057 - 01 (formerly CANSIM table 052 - 0005), “Projected Population, by Projection Scenario, Age and Sex, as of July 1 (x 1,000).” These are projections around a range of potential outcomes. For perspective, the share of the population aged 65 and over ranges from slightly more than 21 percent in the slow-aging scenario to nearly 26 percent in the fast-aging scenario. Historically, such projections have proved to be useful guides. In the mid-1980s, for example, the Canada Pension Plan “case” projection for 2020 was for a population of nearly 35 million and an age 65+ share of 16.2 percent, both not too far off the actual numbers. For this and many other projections, see Canada, Report of the Royal Commission on the Economic Union and Development Prospects for Canada, vol. 2 (Ottawa: Department of Supply and Services, 1985), 52 - 60, tables 7 - 31 to 7 - 35. 2 The term was originally put forward by Alan J. Auerbach, “The U.S. Fiscal Problem: Where We Are, How We Got Here and Where We’re Going,” in Stanley Fischer and Julio J. Rotemberg, eds., NBER Macroeconomics Annual 1994 (Cambridge, MA: MIT Press, 1997), 141 - 86, at 174. Assessing government policy by reference to the fiscal gap is now a well- established approach to fiscal policy analysis. 1086 n canadian tax journal / revue fiscale canadienne (2020) 68:4

FIGURE 1 Government Debt-to-GDP Ratios in Canada (a) General consolidated government debt, 170-201 and 2020 (forecasted) 150

IM forecast for 2020 100 Percent 50

0 10 100 120 140 160 10 2000 2020

(b) ederal and provincial government liabilities, 161-201 0

40 Percent

0 160 170 10 10 2000 2010 2020 ederal government Provincial governments (Figure 1 is concluded on the next page.) rates, a government’s fiscal gap will be roughly equal to its average projected annual primary deficit (the deficit net of interest payments). Projecting revenues and program expenditures is therefore central to the exercise. Governments with projected future deficits face a positive fiscal gap and unsustainable fiscal policy unless revenues increase or spending decreases. The reverse is true for governments with projected future surpluses, as we will see is the case for Canada’s federal government. To construct these projections, I compile detailed provincial data on 12 revenue and 6 expenditure components. I also model federal finances, with additional detail for unique federal areas of expenditure, such as provincial transfers, old age security, child and family benefits, employment insurance (EI), defence spending, and so on. Each budget component is then projected forward by forecasting growth rates of underlying tax bases, revenue sources, and cost pressures using demographic pro jections from Statistics Canada, health-care expenditure data from the Canadian Institute for Health Information (CIHI), population projections from the Office of finances of the nation n 1087

FIGURE 1 Concluded GDP = gross domestic product; IMF = International Monetary Fund. Note: This figure displays data on general government debt in Canada as a share of gross domestic product from 1870 to 2019 and a forecast for 2020. This includes both federal and subnational debt. Federal and provincial governments are separated for the period 1961-2019. Sources: Debt-to-GDP ratio for 1870 to 2016 is from Òscar Jordà, Moritz Schularick, and Alan M. Taylor, “Macrofinancial History and the New Business Cycle Facts,” in Martin Eichenbaum, Erik Hurst, and Jonathan A. Parker, eds., NBER Macroeconomics Annual 2016 (Chicago: University of Chicago Press, 2017), 213-63. The debt ratio is updated from 2017 to 2019 using the change in general government debt from Statistics Canada table 36-10-0580-01 (formerly CANSIM 378-0121), “National Balance Sheet Accounts (x 1,000,000)”; nominal GDP growth rates for 2017 and 2018 are from Statistics Canada table 36-10-0222-01 (formerly CANSIM 384-0038), “Gross Domestic Product, Expenditure-Based, Provincial and Territorial, Annual (x 1,000,000).” Forecast for 2019 is from Canada, Department of Finance, Federal Economic and Fiscal Snapshot 2020 (Ottawa: Department of Finance, July 8, 2020); forecast for 2020 is from the International Monetary Fund, World Economic Outlook Update (Washington, DC: IMF, June 2020). Separate provincial and federal liabilities (book value) are from Statistics Canada tables 36-10- 0535-01 (formerly CANSIM 378-0076) (Archived), “National Balance Sheet, Provincial Governments, Annual, 1961-2011 (x 1,000,000)”; 36-10-0533-01 (formerly CANSIM 378-0074) (Archived), “National Balance Sheet, Federal Government, Annual, 1961-2011 (x 1,000,000),” for 1961 to 1989; and 36-10-0580-01 (formerly CANSIM 378-0121), “National Balance Sheet Accounts (x 1,000,000)” for 1990 to 2019. GDP data are as shown in graph (a), supplemented with Statistics Canada table 36-10-0325-01 (formerly CANSIM 384-0015) (Archived), “Provincial Gross Domestic Product (GDP), Expenditure-Based, Provincial Economic Accounts, Annual, 1961-1980 (x 1,000,000),” for 1961 to 1980.

the Parliamentary Budget OfficerPBO ( ), and numerous other sources. The result is a detailed interconnected model of Canadian government finances. Though in some cases necessarily abstract, the model incorporates sufficient complexity to reveal novel interactions between orders of government, program designs, and economic and fiscal shocks, all within several informative scenarios. The analysis reveals that provincial government finances are not sustainable, with the notable exception of Quebec. Across a projected 75-year time horizon, provincial revenues average nearly 18 percent of GDP versus program and capital spending of nearly 21 percent. These imbalances, appropriately discounted to present-value terms, are equivalent to roughly 170 percent of GDP in debt obligations today. To ensure that debt levels at the end of the projection period are no higher than the levels today, revenues must increase or expenditures must decrease by an immediate and permanent amount equivalent to 2.7 percent of GDP per year. This positive fiscal gap for provincial governments, however, is more than fully offset by a negative fiscal gap of 2.8 percent ofGDP for the federal government. Thus, the general fiscal situation in Canada is sustainable, though there is an imbalance between the two orders of government. The analysis also reveals that the aging population fully accounts for the provincial challenge, with 40 percent being attributable to slowing economic growth and 60 percent to rising health-care costs. 1088 n canadian tax journal / revue fiscale canadienne (2020) 68:4

There is, however, significant variation across provinces. Quebec’s fiscal situation is sustainable, largely owing to that province’s higher-than-average taxes. The situation in Alberta and Saskatchewan is not sustainable; both provinces are projected to experience relatively large fiscal gaps, at 4.8 percent and 3.4 percent ofGDP respectively. These gaps are accounted for by far lower-than-average taxes rather than projected health-care expenditures, which are lower in these provinces than in any other. The Atlantic provinces also face unsustainable finances, although the Maritimes are significantly aided by equalization to an extent that Newfoundland and Labrador is not. The latter province faces the largest fiscal gap by far, at 9.4 percent of GDP. Above-average spending, a rapidly aging population, and the slowest projected economic growth in Canada underlie the significant challenge faced by Newfoundland and Labrador. Finally, while the long-run challenges are large, there is significant concern around the abrupt increase in government debt resulting from COVID-19, especially at the federal level. I find that federal finances remain strongly sustainable despite the shock, and provincial finances are actually improved in the long run as a result. Behind this seemingly counterintuitive result is the way the Canada health transfer (CHT) formula operates in the face of a large short-term shock. I will explain in the main text, but this underscores why federal-provincial transfer arrangements are central to an understanding of long-term provincial sustainability. Before proceeding further, some important caveats are in order. All projections in this article are subject to uncertainty and are not themselves predictions. Nor do these results guide what governments should or should not do to adjust the path of future finances. Instead, they illustrate a potential path that current policy is on and quantify the size of long-term gaps between revenues and program spending. Understanding this is necessary to guide tax and expenditure decisions today. This exercise also reveals how sensitive long-run finances are to changes in underlying assumptions. Indeed, exploring a variety of scenarios is potentially this exercise’s most valuable contribution. Consider a few examples. First, in line with historical experience, the baseline projections incorporate health-care-specific inflation of 1 percentage point above the economy-wide inflation rate of 2 percent per year. But if health-care-specific inflation falls to just 0.5 percentage points above average, for example, the provincial fiscal gap falls from 2.7 percent to 1.3 percent ofGDP . Simi- larly, I find that provincial revenues grow more slowly than the economy overall. But if instead own-source revenues grow in line with GDP, the aggregate fiscal gap declines to 1.5 percent. This analysis therefore demonstrates that gradual health- care spending restraint, combined with modestly higher revenue growth, can fully address the long-term challenges of provincial governments. The analysis also reveals an important role for federal transfers. Current fiscal arrangements contribute an average of 3.4 percent of GDP to provincial finances across the 75-year horizon, and programs like equalization are particularly im portant for the sustainability of lower-income provinces. I propose two potential reforms to federal transfers to help cover aging-related health-care costs. I find that both of the proposed measures have meaningful effects on fiscal gaps and may finances of the nation n 1089 therefore be potentially important reforms to consider. Finally, certain provinces face particularly large fiscal challenges that modest reforms cannot overcome. Alberta, Saskatchewan, and Newfoundland and Labrador face fiscal gaps that persist across nearly all scenarios examined here. These provinces should therefore consider revenue and expenditure changes to address this. Delaying action will merely increase the scale of adjustment required. To be clear, this article is not the first to examine the long-run fiscal future of Canada’s provincial governments. The most important contribution to this area of research is the PBO’s Fiscal Sustainability Report 2020.3 The PBO’s analysis is a timely, thorough, and important examination of all subnational finances; however, it aggregates provincial and municipal finances. This article complements the PBO’s work by focusing on provincial governments only, by enriching the level of detail behind government budget projections, and by including capital spending. By separately modelling a dozen different revenue categories, this article finds a notably larger fiscal gap than that forecasted by thePBO , since I find that provincial own-source revenues will grow more slowly. There is also an important place for analysis that excludes municipalities, since in normal circumstances their long-term finances are sustainable by construction. Local governments do not generally set tax rates and fee levels separately from expenditure decisions. Standard practice is for expenditures to be determined by local councils and then property tax rates are endogenously determined to mechanically balance municipal budgets. In that sense, primary balances are zero by construction, and therefore fiscal gaps are also zero. Finally, the budget model developed here and the wide varieties of scenarios it explores will not only form the basis of the current analysis, but it, and regular updates to it, will also be made available to facilitate future research. Before turning to this detailed model of provincial finances, I begin with a primer on public debt dynamics. Much of this will build on, and contribute to, well- established practices in the literature.4

A PRIMER ON PUBLIC DEBT DYNAMICS At its core, long-run debt sustainability analysis asks two simple questions: Will public debt grow to unmanageable levels? If so, what policy changes are required? To be clear, what is specifically meant by “unmanageable” is a moving target and varies both across jurisdictions and over time. In the Canadian context, Alberta defaulted on its debt in the mid-1930s, with a debt level that was roughly one-third of the province’s GDP. Today, most provinces have debt levels at or above that level

3 Office of the Parliamentary Budget Officer,Fiscal Sustainability Report 2020 (Ottawa: Office of the Parliamentary Budget Officer, 2020). 4 A useful guide to debt sustainability analysis is Julio Escolano, A Practical Guide to Public Debt Dynamics, Fiscal Sustainability, and Cyclical Adjustment of Budgetary Aggregates, International Monetary Fund Technical Notes and Manuals no. 10/02 (Washington, DC: IMF, Fiscal Affairs Department, January 2010). 1090 n canadian tax journal / revue fiscale canadienne (2020) 68:4 with no reasonable risk of default. Internationally, Japan’s debt level is on track to approach 270 percent of its GDP in 2020,5 which is roughly double the level that led Greece into a debt crisis nearly a decade ago. Various factors—interest rates, economic growth rates, domestic versus international holdings, the currency that public debt is denominated in, volatility, and more—all matter. But for Canada’s provinces we must also consider the role of the federal government and fiscal transfers. And behind all such complexities is some basic arithmetic.

Simple Debt Sustainability Arithmetic Public debt rises if spending exceeds revenue. Dollars out, after all, must be balanced by dollars in, from either revenues or new borrowing.6 This is summarized by the government’s budget constraint, expressed as

Gt + rt × Dt - 1 = Rt + DDt, (1) where Gt is program expenditures, rt is the rate of interest on debt Dt - 1 in the last period, Rt is government revenue (from all sources), and DDt = Dt - Dt - 1, the change in public debt (that is, the deficit). If spending (the left-hand side of the equation) exceeds revenue (Rt), borrowing increases debt as DDt > 0. The reverse holds if revenue exceeds spending. Importantly, changes in debt in one period affect the government’s future budget, because the level of debt affects interest costs and therefore future spending. There is therefore a risk that debt may snowball and grow beyond a government’s ability to service it. How much debt can sustainably increase over time depends on economic growth. Without economic growth, public debt cannot indefinitely grow more quickly than the interest rate. If current public debt is rolled over, without principal 2 ever being paid off,D 0 today becomes D0(1 + r) next year, and D0(1 + r) in the year after, and so on. This exponential growth implies that debt will eventually grow beyond the public’s ability to service it. But with a growing economy, the ability to service debt is itself increasing. In this case, if debt is rolled over indefinitely, the burden of debt D0 today becomes D0(1 + r)/(1 + g) next year (where g is the rate of 2 2 growth in the economy), D0(1 + r) /(1 + g) in the year after, and so on. If these values are declining over time—say, because economic growth g exceeds the interest rate r—the debt ratio is mechanically sustainable in perpetuity. Looking at debt ratios (that is, debt to GDP) rather than levels not only allows for easy comparison of debt burdens over time and across jurisdictions, but is also the relevant measure for long- run sustainability analysis.

5 International Monetary Fund, World Economic Outlook Update (Washington, DC: IMF, June 2020). 6 Another option is printing money. This option comes with a risk of rising inflation if it is not used in moderation, so most advanced economies shy away from it, and central banks operate (largely) independently of fiscal authorities. Since printing money is unavailable to provincial governments (the focus of this article), I do not consider it in the analysis. finances of the nation n 1091

Dividing the government’s budget constraint by nominal GDP, and rearranging the terms, yields an expression that governs how debt ratios evolve,

____1 + r t dt = × dt − 1 − pt , (2) (1 + g t) where dt is total debt as a share of GDP and pt is the government’s primary budget balance (revenue Rt minus program spending Gt ) as a share of GDP. If the primary budget is balanced, revenues cover all program spending, and the debt ratio in the next period will evolve over time according to (1 + rt )/(1 + gt). If interest rates exceed growth rates, this ratio will be larger than 1, and the debt burden will rise. If interest rates equal growth rates, the debt ratio will remain stable. And if interest rates are lower than growth rates, the debt ratio will gradually decline to zero over time. The expression in equation 2 also allows one to appreciate what factors matter for long-run sustainability. In the next section, I will unpack this in more detail, but if the debt ratio, interest rates, and growth rates are each stable, so too is the burden of debt. This is sustainable. To achieve this, equation 2 reveals that to achieve dt = dt - 1, the government must run a primary balance equal to

r - g p * = _ × d. (3) (1 + g)

If interest rates exceed growth rates, the government must run a primary surplus to compensate and maintain the debt ratio at d. If the primary balance falls short, a fiscal gap exists. In this case, increases in revenues or decreases in program spending may be required. And if debt levels increase—say, owing to a short-term shock— and if r > g, the government will require a larger primary surplus to ensure sustainability. In this sense, (r - g) captures the fiscal cost of public debt. But if interest rates are less than growth rates, larger debt may create a fiscal benefit by allowing the government to sustainably run larger primary deficits. I will return to this point shortly, but much therefore depends on the interest :growth differential (r - g). Historical experience provides insight around what this differential normally is. Gathering data from a variety of sources, I display the full history of Canada’s long-term interest rates and economic growth rates in figure 2(a). I abstract from the periodic ups and downs over the business cycle to reveal the underlying trend rate of growth. Growth is typically more volatile than long-term interest rates, but robust growth that exceeds interest rates is not uncommon—in fact, it is the norm. The trend rate of annual nominal GDP growth since Confederation has averaged over 6.2 percent. Meanwhile, nominal long-term borrowing rates averaged just over 5 percent, meaning that the average interest : growth differential was −1.2 percent. But, as is evident in figure 2(b), there is significant variation around this average. In the years between 1945 and 1979, the differential averaged over −4 percent, while between 1980 and 2000 it averaged over 3 percent. Since 2000, the average differential has been modestly negative at −0.5 percent. This phenomenon is not unique to Canada. Recent research suggests that interest rates fall below 1092 n canadian tax journal / revue fiscale canadienne (2020) 68:4

FIGURE 2 Comparison of Borrowing Costs and Growth Rates in Canada, 1870-2019 (a) Long-run nominal interest rates and economic growth rates 15

Percent 0

−5

−10 10 100 120 140 160 10 2000 2020 Interest rates Trend growth rates

(b) The interest:growth diﬀerential 15

0 Percent −5

−10

−15 10 100 120 140 160 10 2000 2020 Note: This figure displays the long-run nominal interest rates and the trend annual nominal economic growth rates in Canada from 1870 to 2019. The cyclical component of gross domestic product growth is removed using a Hodrick-Prescott filter. Sources: Long-run nominal interest rates for 1870 to 1975 and GDP growth for 1870 to 1981 are from Òscar Jordà, Katharina Knoll, Dmitry Kuvshinov, Moritz Schularick, and Alan M. Taylor, “The Rate of Return on Everything, 1870-2015” (2019) 134:3 Quarterly Journal of Economics 1225-98 (https://doi.org/10.1093/qje/qjz012); and Òscar Jordà, Moritz Schularick, and Alan M. Taylor, “Macrofinancial History and the New Business Cycle Facts,” in Martin Eichenbaum, Erik Hurst, and Jonathan A. Parker, eds., NBER Macroeconomics Annual 2016 (Chicago: University of Chicago Press, 2017), 213-63. Interest rates for 1976 to 2019 are from Statistics Canada table 10-10-0122-01 (formerly CANSIM 176-0043), “Financial Market Statistics, Last Wednesday Unless Otherwise Stated, Bank of Canada,” vector v122544. Growth rates for 1982 to 2018 are from Statistics Canada table 36-10-0222-01 (formerly CANSIM 384-0038), “Gross Domestic Product, Expenditure- Based, Provincial and Territorial, Annual (x 1,000,000),” vector v62787312, updated to 2019 using Canada, Department of Finance, Federal Economic and Fiscal Snapshot 2020 (Ottawa: Department of Finance, July 8, 2020). finances of the nation n 1093 economic growth rates more frequently than the reverse, often for long stretches.7 These interest : growth differentials also imply that the fiscal cost of public debt can fluctuate and potentially be negative (that is, become a fiscal benefit). I plot this in figure 3. Looking forward, both growth rates and interest rates may continue recent trends. Indeed, they share one particularly important driver, an aging population, which will be the focus of much of the analysis to come. There are a variety of mechanisms at play, but on balance an aging population may lower an economy’s potential rate of growth by decreasing the share of its population employed and may also lower interest rates through changes in saving behaviour over the lifecycle of individuals.8 In Canada, recent evidence suggests that the natural real rate of interest may have been consistently falling over time, as in many other countries.9 Observed rates fluctuate from year to year, to be sure, but a real federal interest rate of 1 percent (3 percent nominal) with provincial borrowing rates roughly 1 percentage point higher is a reasonable rule of thumb that I will use in this article. Current forward rates for government borrowing costs are notably less than this. As for growth rates, if we presume labour productivity growth of 1 percent per year, the analysis to come points to real GDP growth averaging 1.7 percent (3.7 percent nominal). This is consistent with estimates of trend real growth in the literature and among many forecasters. It also implies that future economic growth may very well exceed long-term federal borrowing rates. Does this mean that any level of debt is sustainable? If governments can perpetu- ally roll over debt incurred today, both current and future generations benefit. This is known as “the deficit gamble”—and it may pay off under certain conditions.10 But it comes with risk: If governments fail to roll over debt owing to, say, a large adverse shock, costly fiscal adjustment will be required. It may be prudent to avoid this risk.

7 See, for example, Paolo Mauro and Jing Zhou, r-g<0: Can We Sleep More Soundly? IMF Working Paper no. WP/20/52 (Washington, DC: International Monetary Fund, Fiscal Affairs Department, 2020). 8 For informative research, see Douglas W. Elmendorf and Louise M. Sheiner, “Federal Budget Policy with an Aging Population and Persistently Low Interest Rates” (2017) 31:3 Journal of Economic Perspectives 175 - 94; Kurt G. Lunsford and Kenneth D. West, Some Evidence on Secular Drivers of US Safe Real Rates, Federal Reserve Bank of Cleveland Working Paper no. 17 - 23 (Cleveland, OH: Federal Reserve Bank of Cleveland, December 2017); Gabriele Fiorentini, Alessandro Galesi, Gabriel Pérez-Quirós, and Enrique Sentana, The Rise and Fall of the Natural Interest Rate, CEPR Discussion Paper no. 13042 (London, UK: Centre for Economic Policy Research, July 2018); and Carlos Carvalho, Andrea Ferrero, and Fernanda Nechio, “Demographics and Real Interest Rates: Inspecting the Mechanism” (2016) 88 European Economic Review 208 - 26 (https://doi.org/10.1016/j.euroecorev.2016.04.002). 9 Kathryn Holston, Thomas Laubach, and John C. Williams, “Measuring the Natural Rate of Interest: International Trends and Determinants” (2017) 108 , supplement 1 Journal of International Economics S59 - 75 (https://doi.org/10.1016/j.jinteco.2017.01.004). 10 Laurence Ball, Douglas W. Elmendorf, and N. Gregory Mankiw, “The Deficit Gamble” (1998) 30:4 Journal of Money, Credit and Banking 699 - 720. 1094 n canadian tax journal / revue fiscale canadienne (2020) 68:4

FIGURE 3 The Fiscal Cost of General Government Debt in Canada, 1870-2019

0 hare of GDP () −5

−10

10 100 120 140 160 10 2000 2020 GDP = gross domestic product. Note: This figure displays the fiscal cost of debt as a share of GDP in Canada, based on d0 (r - g)/(1 + g). See the text for details. Source: Author’s calculations from the data in figures 1 and 2.

High debt levels may also lower the probability and duration of periods with favourable interest :growth differentials.11 But regardless, rising debt levels may not be optimal even if there are mechanical fiscal benefits. Taxes necessary to pay interest, for example, may come with additional distortionary effects on the economy, and government bonds may crowd out private investment. These are topics explored by a large (and recently growing) research literature but will not be examined in this article. In any case, with this foundational knowledge and intuition in hand, some additional detail is necessary to quantify the long-run fiscal challenges facing Canada’s governments.

A General Framework for Debt Sustainability Analysis Over long horizons, it is helpful to express future values in present-value terms. A dollar next year, after all, is worth less than a dollar today. Similarly, 1 percent of GDP next year is different than 1 percent today. Given interest rates and growth rates that may change through time, I define the effective discount rate ϕt as

t ____1 + r s ϕ t = ∏ (4) s = 1(1 + g s )

11 For recent evidence on these considerations, see Weicheng Lian, Andrea F. Presbitero, and Ursula Wiriadinata, Public Debt and r − g at Risk, IMF Working Paper no. WP/20/137 (Washington, DC: International Monetary Fund, Research Department, 2020). finances of the nation n 1095

Intuitively, this represents the accumulated interest rates and growth between today and some future year t. Debt of d0 today, for example, will have a future value of ϕt d0 in t years. And the present value of some future primary surplus pt is pt /ϕt today. The effective discount rate is also useful to determine how annual flows -ac cumulate. Specifically, “the sinking fund factor”—the annual amount necessary to accumulate a value equal to 1 percent of GDP by the end of T years—is _ T ϕ _1 -1 -1 ___T σ T = ϕ ∑ ϕ t ≡ , (5) T(t = 1 ) TϕT _ where ϕ T is the (harmonic) mean of the effective discount ratesϕ t over T years. To accumulate an amount equal to 10 percent of GDP, for example, one must raise

10 × σT percent of GDP each year for the next T years. Both ϕt and σT are useful for analyzing public debt dynamics. The former converts between future and present values while the latter converts between stocks and flows. As discussed, public debt evolves according to the government’s budget constraint. Future debt is composed of the accumulated changes derived through repeated substitution of one period’s budget constraint, dt = [(1 + rt)/(1 + g t )] × dt - 1 - pt, into the next. After appropriately rearranging, we have

T -1 dT = ϕT × d0 - ϕT × ∑ ϕ t p t . (6) (t = 1 ) Though the full derivation is omitted, this result is intuitive. The first term following the equals sign is future debt caused by current debt d0. The second term is future debt caused by imbalances between revenue and program spending, summarized in parentheses by the present value of all future primary balances pt. Projecting those future primary balances will occupy the bulk of the analysis to come. With those projections in hand, equation 6 allows us to estimate whether future debt levels dT will exceed current levels d0, and if so, by how much. If those debt levels do differ, a fiscal gap will exist, and changes in revenues or expenditures may be warranted. Consider an immediate and permanent change in revenues or program spending to ensure that dT* = d0, where dT* is the future debt level with fiscal adjustment. That is, I define a fiscal adjustmentf such that

T -1 dT* = ϕT × d0 - ϕT × ∑ ϕ t ( pt + f ) . (7) (t = 1 ) This effectively determines the annual contribution f required to accumulate

(dT - d0) by year T. Using the sinking-fund factor, this is simply

f = (dT - d0) × σT , (8) which will be our measure of a government’s fiscal gap. Notice that the result can be either positive or negative. If a government is projected to run large primary surpluses in the future, and therefore future debt levels will be lower than today’s, then f < 0. This means that there is scope for sustainable tax cuts or spending increases. The opposite is true for a government facing projected primary deficits. 1096 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Some additional intuition may solidify this point. In a special case where interest rates and growth rates are constant over time, the fiscal gap becomes

r - g _ f = d × _ - p , (9) 0 (1 + g) _ where p is the average primary balance from now until time T.12 The intuition here is identical to the simple arithmetic explored earlier. To maintain a stable debt ratio, fiscal policy must adjust to offset primary deficits and any change in the burden of current debt over time. Equation 9 also clarifies the way in which changes in interest rates matter. In general,_ changes in interest rates affect discount rates and therefore potentially affect p . But if revenues and expenditures both grow at the_ same rate as the economy (and therefore pt is constant), they have no effect on p . The effect of interest rates on the fiscal gap therefore depends only on current debt. To illustrate, if debt is 50 percent of GDP and the interest :growth differential rises by 1 percentage point, the fiscal gap increases by 0.5 percent of GDP. This reveals how higher debt today increases risk exposure to future changes in interest :growth differentials. Given this risk, what if we do not want to simply maintain debt but instead to repay it? To achieve dT = 0, we require a larger fiscal adjustment to accumulate d0 by year T. Specifically,

f0 = f + d0σT. (10)

To be clear, neither of the above fiscal gap measures represents optimal policy. These measures are almost certainly not optimal. Any sequence of annual adjustments ( ft ) can achieve the same result as a single uniform_ adjustment so long as, on average, the adjustments are equal and therefore f = f. Governments will need to balance many important tradeoffs when implementing any fiscal policy adjustments. These measures of fiscal gaps are nevertheless useful as digestible metrics by which to quantify the scale of future challenges.

The Effect of a Temporary Fiscal Shock Fiscal adjustment to repay debt is a particularly useful measure for analyzing temporary shocks such as COVID-19. If we (for now) suppose that there are no permanent structural changes in revenues or program spending, we can focus only on changes in debt. Though much remains uncertain, suppose that the pandemic

_ = T w w = ϕ T ϕ 12 In general, p ∑ t = 1_ t pt where the weights are t (1/ t)/( ∑ t = 1 1/ t). The general solution is = ϕ - w - = f d0( T 1) T p _to achieve dT d0. Alternatively, to achieve zero debt by period T, the gap is f0 = d0jT wT - p = f + d0 × wT. Finally, when governments have meaningful levels of financial assets (as is the case_ in Alberta), achieving the same net debt level at time T requires fn = [d0(jT - 1) - Da]wT - p = f - Da × wT , where Da is the change in financial assets as a share of GDP. The two measures are identical if financial assets grow with GDP. finances of the nation n 1097 response increases the public debt by $100 billion (over 4 percent of GDP) provincially and $360 billion (nearly 16 percent of GDP) federally. The combined effect is an increase in general government debt of 20 percent of GDP. While merely illustrative, this matches the International Monetary Fund’s projection for Canada in its June 2020 World Economic Outlook Update.13 What effect does a 20 percentage point increase in the government debt ratio have for long-run finances? If anticipated interest rates and growth rates are unaffected, changes in long-run sustainability depend on changes in debt and the interest :growth differential. Specifically,

r - g Df = Dd × _ . (11) 0 (1 + g) An interest :growth differential of, say, 0.01 implies that a 20 percentage point increase in the public debt would increase the fiscal gap by 0.2 percent of GDP (0.16 percent federally and the rest provincially). This is a relatively minor change, but not trivial. For perspective, it is roughly 0.5 goods and services tax (GST) points in perpetuity. But the result is highly sensitive to the interest :growth differential. If interest rates equal growth, there will be no ongoing fiscal costs from the shock. If interest rates fall below growth (as we have seen is possible for the federal government), there will be fiscal benefits from the higher debt, in the sense that there will be room to lower revenues or increase program expenditures and maintain a stable (though now higher) debt-to-GDP ratio. Simply maintaining the debt ratio in the face of a large current shock, however, may be imprudent. Governments may therefore wish to bring down their debt ratios to pre-crisis levels. How large a fiscal adjustment is required depends on how quickly governments want to bring debt ratios down. If their time frame is a period of T years, the required adjustment is

Df = Dd0 × jT × σT . (12)

More intuitively, if interest rates and growth rates are constant,

-1 r - g -T Dd Df = Dd × _ × 1 - _1 + r ≈ __ 0 . (13) 0 (1 + g) ( (1 + g) ) T

The first term following the equals sign is theCOVID -19 debt shock. The second term is the carrying cost of the incremental debt. The third term reflects how much higher than the carrying cost that payments must be to repay the debt over T years. _ If interest rates and growth rates are equal, however, ϕ T = 1 and therefore Df = Dd0 /T. Retiring COVID-19 -related debt equivalent to 20 percent of GDP would therefore require increasing revenues or decreasing spending by 2 percent of GDP for 10 years, or by 1 percent for 20 years. This is a convenient rule of thumb.

13 Supra note 5. 1098 n canadian tax journal / revue fiscale canadienne (2020) 68:4

With this robust framework for modelling public debt dynamics in hand, we may proceed to a detailed examination of federal and provincial finances in Canada.

PROJECTING GOVERNMENT FINANCES OVER THE LONG RUN The Projection Model To project future primary balances, I disaggregate revenues and expenditures into separate components and project forward their underlying bases or cost drivers. This exercise will be grounded in an initial year that maps directly to data on government finances from Statistics Canada. I summarize the relevant provincial budget components and data sources in figure 4, and in what follows I (briefly) describe the assumptions that I use to project their future values. On the revenue side, I model 12 separate sources. Most revenues grow with the overall economy. Personal income taxes, corporate income taxes, payroll taxes, and consumption and excise taxes are mechanically related to total income and total spending. To the extent that economic growth is shared proportionally across the income distribution, aggregate rates of economic growth are sufficient proxies for rates of growth in these revenue sources. Certain other revenue sources also keep pace with overall economic growth, such as business income and (potentially) natural resource revenues. But many revenue sources grow more slowly. Some provincial governments, such as British Columbia’s, set property tax growth to maintain the real cost per household. I assume that revenue from this source will therefore roughly keep pace with changes in the population and in inflation. Tobacco and gasoline taxes are also likely to grow more slowly—the former as the share of the smoking population declines, and the latter as fuel use declines as a result of technological change. I assume that tobacco taxes grow with inflation and gasoline taxes grow with real GDP. Regarding own-source revenues in the “other revenue” category, I assume that they will grow with rising inflation and population growth. Finally, federal transfers follow an explicit formula. I separately model four distinct components of federal transfers. First, the CHT (Canada health transfer) (the largest major transfer) grows with a moving average of national nominal GDP growth, with a minimum floor growth rate of 3 percent per year. Second, the Canada social transfer (CST) is simpler and grows at a fixed rate of 3 percent per year. Both transfer programs are distributed across provinces according to population. Third, equalization payments are not equal and instead are distributed according to provincial revenue-raising capabilities. The total size of equalization grows with a moving average of national nominal GDP, with no floor growth rate. I model provincial fiscal capacity as evolving from the observed average across fiscal years starting in 2016 to 2018 and over time according to a three-year moving average of provincial nominal GDP. This is a very good approximation of each province’s true fiscal capacity. Fourth, and finally, I assume that provincial revenues from all other transfer programs will increase with population and inflation. finances of the nation n 1099

FIGURE 4 Provincial Government Financial Flows, 2018 (Millions of Dollars)

Taxes Revenues Total inflows and outflows Expenditures Program spending

Personal income: 108,206

Health: 157,425

Total taxes: 253,162

Corporate income: 37,169

Payroll taxes: 15,607 Program spending: 414,711 Education: 59,468 Total revenues: 438,739 Sales taxes: 63,996 Total expenditures: 491,175 Resource royalties: 12,349 Advanced education: 32,026 Dividend income: 5,371 Property taxes: 13,076 Interest income: 13,800 Gas tax: 10,243 Social services: 70,535 Tobacco taxes: 4,865

Federal transfers: 82,711

Capital: 40,112 Other: 95,257

Other revenue: 71,346 Borrowing: 52,436 Interest payments: 36,352

Note: This figure displays the aggregate 2018 fiscal inflows to provincial governments on the left and the fiscal outflows on the right. Sources: Author’s calculations from Statistics Canada tables 10-10-0017-01 (formerly CANSIM 385-0034), “Canadian Government Finance Statistics for the Provincial and Territorial Governments (x 1,000,000)”; and 10-10-0024-01 (formerly CANSIM 385-0040), “Canadian Classification of Functions of Government, by General Government Component (x 1,000,000).” Additional federal budget data, though not displayed, are from Statistics Canada table 10-10-0016-01 (formerly CANSIM 385-0033), “Canadian Government Finance Statistics for the Federal Government (x 1,000,000).”

On the expenditure side, the three largest provincial ministries account for over 60 percent of total program spending. I model each separately. Health care, given its size and its importance for long-run provincial finances, will be discussed in depth shortly. Primary and secondary education spending will grow along with the K- 12 population and inflation, plus a real increase of 0.5 percent per year in the per- student spending. This increment roughly accounts for real wage increases among workers in education that keep pace with the rest of the economy. Similarly, post- secondary education spending will grow along with the relevant population, which I consider to be those aged 20 to 24, plus inflation and a 0.5 percent real per-student increase. I assume that all other program spending grows in line with population plus inflation (a very conservative assumption), and capital spending grows with the provincial economy.14 Provincial primary balances are then calculated as total revenue from all sources minus total program and capital spending.

14 Provincial governments do not normally include infrastructure spending in their calculation of budget deficits since such expenditures are gradually amortized over time. Because the focus of this article is on public debt dynamics, a cash basis for the deficit is more appropriate, so all capital spending is included. 1100 n canadian tax journal / revue fiscale canadienne (2020) 68:4

The federal government is also an important component of the analysis to come. Its revenue sources are simpler and grow faster than those of provincial governments, since federal revenues from income and consumption taxes constitute a significantly larger share of the total. Tax revenues that grow with the economy account for nearly 90 percent of federal revenues. Federal revenues from other sources, such as government business enterprises, are also likely to grow with overall GDP. Mean- while, EI premiums are tied to EI benefit payments. On the spending side, the federal government makes significant transfers to individuals through the payment of benefits to seniors, families with children, and the unemployed. I assume that these transfers will grow with the relevant demographic group plus inflation, andEI benefits will also grow with GDP per worker. I assume that defence spending will grow with the overall economy, consistent with stated goals and North Atlantic Treaty Organization (NATO) guidelines. Transfers to provinces have been discussed above. Other program spending is assumed to grow with population plus inflation. Finally, there are several important macroeconomic variables that drive these budget projections. As discussed earlier, I assume federal borrowing rates of 3 percent and provincial borrowing rates of 4 percent. Actual rates may come in lower or higher, and I explore how sensitive the main results are to alternative assumptions. Total interest costs are endogenous and are determined by the model: public debt of dt implies interest costs of r × dt. To err on the conservative side, I do not assume that interest rates themselves will respond to overall debt levels; rather, I assume that Canadian governments will have access to a large global capital market that does not charge a risk premium if debt levels grow large. Finally, for each province’s overall economy, real GDP is expressed as

Yit = Aitwit Pit, (14)

where Ait is labour productivity, wit is the working-age share of the population, and Pit is the total population. I assume that labour productivity growth is 1 percent per year; the working-age share is from the Statistics Canada population projections cited above. Total GDP across all provinces is then equal to Canada’s total GDP.15

Projecting Future Health-Care Expenditures Health care is the most significant public service delivered by provincial governments. It accounts for nearly 40 percent of overall program spending, and as populations age, this will only grow. Statistics Canada’s latest projection suggests that the share of Canada’s population aged 65 and over may rise from 18 percent today to nearly 24 percent by 2040, while the share aged 75 and over may double

15 This method of calculation excludes economic activity in the three northern territories, which is not quantitatively important for the purposes of the analysis. finances of the nation n 1101 from 7 percent today to 14 percent.16 Changes in population shares map into health- care expenditures using data on average spending by age and gender cohorts. Specifically, average per-capita spending for a province, i, in year t is

c c hit = ∑ hit p it , (15) c

c where h it is the per-capita spending for cohort c (say, men aged 20 to 24 or women c aged 65 to 69) and p it is the share of the province’s population accounted for by this cohort. Using data compiled by the CIHI, I illustrate the full distribution of age- specific health-care spending in figure 5, with the range across all provinces illustrated as “whiskers” around the overall Canadian average. Demographics will affect average spending levels as population shares change. In this analysis, I use Statistics Canada’s population projections for 2018 to 2068, taking its medium-growth (M2) scenario as the baseline case, but report how sensitive the results are to alternative growth assumptions.17 Beyond 2068, I assume that population shares are constant. In any case, holding all other factors constant, health-care costs increase according to a weighted average of cohort-specific population change, expressed as

c c hˆ it = ∑ w i 0 pˆ it , (16) c

c c c where w i0 ∝ h i 0 p i0 is the initial share of total health-care spending accounted for by spending on individuals within cohort c and hats denote relative changes. I find that h ˆ it is nearly 1.28 for British Columbia by 2050, implying that demographics and aging alone will increase health-care spending in that province by 28 percent. Across provinces, this projection ranges from a high of 1.53 in Newfoundland and Labrador to a low of 1.14 in Saskatchewan. Beyond demographics, other factors contribute to health-care spending. To estimate health-care-specific cost inflation over and above the economy-wide c 2 percent per year, I use the same weight w i0 to construct

c c h ˆ it = ∑ w i 0 h ˆ it . (17) c I estimate that since 1998 health-care-specific cost inflation grew at roughly 1.3 percent per year, though there is significant variation over time and across provinces. From 1998 to 2010, for example, the average was 2.3 percent across Canada, falling to roughly zero in the years following. Over the whole period since 1998, this measure was lowest in Quebec, at 0.85 percent per year, and highest in Alberta, at 2.3 percent. Looking ahead, I assume 1 percent health-care-specific inflation

16 Statistics Canada table 17 - 10 - 0057 - 01 (formerly CANSIM 052 - 0005), “Projected Population, by Projection Scenario, Age and Sex, as of July 1 (x 1,000).” 17 Ibid. 1102 n canadian tax journal / revue fiscale canadienne (2020) 68:4

FIGURE 5 Per-Capita Health-Care Spending in Canada by Age Group, 2017 50,000

ange across provinces 40,000

30,000

20,000 Dollars per capita

10,000

0 < 1 1-4 5- 0+ 10-14 15-1 20-24 25-2 30-34 35-3 40-44 45-4 50-54 55-5 60-64 65-6 70-74 75-7 0-4 5- Age group

Note: This figure displays the average level of health-care spending in Canada per capita across different age groups. The range across provinces represents the gap between the provinces with the lowest average spending and those with the highest average spending. Source: Author’s calculations from the Canadian Institute for Health Information, National Health Expenditure Trends, 1975 to 2019 (Ottawa: CIHI, 2019). as the baseline case, but report results across a range of values. This measure assumes the same rate of health-care-specific inflation across all age and gender cohorts.18 Combining both demographic change and health-care-specific inflation, I project forward overall per-capita health-care costs for each province. While I do not display all individual provincial projections here, I report the average per-capita spending levels in figure 6. From an overall level of roughly $4,500 per capita in 2018, the projection using demographics alone rises to over $5,000 (in 2018 dollars per capita) by 2040 and nearly $5,300 by 2050. Including health-care-specific inflation, these estimates rise to over $6,200 by 2040 and $7,200 by 2050. Over the next 30 years, this increase represents an average annual growth rate of 1.8 percent per year for health-care costs attributable to demographics alone and incremental health-care- specific inflation of 1 percent per year. This analysis suggests that provincial health-care spending in Canada will rise from just over 7 percent of GDP today to nearly 9 percent by 2040 and to 10 percent by 2050. Health-care spending plateaus at roughly this level for the remaining years. This 3 percentage point increase in provincial health-care spending is reasonable, though it is somewhat larger than the projection in the latest PBO fiscal sustainability report.19 To be clear, there are several sources of uncertainty. Technological developments in health-care delivery may

18 The magnitude of future price changes is unlikely to be uniform, and historically the relative cost increases for older age cohorts have been smaller than for younger cohorts; however, 1 percent is a reasonable approximation. 19 See supra note 3. finances of the nation n 1103

FIGURE 6 Per-Capita Health-Care Spending by Provincial Governments, Actual and Projected, 1975-2055 Onward

10,125.00

6,750.00

4,500.00

3,000.00 Projection

Dollars per capita, log-scale 2,000.00

175 15 2015 2035 2055 Demographics + 1% health-care-speciﬁc inﬂation Demographics only

Note: This figure displays the average level of health-care spending by provincial governments per capita historically between 1975 and 2019 and the author’s baseline projection from 2018 onward. The projection separately reports health-care costs with and without a 1 percentage point increase in the rate of inflation specific to those costs. Source: Author’s calculations from the Canadian Institute for Health Information, National Health Expenditure Trends, 1975 to 2019 (Ottawa: CIHI, 2019) for the period 1975-2019. The projection incorporates several data sources and methods. See the text for details. increase or decrease costs. Immigration patterns may dampen the pace at which Canada’s population ages. And health-care spending is endogenous to other government policies, such as supports for low-income individuals, housing, promotion of health and wellness, and so on. Nevertheless, I suggest that the projections presented here are a conservative illustration of a potential future.

THE LONG-RUN FISCAL SUSTAINABILITY OF CANADA’S PROVINCES Combining all revenue and expenditure projections described in the previous section, we may proceed to estimating the long-run fiscal future of Canada’s various governments. Overall, the federal government is in a much stronger position than provincial governments. Federal revenue growth averages roughly 3.7 percent per year consistently across the 75-year forecast horizon. The ratio of federal revenue to GDP is therefore stable, since this is also the growth rate of the national economy. Provinces, however, will see more modest revenue growth ranging between 3.3 percent and 3.4 percent; therefore, the revenue-to-GDP ratio declines from the current level of 20 percent to 17.4 percent by 2040 and 16 percent by 2060. In terms of program expenditures, the federal government may see growth averaging 3.4 percent per year to 2040, declining somewhat thereafter. Provincial governments may see much more rapid growth, with program expenditures rising by an 1104 n canadian tax journal / revue fiscale canadienne (2020) 68:4 average of 3.8 percent to 2040 and roughly 3.5 percent thereafter. Health-care spending is a core driver, with growth at nearly 5 percent per year over the next two decades and a more modest 4 percent thereafter. With provincial revenues failing to keep pace with expenditures, deficits will rise and debt will mount. The federal government will see the opposite. Using the debt dynamics expressions derived earlier, I summarize the average annual primary deficits and the accumulated debt_ that those deficits create in table 1. Specifically, these values correspond to -p T -1 and -∑ t = 1 ϕ t pt , respectively. Most provincial governments face large and persistent gaps between their projected revenues and program expenditures. Quebec is the notable exception. Between 2018 and 2050, for example, Quebec’s average annual primary surplus is 1.3 percent of GDP. By comparison, all other provincial governments have an average annual primary deficit of no less than 1.1 percent, and Newfoundland and Labrador has an average deficit of 7.7 percent. In present-value terms, the projected provincial primary imbalances from 2018 to 2050 are collectively equivalent to two- thirds of GDP today. For the period 2018 - 2090, I estimate that these imbalances are equivalent to 168 percent of GDP today. For comparison, the total stock of current provincial gross debt is 42 percent of GDP. Projected future imbalances are therefore significantly larger than the current debt-to-GDP ratio and dwarf the short-term debt increases attributable to COVID-19. In contrast, the federal government is expected to record long-run primary surpluses averaging 1 percent of GDP between 2018 and 2050. In subsequent years, the projected federal surplus is even larger. Between 2018 and 2090, the average federal primary surplus is 2.5 percent of GDP—the equivalent of roughly 8 GST points today. The present value of such surpluses approaches 230 percent of GDP. Projected provincial debt levels are not merely large but also unsustainable. For some provinces, they would not likely be possible, and a fiscal crisis would occur before the end of the forecast horizon. By the end of the 75-year period, for example, I project a debt-to-GDP ratio of nearly 350 percent of GDP for Alberta. At 4 percent interest, this would require that 14 percent of Alberta’s entire economy be directed toward debt service payments. Given the revenue instruments available to the government in this projection, interest costs would amount to 125 percent of revenue. Something would have to give long before this situation materialized. One measure of long-run sustainability is the fiscal adjustment required, starting immediately, to stabilize debt as a percentage of GDP. This adjustment is equivalent to the average primary imbalances reported in table 1 plus a measure of the burden of current debt. The projected fiscal gap is 2.7 percent of GDP for provincial governments collectively and −2.8 percent for the federal government. Importantly, the combined federal and provincial finances are sustainable in the long run, but there exists a persistent imbalance between the two orders of government as well as large differences between provinces. I report gaps for each province in table 2. finances of the nation n 1105 2090 306.0 167.0 120.0 114.0 248.0 131.0 418.0 240.0 107.0 188.0 − 49.0 − 229.0 98.0 88.0 89.0 74.0 2070 253.0 120.0 180.0 289.0 173.0 139.0 − 40.0 − 104.0 66.0 64.0 51.0 97.0 44.0 35.0 82.0 2050 173.0 163.0 101.0 − 34.0 − 35.0 for the period 2018 to . . . Accumulated primary deficits 9.0 7.0 3.0 16.0 18.0 20.0 60.0 52.0 32.0 24.0 2030 − 5.0 − 25.0 over various time horizons, and the accumulated

_ p 0034), “Canadian Government Finance Statistics for the - - 2.6 2.6 2.4 3.9 8.4 2.1 4.9 3.4 1.6 3.0 2090 − 2.5 − 0.9 2.5 2.7 2.3 3.8 8.3 1.9 4.9 3.3 1.5 2.9 2070 − 1.7 − 0.9 2090 (Percent of GDP) - 01 (formerly CANSIM 385 2.2 2.5 2.0 3.2 7.7 1.5 4.6 3.1 1.1 2.7 - 2050 − 1.0 − 1.3 for the period 2018 to . . . 0017 - Average annual primary deficit Average 10 - 1.4 1.7 0.8 1.7 5.8 0.6 4.2 2.7 0.2 2.1 2030 − 0.4 − 2.2 42.0 35.0 70.0 43.0 38.0 46.0 22.0 25.0 29.0 74.0 47.0 52.0 debt . t p Current - 1 t ϕ

= 1 t T ∑ - . . 2018 Governments, for Canada’s Fiscal Projections Long-Run . . . Provincial and Territorial Governments (x 1,000,000).” Other values are the author’s calculations. See the text for details. Governments (x 1,000,000).” Other values are the author’s Provincial and Territorial Sources: Current debt is from Statistics Canada table 10 TABLE 1 TABLE present value of those deficits, GDP = gross domestic product. Note: This table reports the current gross debt ratio for 2018, average primary budget deficits Provinces . Federal New Brunswick . Nova Scotia . Prince Edward Island . Newfoundland and Labrador . Province British Columbia Alberta Saskatchewan Manitoba . Ontario Quebec . 1106 n canadian tax journal / revue fiscale canadienne (2020) 68:4

TABLE 2 Long-Run Fiscal Gaps for Canada’s Provinces (Percent of GDP)

Fiscal adjustment to meet different debt targets over a 75 -year horizon Gross debt Net debt Same net Same gross Province in 2018 in 2018 Zero debt debt debt

British Columbia ...... 22.5 14.4 2.5 2.3 2.2 Alberta ...... 25.5 8.0 5.2 5.1 4.8 Saskatchewan ...... 29.4 14.7 3.8 3.6 3.4 Manitoba ...... 74.4 34.4 2.6 2.2 1.7 Ontario ...... 47.8 39.5 3.7 3.2 3.1 Quebec ...... 51.7 39.3 0.0 −0.4 −0.6 New Brunswick ...... 69.7 37.8 4.1 3.7 3.5 Nova Scotia ...... 43.4 33.8 3.3 3.0 3.0 Prince Edward Island . . . . . 38.5 30.4 4.5 4.1 4.0 Newfoundland and Labrador . . 46.1 46.1 9.7 9.4 9.4 Provinces ...... 41.7 30.0 3.2 2.9 2.7 Federal ...... 35.1 32.6 −2.2 −2.8 −2.8

GDP = gross domestic product. Note: This table displays the initial debt and net debt as a share of GDP in 2018 and three measures of the 75 -year horizon fiscal gap. Zero debt reports the permanent fiscal adjustment necessary to achieve zero debt at the end of the 75-year forecast horizon. Same gross and net debt report the adjustment necessary to achieve either the same gross or the same net debt as in 2018. The bottom two rows report the estimates for the aggregate of all 10 provinces and the federal government, respectively. Sources: Current debt is from Statistics Canada table 10 -10 - 0017- 01 (formerly CANSIM 385 - 0034), “Canadian Government Finance Statistics for the Provincial and Territorial Governments (x 1,000,000)”; net debt is from Canada, Department of Finance, Fiscal Reference Tables 2019 (Ottawa: Department of Finance, September 2019), tables 18 through 27. Other values are the author’s calculations. See the text for details.

What Affects Provincial Fiscal Gaps? It is instructive to investigate the drivers of provincial primary balances in the long run. First, consider macroeconomic developments. Interest rates and economic growth rates both matter, but the latter more than the former. For a range of federal borrowing rates between 2 percent and 5 percent, and provincial rates between 3 percent and 6 percent, fiscal gaps range from− 3.4 percent to −1.5 percent federally and 2.6 percent to 3.4 percent provincially. For a range of labour productivity growth rates from 0.5 percent to 1.5 percent per year, fiscal gaps range from −4.5 percent to −1.5 percent federally and 0.3 percent to 5.7 percent provincially. To reinforce this point, I illustrate in figure 7 the projected debt-to-GDP ratios for provincial governments given an annual rate of growth in labour productivity that is 0.5 percentage points higher than the baseline rate of 1 percent. Sustained increases in productivity growth, though difficult for governments to influence directly, are crucial for long-run sustainability. Alternative demographic assumptions also matter, but only slightly. Using Statistics Canada’s slow-aging scenario, I finances of the nation n 1107

FIGURE 7 Projected Provincial Debt-to-GDP Ratios with Higher Productivity Growth, 2020-2090 300

200 Percent 100

0 2020 2030 2040 2050 2060 2070 200 200 +0.5 labour productivity growth Baseline projection

GDP = gross domestic product. Note: This figure displays the projected level of aggregate provincial debt as a share of GDP with 0.5 percentage points higher annual labour productivity growth than in the baseline projection of 1 percent per year. Source: Author’s calculations. See the text for details. estimate an average provincial fiscal gap of 2.0 percent and a federal gapof −3.2 percent. Using the fast-aging scenario, these estimates become 3.4 percent and −2.5 percent, respectively. Our framework also allows_ for a simple additive decomposition of the average annual provincial balance p and therefore a decomposition of the long-run provincial fiscal gap f. While we can investigate the contribution from each of the individual revenue and expenditure components contained in the analysis (see figure 4), I combine some of those components for ease of presentation in figure 8. Each represents the additive contribution to primary balances and therefore the fiscal gap. Overall, provincial revenues equivalent to nearly 18 percent of GDP are more than offset by expenditures of 21 percent of GDP, nearly half of which is for health care. This visual puts magnitudes in proper perspective. Own-source revenues average 14.3 percent across the 75-year horizon, so the fiscal gap is equivalent to nearly one-fifth of those revenues. This would be the increase in revenues necessary to eliminate the fiscal gap. And on the spending side, the long-run gap is equivalent to roughly one-seventh of current program expenditures or one-quarter of non- health-care spending. Not only do fiscal gaps differ widely across provinces, but so too do the underlying drivers. I report the magnitude of each component in table 3. Comparing each province with the 10 -province average reveals some of the important underlying causes of provincial fiscal gaps. Alberta’s 4.8 percent gap, the second-largest of all provinces, is not due to above-average levels of expenditures over the projection period. Instead, total tax revenues are roughly 5 percentage points of GDP below the 1108 n canadian tax journal / revue fiscale canadienne (2020) 68:4

FIGURE 8 Decomposing the Long-Run Provincial Fiscal Gap 3.0

0.0

−3.0

−6.0

−.0

hare of GDP −12.0

−15.0

−1.0 Current Ta x Other ederal Health Other Capital Long-run provincial revenue own-source transfers spending program spending ﬁscal gap debt revenue spending

GDP = gross domestic product. Note: This figure displays the relative contributions of various budget components to the aggregate provincial fiscal gap across a 75-year horizon. The size of each bar corresponds to the average annual amount represented by each component as a share of GDP. Negative values shrink the fiscal gap while positive values enlarge it. See the text for details. Source: Author’s calculations. See the text for details. national average, fully accounting for the province’s long-run fiscal gap. Other own- source revenues are larger than average and include resource revenues. These estimates suggest that if Alberta had average taxes, its finances would be sustainable in the long run. Newfoundland and Labrador, the province with the largest long- run challenge by a wide margin, is different. This province is projected to have revenues that are more than 4 percentage points of GDP higher than the average. But its expenditure levels, especially for health care, more than offset that advantage. Interestingly, other Atlantic provinces also have significantly higher health-care expenditures yet a far lower fiscal gap. On the revenue side, the maritime provinces benefit significantly more from federal transfers compared to Newfoundland and Labrador. We will soon see that this difference is entirely the result of Canada’s equalization program. Finally, Quebec is the only province with solidly sustainable finances over the projection period. Its position is not due to lower expenditure demands (which, on the contrary, are above average in all categories) but to significantly higher taxes than elsewhere—nearly a full 5 percentage points of GDP higher on average across the 75-year horizon. So far, this analysis has been a mere accounting exercise. We can push further in understanding the long-term challenges faced by provinces by experimenting with alternative scenarios where provincial finances are resimulated under different -as sumptions. This can be a powerful means of identifying fundamental causes of fiscal challenges. For instance, by holding fixed the demographic composition of provinces at the observed 2018 levels, we can show that an aging population fully finances of the nation n 1109 a 2.2 4.8 3.4 1.7 3.1 3.5 3.0 4.0 9.4 2.7 − 0.6 Fiscal gap 1.8 1.6 2.0 1.7 1.7 2.1 3.3 2.2 2.1 2.1 1.8 Capital Expenditures 8.4 8.1 9.3 8.2 9.2 10.0 12.4 12.0 13.6 12.8 13.9 Non-health 9.7 7.8 7.9 8.9 9.4 . Negative numbers shrink the fiscal gap while 10.8 11.4 14.2 15.4 15.4 12.7 Health − 2.9 − 2.0 − 2.7 − 6.0 − 2.7 − 5.6 − 4.9 − 3.4 − 10.0 − 10.9 − 11.0 Federal transfers Other − 3.9 − 4.9 − 5.4 − 3.1 − 1.9 − 5.9 − 3.6 − 3.1 − 5.9 − 3.7 − 4.2 Revenues own-source − 5.7 − 8.4 Taxes − 10.9 − 11.0 − 11.2 − 15.4 − 12.4 − 13.2 − 11.1 − 10.6 − 12.1 0.1 0.0 0.1 0.2 0.3 0.5 0.1 1.0 0.1 0.9 − 0.1 Effect of current debt . . (Percent for Individual Provinces Fiscal Gap Decomposing the Long-Run of GDP) . . Fiscal gap measure based on gross debt ratios. calculations. See the text for details. Source: Author’s . Province TABLE 3 TABLE British Columbia Alberta Saskatchewan Manitoba . Ontario Quebec . Nova Scotia . Prince Edward Island . Newfoundland and Labrador . All GDP = gross domestic product. Note: This table displays the relative contributions of various budget components to fiscal gap over a 75-year horizon for each province. Each number corresponds to the average annual magnitude of each revenue and expenditure category as a share GDP positive numbers enlarge it. The final column may not exactly equal the sum of other seven because rounding. See text for details. a New Brunswick . 1110 n canadian tax journal / revue fiscale canadienne (2020) 68:4 accounts for the long-run fiscal gaps facing provincial governments. Population growth is unaffected overall, but the fraction of the population in each age category is fixed through time. This has two effects. First, because the working-age share of the population is no longer declining, economic growth rates are higher, averaging 4 percent per year until 2040 and 3.8 percent thereafter. Second, health-care expenditures grow more slowly, roughly maintaining a level just over 7 percent of GDP instead of gradually increasing to 10 percent in the baseline estimates. Both of these factors matter for the long-run financial health of provinces. Decomposing the effect of demographics on the long-run financial position of provinces, I find that over 40 percent of the fiscal gap is from slower economic growth, nearly 60 percent is from rising health-care costs, and lower education costs provide a modest offsetting effect.20 In aggregate, without any demographic change, the measured fiscal gap declines to only 0.2 percent ofGDP . I display these changes in figure 9. While an aging population is the central driver of provincial long-run fiscal challenges overall, this is not true for all provinces. In figure 10,Ireport the fiscal gap estimates for each province with and without demographic changes. It is evident that the oil-producing regions of Alberta, Saskatchewan, and New- foundland and Labrador face unique challenges. In part, these reflect the challenges that are particular to these provinces. All three of them have yet to address their overreliance on natural resource revenues. If this revenue source does not grow significantly faster than the province’s overall economy, large structural deficits will persist. For Newfoundland and Labrador, slow underlying rates of economic growth reflect demographics—as seen in the large change between the baseline results and the scenario holding demographics fixed—but that province also has a structural challenge. In Newfoundland and Labrador’s case, there may be scope for federal support since the scale of the challenge may exceed the provincial government’s own capacity to achieve a healthier fiscal balance. The potential role of the federal government in improving the sustainability of provincial finances will be a theme in much of the analysis to come. Policy Options To Improve Fiscal Outlooks An aggregate fiscal gap of 2.9 percent, which implies immediately and permanently raising revenues by the equivalent of nearly 8 GST points or decreasing spending by nearly 15 percent, may appear daunting. The gap is certainly not small. But it can be reduced through less abrupt action—that is, through gradual and sustained changes to revenue and spending policies. In the discussion that follows, I will present examples of a few such options. These examples will also serve to illustrate the sensitivity of fiscal gap estimates to the underlying projection assumptions.

20 The effects are not strictly additive. The change in provincial fiscal gaps owing to demographic-related education or health-care costs, for example, is different when economic

growth is also affected. This follows from the underlying effective discount factor jt being different across scenarios. I report here the average marginal contribution of each factor across all 12 possible orderings of the three factors. finances of the nation n 1111

FIGURE 9 Decomposing the Effect of Demographics on Provincial Fiscal Gaps 3.0

2.5

2.0

1.5

hare of GDP 1.0

0.5

0 Baseline No effect on No effect on No effect on Provincial provincial economic health-care education fiscal gap fiscal gap growth costs costs excluding demographics

GDP = gross domestic product. Note: This figure displays the relative contributions of three ways in which demographic change affects long-run provincial finances. This illustrates the fiscal adjustment required over a 75-year horizon. Source: Author’s calculations. See the text for details.

FIGURE 10 The Effect of an Aging Population on Provincial Fiscal Gaps 10

hare of GDP per year () −2

−4

New Nova British Alberta Ontario Quebec Manitoba cotia Island Columbia Brunswick askatchewan Prince dwardNewfoundland and Labrador Baseline scenario Holding demographics ﬁxed

GDP = gross domestic product. Note: This figure displays the estimate of the long-run fiscal gap (corresponding to a 75-year horizon) for each province, with and without an aging population. Source: Author’s calculations. See the text for details. 1112 n canadian tax journal / revue fiscale canadienne (2020) 68:4

If provinces could lower the health-care-specific inflation rate from 1 percent to, say, 0.5 percent per year, the aggregate provincial fiscal gap would decline to 1.3 percent—less than half of the baseline estimate presented above. This approach would mark a material departure from the past, but it may not be infeasible. It would imply that health-care spending would rise to a peak of 8.4 percent of GDP by the mid- 2040s and fall thereafter. On the revenue side, as we have seen, there are many revenue sources that will not keep pace with overall economic growth. Over the entire projection period, total revenue grows at a rate that is roughly 0.3 percentage points lower than GDP growth. If gradual reforms over time, such as small changes in tax rates, fee schedules, and so on, kept provincial own-source revenue growth in line with GDP, the aggregate fiscal gap would decline to 1.5 percent. If provincial governments could achieve both modestly lower health-care spending growth and modestly higher own-source revenue growth, the entire fiscal gap might be closed. Cutting health-care-specific inflation to 0.5 percent and growing revenues in line with GDP would lead the aggregate fiscal gap to decline to −0.1 percent, achieving sustainability within the 75-year horizon. While the overall picture in this scenario is hopeful, Alberta and Newfound- land and Labrador remain in an unsustainable position—and, to a lesser extent, so too does Ontario. To be clear, Alberta and Ontario have more options at their disposal to close this remaining gap. But Newfoundland and Labrador does not appear to have many easy options available. To illustrate one potential scenario, consider (1) lowering health-care-specific inflation to 0.5 percent, (2) growing provincial own-source revenues with GDP, (3) allocating CHT payments on the basis of the population aged 65 and over, and (4) removing resource revenues from the equalization program. Under this scenario, aggregate provincial finances are fully sustainable and so too are Newfoundland and Labrador’s. The first two components of this package are for the province to implement while the latter two require federal reforms. While this scenario achieves sustainability within the 75- year horizon, unfortunately the transition path may not be feasible. Debt will accumulate substantially in the meantime and will exceed 100 percent by the mid- 2030s. There may be no avoiding a more aggressive approach to fiscal consolidation in Newfoundland and Labrador. The province’s average program expenditures as a share of GDP are fully 5 percentage points higher than the national average. It will be necessary to bring spending in line with that of other provinces, and perhaps to modestly increase Newfoundland and Labrador’s own-source revenues (which are already above average), in combination with other gradual and ongoing fiscal reforms. Federal transfers will also be important. I turn to this area of the fiscal landscape next and explore it in depth.

Federal Transfers and Provincial Financial Sustainability Federal transfers play an essential role in ensuring that provincial governments have the fiscal capacity necessary to deliver key public services. Programs like the CHT finances of the nation n 1113 and the CST are allocated across provinces on an equal per-capita basis, while fiscal equalization payments top up provinces with below-average ability to raise their own revenues. Provinces with weak economies tend to have smaller tax bases, and therefore less capacity to raise revenue. Over time, the economic prospects of some provinces are also stronger than the prospects of others. For example, the populations of the Atlantic provinces are aging more quickly, and this trend may dampen the economic growth rates in those provinces. In the baseline scenario explored in this article, I estimate average real GDP growth rates of 0.8 percent per year in New Brunswick and Nova Scotia, and near zero in Newfoundland and Labrador. Mean- while, Ontario averages growth of 1.6 percent per year, and Alberta averages 2.4 percent. These growth differentials will, over time, affect the relative revenue- raising capabilities of provincial governments. Equalization will therefore help to fill that gap. To estimate the effect of equalization on provincial debt sustainability, I estimate fiscal gaps under a scenario where equalization is replaced with an equal per-capita transfer. That is, equalization is eliminated and the proceeds are used to proportionally increase the CHT and CST. This is not a proposal under serious consideration by any federal political party, but proposals along these lines are regularly advanced. Saskatchewan Premier Scott Moe, for example, recently pitched a 50/50 plan whereby equalization would be cut in half and the proceeds redirected toward equal per-capita allocations.21 In any case, there are large implications of this change for lower-income regions. I estimate that the fiscal gap in New Brunswick would increase to 9.5 percent of GDP from its baseline level of 3.7 percent. Nova Scotia and Prince Edward Island would also see significant increases. Quebec and Manitoba— the other large equalization recipients—would each see their fiscal gap increase by more than 2 percentage points. Higher-income regions that typically do not receive equalization would benefit since the per-capita grants would increase. Alberta’s fiscal gap would decline from 5.1 percent to 4.4 percent; Saskatchewan’s, from 3.6 percent to 2.9 percent; British Columbia’s, from 2.3 percent to 1.4 percent; and Ontario’s, from 3.2 percent to 1.8 percent. I report these results, along with other scenarios for federal transfers, in table 4. More generally, the contribution of current fiscal arrangements to long-run provincial sustainability may also be quantified. I estimate fiscal gaps assuming that all federal transfers were set at zero in the third column of table 4. Without transfers, the aggregate provincial fiscal gap would be 6.3 percent, more than double the baseline estimate. This suggests that as of 2018, federal transfers cover nearly 60 percent of provincial fiscal gaps that would exist in the absence of transfers. To be sure, provincial governments would have made very different tax and expenditure decisions in such a situation. But over the 75-year period examined here, federal transfers contribute the equivalent of 3.4 percent of GDP to provincial revenues.

21 Government of Saskatchewan, “Premier Scott Moe Calls for Changes to Equalization Program,” press release, June 20, 2018 (https://www.saskatchewan.ca/government/news-and -media/2018/june/20/equalization-program). 1114 n canadian tax journal / revue fiscale canadienne (2020) 68:4 0.4 3.1 1.7 0.2 1.3 1.8 1.1 2.2 7.5 0.9 − 2.4 − 0.9 to provinces Transfer GST Transfer 1.0 1.6 4.6 2.9 6.6 2.5 0.5 1.4 8.2 2.0 − 1.8 − 2.0 25 percent Boost cash transfers by Increase transfers 2.6 2.0 4.9 3.3 1.6 2.9 2.0 3.0 8.9 2.5 − 1.0 − 2.5 -year time horizon. 10 percent Boost cash transfers by 5.2 7.1 6.3 8.2 5.9 5.1 6.3 14.7 13.1 15.1 14.3 − 6.2 Eliminate all federal transfers Effect of transfers 9.5 1.4 4.4 2.9 2.3 1.8 1.8 7.5 9.5 9.6 2.9 − 2.8 Replace and CST equalization with larger CHT 3.7 2.3 5.1 3.6 2.2 3.2 3.0 4.1 9.4 2.9 − 0.4 − 2.8 Baseline fiscal gap (Percent Governments Sustainability of Provincial on the Long-Run of GDP) Transfers . . The Effect of Federal . . . Source: Author’s calculations. See the text for details. Source: Author’s New Brunswick . British Columbia Alberta Saskatchewan Manitoba . Ontario Quebec . Nova Scotia . Prince Edward Island . Newfoundland and Labrador . Provinces . Federal Province CHT = Canada health transfer; CST social GDP gross domestic product; GST goods and services tax. Note: This table displays the fiscal gap estimate corresponding to stable net debt-to-GDP ratios over a 75 TABLE 4 TABLE finances of the nation n 1115

Reforms that potentially increase this contribution may be important to address provincial fiscal gaps in future years. Increasing federal transfers is feasible, given the relatively large fiscal space available to the federal government. As shown in table 4, if the size of cash transfers were increased by 10 percent (over $8 billion in 2020 - 21, for perspective), the aggregate provincial fiscal gap would be reduced to 2.5 percent ofGDP . Increasing transfers by 25 percent would decrease the provincial fiscal gap to just under 2 percent. The federal government can also transfer tax points instead of cash to the provincial governments. Table 4 illustrates the effect if the federal government vacated the entire sales tax field, leaving it to the provinces: the provincial fiscal gap would shrink from 2.9 percent to 0.9 percent. Historically, tax point transfers were central to fiscal arrangements in Canada, although we have moved away from this approach in recent years.22 The increases imagined in the foregoing scenarios are undeniably large—larger than is realistically on offer—but provide an important sense of scale. Finally, some targeted measures to support Newfoundland and Labrador may be necessary given that province’s precarious fiscal position. I estimate that without the federal government’s stream of payments to Newfoundland and Labrador, totalling $2.5 billion under the 2019 Atlantic accord, the province’s fiscal gap would be roughly 0.2 percentage points higher. Most important, however, is the lack of equalization payments to Newfoundland and Labrador, compared to the other Atlantic provinces. If natural resource revenues were excluded, Newfoundland and Labrador would qualify for equalization. The province’s average income is higher than that in the other Atlantic provinces, so it would not receive as much, but I find that its fiscal gap would decline to 8 percent. Given that demographics and health-care costs are such an important driver of provincial fiscal challenges, specific changes in federal support for health-care -ex penditures may be warranted. In the 2019 federal election campaign, the Bloc Québécois (BQ) put forward a proposal to allocate the CHT on the basis of the provincial population aged 65 and over, rather than the current per-capita allocation.23 This “needs-based” approach to the CHT would benefit provinces with older populations, cost those with younger populations, and leave the aggregate provincial fiscal gap unaffected. To shrink the aggregate gap, faster growth is necessary. TheBQ also proposed an increase in the CHT growth rate above current levels. In table 5, I report the effects of a change in the allocation and a change in the pace of growth. A sustained increase of 2 percentage points per year in CHT transfers would shrink the aggregate provincial fiscal gap to 0.7 percent and consume almost all of the

22 For a comprehensive review of the history of federal-provincial transfers, including tax point transfers, see Trevor Tombe, “‘Final and Unalterable’—But Up for Negotiation: Federal- Provincial Transfers in Canada,” Finances of the Nation feature (2018) 66:4 Canadian Tax Journal 871 - 917. 23 Bloc Québécois, Le Québec, c’est nous : Plateforme politique du Bloc Québécois (Quebec: BQ, 2019) (www.blocquebecois.org/wp-content/uploads/2019/09/Plateforme_Bloc2019_web.pdf ). 1116 n canadian tax journal / revue fiscale canadienne (2020) 68:4 0.9 4.3 3.1 1.3 2.1 1.5 1.1 2.4 7.4 1.7 − 1.8 − 1.6 related to an Cover all costs aging population Deeper reforms 4.8 3.4 1.9 2.8 2.4 3.5 8.7 2.5 1.8 3.1 − 0.9 − 2.4 Increment to CHT growth 3.3 1.7 1.0 0.4 1.3 7.6 0.7 0.1 1.2 -year time horizon. − 0.3 − 2.9 − 0.3 per year 2 percent Increase of Faster CHT growth 4.4 2.9 1.2 2.4 2.0 3.0 8.7 2.0 1.5 2.8 − 1.4 − 1.9 per year 1 percent Increase of 5.3 3.9 2.5 3.2 2.5 3.8 8.7 2.9 2.1 3.1 − 0.6 − 2.8 population Allocate CHT on basis of 65 + 5.1 3.6 2.2 3.2 3.0 4.1 9.4 2.9 2.3 3.7 − 0.4 − 2.8 Baseline fiscal gap (Percent Financing Reforms Health-Care of GDP) Various . . Estimates for Fiscal Gap . . . Source: Author’s calculations. See the text for details. Source: Author’s TABLE 5 TABLE Province Alberta Saskatchewan Manitoba . Ontario Quebec . Nova Scotia . Prince Edward Island . Newfoundland and Labrador . Provinces . Federal British Columbia New Brunswick . CHT = Canada health transfer; GDP gross domestic product. Note: This table displays the fiscal gap estimate corresponding to stable net debt-to-GDP ratios over a 75 finances of the nation n 1117 long-run fiscal space available to the federal government. This would be a significant increase that, over time, would increase the federal share of health-care expenditures to 30 percent by 2040 and to 40 percent by 2060. Beyond these simple options, more fundamental reforms are worth considering. In the two options that follow, I take care not to propose policies that directly expose the federal government to the spending decisions of any specific provincial government. Such policies would have the effect of subsidizing provincial spending increases and potentially lead to greater inefficiencies in the delivery of important public services.

Index CHT Growth to Demographics Currently, the CHT grows at the same rate for all provinces, but different provinces have different rates of population growth for different age cohorts. Population trends are, to some extent, beyond the provincial government’s control. However, health-care spending by age category is a policy choice and depends on related decisions concerning public-sector compensation, hospital capacity and location, and so on. So, instead of assuming uniform CHT growth, we could measure cost pressures using a nationally representative measure of health-care spending and provincial demographic changes. For Canada as a whole, an aging population adds to health- care costs when the population shifts toward higher-spending cohorts, as illustrated c c in figure 5. The rate of increase in national health-care expenditures is ∑ c h 0 p t , c c where h 0 is the initial period of health-care spending on cohort c and p t is the population share accounted for by that cohort. A national average health-care spending measure applied to each province’s population shares could be a way to grow health- care spending in a relatively exogenous manner. Specifically,

h c p c ∑______c 0 it ′ c c CHT growth increment = c c ≡ ∑ c w 0 pˆ it . (18) ∑ c h0 p it This expression mirrors the province-specific measure of demographic cost pressures h ˆ it defined earlier, but uses national average health-care spending per capita by cohort instead. This increment would see CHT transfers grow faster for all provinces, but at different rates. By 2040, the increase in CHT transfers would range from 11 percent more than the baseline projection for Saskatchewan to 40 percent more for Newfoundland and Labrador. The federal share of health-care spending would decline gradually in the coming years, to less than 23 percent by 2040. Therefore, this option is a relatively modest approach to indexing the pace of CHT growth to demographic factors. It would lower the aggregate fiscal gap for provinces by roughly 0.4 percent of GDP.

Supplement the CHT To Cover All Demographic Costs Real per-capita spending on health care is projected to rise in response to changing demographics and health-care-specific inflation over and above the general rate of 2 percent per year. Since the latter is more a policy choice than the former, the federal government could cover more of the costs related to population aging than 1118 n canadian tax journal / revue fiscale canadienne (2020) 68:4 it does through the current system of funding health-care costs generally. Many elderly individuals move into other provinces in their retirement years, and to the extent that they do, the case for federal support to provinces to which those individuals relocate is potentially strong. One option is for the federal government to fund health-care costs related to aging but not other provincial health-care spending decisions. This approach takes the CHT increment proposed above and shifts incremental aging costs entirely to the federal government. Specifically, _ c c CHT supplement to fully cover aging costs = h 0 × ( ∑ c w 0 pˆ it - 1), ( 1 9 ) _ where h 0 is the national average real per-capita initial level of health-care spending. This option would result in a large increase in federal transfers, but it would provide federal support for an aging population without being susceptible to provincial health-care spending decisions. Only population shares would change over time, and those would result largely from the decisions of individual Canadians. The grant under this formula would gradually increase to nearly 1.5 percent of GDP by 2050 and decline thereafter. This level of support is less than the total projected health-care expenditure increase since it does not compensate for health- care-specific inflation, but only for aging-related cost increases. But the contribution is still large. The federal government’s share of total health-care spending would increase from the current level of one-quarter to a peak of one-third by the mid- 2040s. Without such a transfer, the CHT is on track to grow at a slower rate than health- care spending, and therefore its share of the total will decline to roughly 18 percent by 2050. Expressed another way, the provincial share of health-care spending will rise from the current level of 5.5 percent of GDP to 6.5 percent by 2050 and 7 percent by 2070. In the baseline projection without the age-related supplement, this share rises to 8 percent by 2050 and nearly 8.5 percent by 2070. This option could result in a significant improvement in provincial debt sustainability. I estimate that the aggregate provincial fiscal gap would decline to 1.7 percent of GDP from the baseline 2.9 percent. If, along with this new transfer, provincial governments ensured that their own-source revenues kept pace with economic growth, or if health-care- specific inflation were limited to 0.5 percent, the aggregate provincial fiscal gap would be reduced to almost zero. These exercises, it must be said, do not account for the important behavioural changes that increased federal transfers may induce among provincial governments. Easy money from Ottawa may be as open to abuse as natural resource revenues have proved to be—transfers may encourage provincial governments to increase their spending in order to gain a short-term political advantage. Federal transfer arrangements that are impervious to such abuse are difficult to design. Canada has grappled with this challenge since Confederation, and it is a core challenge of fiscal federalism generally.24

24 For a comprehensive review of this issue, see Jonathan A. Rodden, Hamilton’s Paradox: The Promise and Peril of Fiscal Federalism (Cambridge, UK: Cambridge University Press, 2005). finances of the nation n 1119

THE EFFECT OF COVID-19 AND OTHER MACROECONOMIC DEVELOPMENTS No analysis of public debt sustainability today can ignore the effect of COVID-19. The pandemic has caused the largest disruption to economic activity, government finances, labour markets, business operations, and indeed daily life since the Second World War. The consequences of this shock will be felt for many years to come. And the large deficits that governments are incurring in 2020, and potentially for many years to come, are reasonably raising concerns over the long-run sustainability of public debt. At the time of writing, much about the broader fiscal and economic disruptions associated with COVID-19 and the public health response to it is unknown. Canada’s Federal Fiscal and Economic Snapshot 2020 25 provides rich detail, but substantial uncertainty remains. In this section, I propose a first attempt to quantify the effect of a large-scale economic shock on provincial (and federal) debt sustainability. This exercise should be interpreted as illustrative in nature, and therefore distinct from the main analysis. It has value nonetheless. I show that even a shock as large as COVID-19, which has led to the largest deficits since the Second World War and economic disruptions rivalled only by those seen during the Great Depression, may not have as great an impact on long-run sustainability as one might initially imagine. Perhaps counterintuitively, I demonstrate that COVID-19 may have improved the long-run position of provincial governments despite its short-run costs. I will begin by describing some of the details behind the fiscal scenario. I model two components of the COVID-19 shock. First, all provinces have experienced a large and persistent reduction in the level of economic activity. Current projections for the effect of the pandemic on nominal GDP vary, but I use the June 10, 2020 projections for provincial GDP in 2020 and 2021 from the Royal Bank of Canada.26 From 2022 onward, I assume a gradual recovery to the pre-COVID-19 baseline trajectory of nominal GDP by assuming that one-third of the remaining gap in each period is closed through above-normal growth. While prospects are potentially optimistic, the economic shock of the pandemic will continue to be felt until 2029, the earliest estimated date for the return of all provinces to GDP within 1 percent of their pre-COVID-19 path. Second, I assume a large federal spending response, amounting to $300 billion, that both supports individuals and businesses, and (importantly for our purposes here) cushions provinces against incremental expenditure pressures from the pandemic. The underlying presumption is that the federal government intends to provide funding for the recovery far beyond the $19 billion safe restart agreement already committed to. This exercise serves to illustrate potential

25 Canada, Department of Finance, Federal Economic and Fiscal Snapshot 2020 (Ottawa: Department of Finance, July 8, 2020). 26 Robert Hogue, “Reopening of Provincial Economies: Different Speed, Scale and Outcomes,” Royal Bank of Canada Economics, June 10, 2020 (https://thoughtleadership.rbc.com/reopening -of-provincial-economies-different-speed-scale-and-outcomes). 1120 n canadian tax journal / revue fiscale canadienne (2020) 68:4 magnitudes and is a reasonable approximation of the true fiscal and economic shock based on current information. Accordingly, this scenario anticipates a significant negative shock with long- lasting effects. In the short term, the federal debt ratio rises to 52 percent of GDP by 2021 and provincial debt rises to nearly 49 percent. The combined effect is a 24 percentage point increase in government debt-to-GDP ratios—a modestly larger effect than the latest projections from the IMF. Looking ahead, I estimate that federal primary balances do not return to surplus until 2023 but remain permanently below the pre-COVID-19 baseline. By 2030, the federal primary balance is roughly 0.15 percentage points of GDP below the pre-COVID-19 trajectory and remains roughly 0.12 percentage points below across the entire forecast horizon. Meanwhile, the provinces see worsened primary balances until the late 2020s but afterward have a smaller primary deficit than previously projected. In terms of the 75-year horizon fiscal gaps, under this scenario, the federal position declines to −2.5 percent from the baseline −2.8 percent while the provinces’ aggregate fiscal gap decreases by 0.1 percentage points. I illustrate these results in figure 11. An important aspect of this scenario is the interaction between large economic shocks and federal transfers—specifically, the benefit that provinces derive from health transfers that automatically grow larger. The CHT grows with a three-year moving average of nominal GDP growth but, importantly, has a minimum growth of 3 percent per year. During particularly severe periods of economic contraction, such as that experienced in 2020, the moving average growth in GDP will decline below the 3 percent minimum threshold. It may remain bound by this growth floor until 2023, when the sharp 2020 contraction will be dropped from the moving average. At that point, average growth should exceed the previous baseline growth because Canada will continue to be recovering to its potential level of output. In order to return to the pre-COVID-19 baseline path of economic activity, some above-normal growth is necessary. This will then result in larger growth in the CHT than would have been the case from 2023 until recovery is complete and normal growth returns. In this scenario, I find that by 2030 the total CHT is nearly 5 percent larger than it would have been absent the COVID-19 shock. This is meaningful and represents an increase of roughly 1 percentage point in the share of total health spending covered by the federal government. By dropping large contractions but counting recovery growth rates, the CHT is set to ratchet permanently up to a modestly higher level. In time, this more than offsets the short-term debt that provincial governments incur because of COVID-19.

CONCLUSION This article develops a comprehensive model of provincial and federal finances, and projects future debt ratios over a wide variety of scenarios. I find that most provinces, with the notable exception of Quebec, face significant long-run challenges owing to an aging population, falling rates of economic growth, and rising health-care costs. Meanwhile, the federal government faces an excess of fiscal capacity and enjoys finances of the nation n 1121

FIGURE 11 Post-COVID-19 Debt Sustainability in Canada, 2018-2050 (Forecasted) (a) Projected debt-to-GDP ratios 125

100

50 hare of GDP () 25

0 2020 2025 2030 2035 2040 2045 2050 ederal government Provincial governments ederal pre-COVID-1 Provincial pre-COVID-1 baseline baseline

(b) Projected aggregate primary balances of provincial governments 0

−1

−2 hare of GDP ()

−3 2020 2025 2030 2035 2040 2045 2050 Current projection Pre-COVID-1 baseline

GDP = gross domestic product. Note: This figure displays the projected debt-to-GDP ratio for Canada’s governments in the baseline scenario (dashed line) compared to the post-COVID-19 scenario. Debt to GDP for 2018 is from actual data; values for 2019 onward are fiscal projections. The COVID-19 shock occurs in 2020. Negative primary balances are deficits. Source: Author’s calculations. See the text for details. 1122 n canadian tax journal / revue fiscale canadienne (2020) 68:4 a sustainable financial position despite the massive debt accumulated in response to COVID-19. Combined, Canada’s general government (federal plus provincial) is sustainable in the long run, making the challenge for provinces one that may involve changes in federal-provincial fiscal arrangements, ranging from increased and reformed cash transfers to tax point transfers. Current transfers have significantly contributed to provincial finances, and for relatively lower-income regions (the Maritimes in particular) equalization mitigates what would otherwise be potentially intractable financial challenges. While Newfoundland and Labrador may require a unique approach, there are a wide variety of gradual policy options available to the provinces, and to Ottawa, to overcome the fiscal challenge presented by an aging population. The scenarios set out in this article are but a small sample of the potential fiscal futures that might unfold. But however the fiscal reality evolves, it is important to consider carefully the potential implications of current policy choices for future fiscal outcomes. Policy makers today can take gradual and sustained actionto avoid more dramatic changes later. And such action should be guided by the kind of analysis put forward in this article. Whether rebuilding fiscal capacity following a short-term shock or preparing for predictable long-term pressures, governments have both the tools and the data to respond today to the fiscal challenges of tomorrow. canadian tax journal / revue fiscale canadienne (2020) 68:4, 1159 - 72 https://doi.org/10.32721/ctj.2020.68.4.ustd

Selected US Tax Developments Co-Editors: Peter A. Glicklich* and Michael J. Miller**

NEW PROPOSED REGULATIONS UNDER THE SECTION 1061 CARRIED INTEREST RULES Peter A. Glicklich and Gregg M. Benson***

The Tax Cuts and Jobs Act of 2017 added new section 1061 to the Internal Revenue Code. That provision added a three-year holding period for fund managers holding “carried interests” to qualify for long-term capital gain treatment. On August 14, 2020, the Internal Revenue Service published proposed regulations that provide much-needed guidance. As proposed, however, the new rules are quite complex and include a number of traps for the unwary. We discuss some of the key provisions of the proposed regulations and consider their impact for Canadian funds and their Canadian and US managers. KEYWORDS: PARTNERSHIPS n FUNDS n CAPITAL GAINS n RECHARACTERIZATION n REGULATIONS n INTERNAL REVENUE CODE

* Of Davies Ward Phillips & Vineberg LLP, New York (e-mail: [email protected]). ** Of Roberts & Holland LLP, New York and Washington, DC (e-mail: [email protected]). *** Of Davies Ward Phillips & Vineberg LLP, New York.

1159 canadian tax journal / revue fiscale canadienne (2020) 68:4, 1173 - 89 https://doi.org/10.32721/ctj.2020.68.4.ctr

Current Tax Reading Co-Editors: Robin Boadway, Kim Brooks, Jinyan Li, and Alan Macnaughton*

Trevor Tombe, “An (Overdue) Review of Canada’s Fiscal Stabilization Program,” IRPP Insight no. 31, February 2020, 1-25 The title of this paper is apt. The fiscal stabilization program is due for reform. Trevor Tombe explains why, and in the process he offers a timely lesson in fiscal stabilization. This program, under which the federal government compensates provinces whose revenues decline extraordinarily in a given year, has been un- changed for some time and has two key features that some find objectionable. First, annual stabilization payments are limited to $60 per capita, an amount that, besides being small, is not indexed and so is declining in real terms. Second, only 50 percent of a decline in natural resource revenues is eligible for stabilization, compared with 95 percent for a decline in non-resource revenues. These two restrictions seem particularly stringent for resource-producing provinces whose revenues are subject to large shifts from time to time. Tombe outlines two reasons for the federal stabilization of provincial revenues. One is a standard insurance argument based on risk pooling. The federal government, because of its size and fiscal capacity, is able to pool fiscal shocks faced by the provinces. In so doing, it lowers provincial borrowing costs and facilitates provincial borrowing. Another argument for federal stabilization of provincial revenues is based on equity. The nation as a whole collectively bears the burden of shocks that are beyond the control of the provinces, particularly low-income provinces. Tombe notes that these arguments for fiscal stabilization are increasingly important because provincial GDPs have become increasingly variable. The volatility of provincial GDP growth rates relative to the national average has increased significantly in the last five decades. A shortcoming of stabilization, arguably, is that some higher-income provinces (resource-rich provinces, for example) that could self-insure against shocks choose not to do so. A more general shortcoming is that stabilization, like any insurance

* Robin Boadway is of the Department of Economics, Queen’s University, Kingston, Ontario (e-mail: [email protected]). Kim Brooks is of the Schulich School of Law, Dalhousie University, Halifax (e-mail: [email protected]). Jinyan Li is of Osgoode Hall Law School, York University, Toronto (e-mail: [email protected]). Alan Macnaughton is of the School of Accounting and Finance, University of Waterloo (e-mail: amacnaughton@ uwaterloo.ca).

1173 1174 n canadian tax journal / revue fiscale canadienne (2020) 68:4 policy, can induce moral hazard. Provinces may take excessive fiscal risks, for example, when the fiscal consequences are borne by others. Tombe notes that this is particularly true for resource-rich provinces, and it is the reason why the fiscal stabilization of resource revenues is subject to a deductible. Tombe provides a succinct historical summary of the evolution of the fiscal stabilization system. He notes that the $60 cap was introduced with little analysis and was originally meant to be the dividing line between grants and interest-free loans, although the latter was never used. He summarizes the anomalies of the existing system, using Alberta as an example, and he discusses briefly how equalization provides at least partial stabilization. The author would have done well to emphasize the latter effect. Studies have shown that the equalization system, far from providing stabilization, can increase revenue volatility for some provinces because the national average revenue capacity on which equalization payments are based is itself volatile. That puts more onus on fiscal stabilization as an insurance device. Tombe offers three policy alternatives. The first approach would be to reform the existing system to address the shortcomings discussed above. This would involve changing the per capita ceiling or lowering the deductible. He illustrates, using the example of Alberta for the fiscal years 2015-16 and 2016-17, the effects of increasing the ceiling to $120 and $170 per capita and of eliminating the ceiling altogether. He then considers three deductible combinations: 5 percent for all revenue sources, zero, and zero for non-revenue sources combined with 50 percent for resource revenues. He provides an interesting discussion of considerations that determine the appropriate deductible. The second approach would harmonize fiscal stabilization with equalization. In- stead of stabilizing actual provincial revenues, fiscal stabilization would be based on changes in fiscal capacity as measured by the equalization system. In measuring provincial fiscal capacity, the existing equalization system includes 50 percent of resource revenues. For the Alberta case, Tombe shows what fiscal stabilization payments would be under different rates of inclusion—50 percent, 100 percent, and zero, assuming a 5 percent deductible. He also considers the consequences of excluding property taxes from fiscal capacity and of using a two-year moving average. The final approach determines fiscal stabilization according to a macro-based formula, such as provincial GDP. Such an approach is simpler than one that uses revenue-raising capacity, and it minimizes moral hazard problems. At the same time, macro-indicators are imperfect measures of provincial fiscal capacity. Tombe concludes with his preferred policy prescription. He would eliminate the per capita ceiling on stabilization payments or—if that is deemed undesirable—ease the ceiling to reflect what it would have been had the ceiling been indexed to inflation from the outset. At the same time, he would leave the 50 percent deductible in place for resource revenues, on the grounds that the resource-rich provinces have the capacity to self-insure by saving resource revenues. More stabilization would discourage these provinces from using a sovereign wealth fund to smooth their resource revenues over time and from implementing more stable revenue sources current tax reading n 1175 such as a harmonized sales tax (HST). He also proposes that the federal government initiate a comprehensive review of the fiscal stabilization program, a process of reform that could include aligning the program more closely with equalization. Equalization would address persistent shocks, while fiscal stabilization would address temporary ones. R.B.

David Green, Jonathan Rhys Kesselman, and Lindsay Tedds, “Considerations for Basic Income as a Covid-19 Response” (2020) 13:11 SPP Briefing Papers [University of Calgary School of Public Policy] 1-15 (http://dx.doi.org/10.11575/sppp.v13i0.70353) Discussions of basic income have been prominent since the government introduced, in March 2019, the Canada Emergency Response Benefit (CERB) in order to cushion the income shock that befell many workers as businesses shut down because of the COVID-19 lockdown. During the same period, members of the BC expert panel on basic income were in the midst of their research and deliberations about whether a basic income was a suitable policy instrument for the province to use in its poverty reduction strategy. The three co-commissioners of the BC panel, David Green, Jonathan Rhys Kesselman, and Lindsay Tedds, have yet to settle on recommendations, but they have studied the issue enough to set out what they consider to be key considerations in proposing a way forward. They are quick to note that they have not yet formed a judgment on arguments for and against basic income, and they emphasize that a first step toward such judgment is understanding and analyzing the issue. In that spirit, the authors document the main questions turned up by their analysis. The top priority is to establish the overarching principles of basic income policy, such as fairness, social justice, dignity, security, and simplicity. More specifically, what are the objectives? The objectives, which may be multi-dimensional and even conflicting, can include reducing poverty, responding to changing labour-market conditions, and rationalizing the transfer system. The next step is to specify the details of a basic income program. Should it be universal, or conditioned on income? Should it be delivered through the income tax system, like refundable tax credits and the CERB? Criteria for eligibility must be specified—based on, for example, age, income, assets, and, possibly, work search activities. Should the level of basic income be related to poverty measures such as the market basket measure, and should it differ for those with disabilities? How should basic income interact with other transfer programs such as social assistance and, presumably, employment insurance (EI)? Should benefits be paid to individuals or to families, and, if the latter, should the benefits’ size be related to family size? In addition, administrative details need to be decided, such as how recipients should be identified—whether by income tax filing, for example, or some other measure. As for a delivery system, the benefits could be delivered by bank transfers, 1176 n canadian tax journal / revue fiscale canadienne (2020) 68:4 though special measures will be required for recipients without bank accounts. The frequency of payments must be decided, along with procedures for dealing with changes in individual recipients’ circumstances. Another important question is: How will the program and changes made to it be communicated to the public? A basic income involves two other critical design issues. One is what benefit reduction rate to choose, a decision that affects both the cost of the program and individuals’ responses to it. The benefit reduction rate could be fixed, or it could vary if an objective is, for example, to influence recipients’ decisions about participation in the labour market. The second design issue is how a basic income program would be integrated into the fiscal federalism system. This is important, because both federal and provincial-territorial levels of government currently deliver income transfers to different segments of the population. Another major issue concerns how much a basic income would cost and how it would be financed. Existing studies have shown that much of the financing could be achieved, in principle, by a reallocation of existing refundable and non-refundable tax credits and other transfer programs. Other options would include eliminating some tax expenditures and using other sources of tax revenue. The costing of a basic income would be affected by how individual and firm behaviour will respond to the program, and little can be known about such responses beforehand. Finally, the authors ask whether a basic income—as opposed to, say, a reform of the myriad existing transfer programs and public services—is the best way to achieve the stated objectives. The authors suggest that until these questions are answered, decisions about basic income will be premature. The exhaustive analysis that they recommend is a tall order. In fact, many of these issues have already been recognized in the literature and studied to some extent. As with any major reform, policy decisions about the basic income must be based on imperfect information and judgments formed according to best practices. Perfection is impossible, especially when it comes to addressing administrative issues. Options such as gradual implementation should not be off the table. This paper provides a very useful catalogue of issues relevant to a decision whether to proceed with a basic income. That said, a few considerations are missing from this catalogue. One consideration is whether the basic income may be viewed as an investment, with recognizable social benefits such as improvements in health, education, economic (and nutritional) security, reductions in criminal and anti-social behaviour, and the breaking up of the intergenerational poverty cycle. A second consideration is implementation. Is there a case for implementing the basic income through incremental reforms from which one might learn a great deal? Finally, it is important to consider that the inclusion of First Nations members in a basic income—a very relevant concern, given this population’s need for income support— poses important jurisdictional and delivery issues. R.B. current tax reading n 1177

Canada, Office of the Parliamentary Budget Officer, Costing a Guaranteed Basic Income During the COVID Pandemic (Ottawa: PBO, July 2020) (www.pbo-dpb.gc.ca/web/default/files/Documents/Reports/ RP-2021-014-M/RP-2021-014-M_en.pdf )

Basic income is in the news as a potential successor to the CERB, as the latter program expires. Most proposals consider basic income as a form of refundable tax credit, delivered through the tax system, that replaces some existing refundable and non-refundable tax credits. Accordingly, the implementation of a basic income would be akin to a tax reform. This Parliamentary Budget Officer (PBO) report has adopted an existing methodology to estimate what the cost of a basic income guar- antee (BIG) would be in the immediate aftermath of the pandemic, when the CERB and other federal government assistance programs end. The program being costed by the PBO involves the provision of a nationwide BIG to eligible adults aged 18 to 64 for the relevant six-month period (October 2020 to March 2021), delivered through the income tax system. The BIG benefit level is set at about 75 percent of the “low-income measure” poverty level. Three different tax-back (benefit reduction) rates are used to reduce benefits as taxable income rises: 50 percent, 25 percent, and 15 percent. The report follows several existing studies that use Statistics Canada’s SPSD/M model to cost an income-tested BIG program on the basis of detailed tax information. A major difference is that the model data are scaled up to take account of the depressed income levels caused by the pandemic lockdown. The PBO reports two main findings. The first is the gross cost of the BIG—that is, the total value of the transfers that would be paid out—according to the various tax-back scenarios. With a tax-back rate of 50 percent, the gross cost is about $46 billion for the six-month period. The cost rises to $72 billion with a tax-back rate of 25 percent, and to $96 billion with a 15 percent rate. In its second main finding, the report calculates the amount of program financing that could be found by cancelling certain federal and provincial programs. In the base-case estimate, these programs include provincial social assistance, most refundable tax credits (except the Canada Child Benefit), and non-refundable tax credits that apply mostly to disability, caregiving, and medical expenses. The value of the offsets is about $15 billion, which is a very small proportion of gross BIG costs. In an alternative case, the report includes cancellation of the basic personal amount as a source of financing. This results in offsets of about $46 billion, which covers the cost of a BIG with a tax-back rate of 50 percent. These estimates are comparable to those obtained in previous BIG costing exercises that interpret the implementation of a BIG program as a tax reform exercise. The approach in the report raises a number of issues. Because the study applies only to the immediate post-pandemic period, it overstates the costs of a BIG over the longer term, when incomes are higher. The benefit level assumed is comparable to 1178 n canadian tax journal / revue fiscale canadienne (2020) 68:4 the level in the Ontario basic income pilot project, but it is lower than the poverty level used in other studies. Similarly, the lower two of the benefit-reduction rates are well below those typically assumed. The report is silent on the question of provincial participation, including whether it is likely that the provinces would choose different provincialBIG rates. The choice of which tax credits to eliminate is unfor- tunate. As mentioned above, the ones chosen include those applicable to disability, caregiving, and medical expenses, which arguably complement a BIG. The effects of the BIG on inequality and on taxpayer behaviour are not addressed, although these effects are challenging to estimate. Perhaps the biggest limitation of the PBO study is that it focuses solely on the costs of a BIG and ignores the benefits to society. The latter include benefits that arise from nutritional and health improvements for the poor, greater educational attainment, enhanced opportunities for the children of the poor, and reductions in crime. The extent of these benefits is difficult to measure, of course, and they have typically been omitted from previous studies. R.B.

Jennifer Robson, “Radical Incrementalism and Trust in the Citizen: Income Security in Canada in the Time of COVID-19” (2020) 46, Supplement 1 Canadian Public Policy S1-18 (https://doi.org/10.3138/cpp.2020-080) One of the key innovations in tax administration has been the use of the Canada Revenue Agency (CRA) to deliver targeted transfers. Refundable tax credits such as the goods and services tax/harmonized sales tax (GST/HST) credit, the Canada Child Benefit (CCB), and the Canada Worker Benefit, along with various income- tested non-refundable tax credits, have become important elements of social policy. What distinguishes their administration is that they are based on self-reporting by taxpayers that is backed up by auditing to encourage compliance. In contrast, social assistance, EI, and the Canada Pension Plan rely on ex ante application to determine eligibility. In this paper, Jennifer Robson refers to the self-reporting approach as a trust-but-verify system, as opposed to a verify-then-trust gatekeeping system. The trust-but-verify approach was put to the test in the delivery of the income security policies that were enacted to deal with the income losses caused by the COVID-19 lockdown. Robson discusses why the federal government opted for this approach, considers some of the challenges involved in this approach, and opines on its suitability for the reforms to income security that may be coming after the COVID-19 crisis subsides. Robson first reviews and evaluates the federal government’s main policy responses to the income shocks wrought by the crisis. She identifies three phases of policy action, which occurred sequentially from March to mid-June 2020. Phase 1 consisted of two relatively minor changes to existing policies. One of these changes simplified the application procedure for EI sickness benefits by waiving the one-week waiting period and the requirement of a medical certificate. This change, which increased current tax reading n 1179 the speed at which EI-eligible workers could receive benefits, marked a tentative move in the direction of trust-but-verify. The second change was to extend work- sharing agreements in order to reduce layoffs. Phase 2 was more substantial and involved significant amendments to existing policies. The amendments were a partial response to the fact that EI applications were increasing rapidly and creating a sizable backlog, and that many workers who were losing their incomes were not eligible for EI. The government took some measures to reduce financial obligations that individuals would have trouble meeting. For example, the government suspended payments of Canada student loans, reduced required registered retirement income fund (RRIF) withdrawals for older Canadians, deferred the financial loan and mortgage obligations of large financial institutions, and postponed the income tax filing-due date. Another measure involved a temporary top-up of the GST credit and the CCB. The federal government also instituted a modest temporary wage subsidy (TWS) of 10 percent up to a specified limit. The purpose of the TWS was to encourage firms to retain workers, but the takeup rate was disappointingly low. In phase 3, new income support programs were implemented to deal with the serious financial consequences that Canadian households faced because of public health measures imposed by the provincial government. Four provincial governments provided temporary cash payments to assist those hit hardest. Rather than deliver these payments through the social assistance system, the governments chose to rely mainly on self-identification by the recipients in order to get the cash to households quickly. Many provinces also suspended rent evictions, provided some rent assistance, and legislated leave-without-pay provisions to protect jobs. The most substantial income security measures, however, involved two new temporary federal programs, the CERB and the Canadian Emergency Wage Subsidy (CEWS). The CERB program provided a taxable payment of $2,000 per month for up to four months to those who had lost their jobs and who had earned at least $5,000 in the previous year. Neither those who had quit their jobs nor those drawing EI were eligible, and claimants were allowed to earn up to $1,000 while receiving the CERB. The program was rolled out very quickly by comparison with other transfer or social insurance programs. This speed was owing to minimal, simple eligibility rules and to the trust-but-verify basis of the program’s administration. The CRA made payments on the basis of applicants’ attestations rather than waiting for their documents to be vetted. Robson carefully discusses the details of the trust-but-verify approach (for example, the takeup rate) and its possible consequences (for example, the potential for fraudulent claims). One weakness in the CERB program was that many students and recent graduates were not eligible for it, because of the “prior earnings” requirement. To address that, the federal government introduced the Canada emergency student benefit CESB( ), which was a smaller version of the CERB that paid $1,250 per month. Unlike CERB applicants, CESB applicants were required to search for work. Like the CERB, the CESB was administered by the CRA, which required that a tax return be filed. Since 1180 n canadian tax journal / revue fiscale canadienne (2020) 68:4 many students had not filed a return, this requirement represented a small obstacle that may prove important in future transfer programs administered by the CRA. The CEWS was intended to reduce job loss. It paid 75 percent of an employee’s wages up to a maximum of $847 per week. It was available to firms whose revenues declined by 30 percent and to employees who continued to be paid though they were not necessarily working. The CEWS program, like the CERB, is administered by the CRA, and eligibility is based on a trust-but-verify procedure, although firms that apply for it must provide detailed statements as part of their application. Docu- mentary evidence need not be provided by applicants, but it must be available on request. The CERB and the CEWS cannot be paid to the same employee, and Robson notes that enforcement of this rule may be complicated for the CRA. To discourage double dipping, the CRA could audit randomly selected recipients, but Robson suggests that past reviews of the CRA by the Office of the Taxpayers’ Ombudsman point to the possibility of inconsistent and unequal treatment of taxpayers. Robson concludes with a broad evaluation of emergency income-support policies. She lauds the incremental process whereby changes were gradually phased in and changes were enacted in response to initial experience. Revising programs on the fly was an effective strategy for policies that represented significant departures from existing ones and were enacted quickly. Perhaps the most important inno vation was the reliance on a trust-but-verify administrative approach. This was not without precedent, given that the income tax system and the accompanying refundable tax credits are based on the same approach. But this approach represents a significant departure from other major transfer systems such as social assistance and EI, and it has been touted by proponents as a model for a BIG. In Robson’s judgment, the trust-but-verify approach could be suitable for future reforms to the income support system, particularly those delivered by the CRA. Her main concerns are twofold. First, when it comes to verifying the applicants’ eligibility, procedural fairness will have to be ensured. Second, in order for the CRA to use a trust-but-verify approach, it will be important to verify information on the recipient’s within-year income, so that timely responsiveness to changes in that income are assured. R.B.

Stephen Gordon, “The Incidence of Income Taxes on High Earners in Canada” (2020) 53:2 Canadian Journal of Economics 437-59 (https://doi.org/10.1111/caje.12433) The persistently high levels of after-tax income inequality (especially the concentration of top incomes) in Canada and their effect on tax policy have stimulated the current debate on tax reform. Policy options include increasing top marginal tax rates, broadening the tax base, closing tax loopholes, and instituting a wealth tax or inheritance tax. Standard arguments for limiting increases in tax progressivity have relied on the elasticity of taxable income—that is, the proportionate fall in reported current tax reading n 1181 income with an increase in the marginal tax rate. Relatively high estimates of the elasticity at top income levels can partially frustrate the effect of increasing top tax rates by causing the base to shrink. It appears that high-income earners’ opportunities for tax avoidance largely account for these earners’ higher elasticities. Gordon takes a broader approach. He argues that tax avoidance, along with the ability of high-income earners to move between Canada and the United States, can render an increase in top tax rates ineffective in addressing top-income concentration. In fact, higher top tax rates can lead to offsetting increases in before-tax top incomes, and can worsen income inequality. A higher tax rate can cause top-income earners in Canada to bargain for higher incomes, with a view to maintaining parity with what their earnings would be if they took a job in the United States. At the same time, the taxable incomes of high earners fall to the extent that they can avoid paying the higher taxes by reporting a smaller proportion of their incomes. If these two effects become significant enough, the tax increases are shifted away from top- income earners, and their after-tax incomes do not fall. The paper constructs a simple economic model to capture these influences, and uses it as a basis for empirical estimation. Two parameter estimates are relevant. One is a bargaining-power parameter, β, that relates proportionate changes in real income earned in Canada to proportionate changes in real income in the United States, adjusted for relative tax rates and the Canada-US exchange rate. The other is the share of income q that is not reported as income for tax purposes. The bargaining-power parameter β is estimated by regressing the log of real incomes in Canada at given quantiles of the income distribution against the log of tax-adjusted real incomes in the United States measured in Canadian dollars— that is,

* __Yt 1____ - t * S____t Yt log = a + blog ⋅ + et, [Pt ] [ 1 - t Pt ] where Yt /Pt is real income in Canada at time t, τ is the Canadian top marginal income tax rate, St is the exchange rate, and an asterisk indicates US values. The estimated value of β at the top 0.01 percent of incomes is about 0.7, falling to close to zero for the top 10 percent. The value for q is adopted from existing studies and is about 0.33 for the top 0.01 percent, and 0.11 for the top 10 percent. Gordon shows that after-tax incomes in a given income class will increase with an increase in the tax rate if β + q > 1. This condition is satisfied for those in the top 0.01 percent, and nearly satisfied for those near the top. For these high-income earners, increasing the top marginal tax rate will apparently have virtually no effect on after-tax income and therefore no effect on narrowing the income distribution. These results are highly suggestive and warrant at least a cautionary note about the ability of changes to the top tax rate alone to address the issue of top-income concentration. At the same time, there are some caveats. One is that the study, in determining the bargaining power of mobility, relies on an estimation of the effect of the top marginal income tax rate on reported incomes. The overall tax burden is 1182 n canadian tax journal / revue fiscale canadienne (2020) 68:4 probably the more important factor (for example, the average tax rate). Another caveat is that reducing tax-avoidance opportunities by broadening the income tax base would make tax changes more effective. As Gordon concludes, “This exercise is best viewed as a rough first pass at the question of the economic incidence of taxes on high earners in Canada.”1 It is an important exercise nonetheless. R.B.

Jennifer Winter, “Carbon Pricing in a Federal State: The Case of Canada” (2020) 18:1 ifo Dice Report 13-19 This short article offers a useful and succinct review of the political economy of carbon pricing in Canada. Though it was written for an international audience, Canadian readers will find enlightening its perspective on the peculiarities of how political opportunity, in combination with federal-provincial jurisdictional ambiguities, has shaped the evolution of carbon pricing. Winter begins by reminding us that the constitution is silent about jurisdictional responsibility for environmental policy. While the federal government has residual power over responsibilities that have not been assigned, provincial authority over civil and property rights suggests a provincial role in environmental policy. Supreme Court rulings have lent support to the notion that environmental policy is a shared responsibility. A shared responsibility is appropriate, given that some forms of environmental damage span provincial borders and therefore justify some collective national action. Like other policy areas, carbon pricing was originally a provincial initiative. Ironically, Alberta legislated the first emissions reduction policy in 2007, and it was followed by British Columbia’s carbon tax scheme in 2008. Quebec implemented a cap-and-trade program in 2013. Politics subsequently played a decisive role. The New Democratic Party (NDP) took office in Alberta in 2015 and announced a new climate plan that included a carbon tax on consumers and output-based pricing for large emitters. At the federal level, the Liberals came to power already committed to carbon pricing. In 2016, most of the first ministers agreed to implement a Canada- wide emissions pricing scheme, which included a minimum required level of emissions pricing, flexibility in the provinces’ choice of pricing mechanism, and a federal backstop when provinces failed to act. Ontario and New Brunswick subsequently joined in 2018 and 2019, respectively, but Alberta repealed its carbon tax when the Conservatives gained power. Carbon-pricing policies vary among provinces. They include carbon taxation, cap and trade, and a hybrid of carbon pricing and output-based pricing. Six provinces have carbon-pricing systems that meet the federal standard. The other provinces are subject to a federal backstop in whole or in part. It is a typically Canadian federal- provincial compromise.

1 At 457. current tax reading n 1183

As Winter points out, considerable policy convergence has been achieved, but divergence is beginning to show. The divergence is the result of four factors: (1) political distaste for carbon taxes, (2) tension between resource-development policy and emissions policy, (3) federal-provincial negotiations over whether provincial policies have met the minimum standard, and (4) changes in government. The strains in federal-provincial carbon-pricing policy are shown by the ongoing constitutional challenges that some provinces have mounted against the federal jurisdiction. Winter characterizes federal carbon pricing as a combination of reward- ing cooperative provinces and punishing non-cooperative ones. She sees a clear lesson for other federations, applicable when shared jurisdictions are at stake: federal policy action requires some agreement from provincial jurisdictions, but the inverse is not true. R.B.

Susann Sturm, Tax Complexity in Canada: A Comparative Perspective, TRR 266 Accounting for Transparency Working Paper series no. 20 (Paderborn, Germany: TRR 266 Accounting for Transparency, February 2020) (https://ssrn.com/abstract=3544366) Canada’s tax system is frequently characterized as complex, but the research on tax complexity in Canada lags behind equivalent work on complexity in the Australian and UK systems. Sturm’s paper fills the void. She compares the complexity of Can- ada’s income tax system, in the context of resident multinationals in 2016, with the systems in the other Organisation for Economic Co-operation and Development (OECD) countries. Sturm divides complexity into “tax code complexity” and “tax framework complexity.” With respect to code complexity, she concludes that the provisions on corporate reorganization, transfer pricing, and controlled foreign corporations are the most complex. With respect to framework complexity, she identifies tax audits, tax law enactment, and tax guidance as culprits. Sturm determines that Canada’s tax code complexity is slightly higher than the OECD average and that its tax framework complexity is similar to the OECD average. Her comparative work identifies some areas that require attention: the quality of drafts of tax legislation in the enactment process, the time period between the filing of an appeal and the rendering of a decision in the appeals process, and the skills of tax officers in the tax audit process. But she does not see the complexity of Canada’s corporate income tax system as particularly divergent from the OECD norm. Sturm’s paper is useful in several respects. First, she offers a review of the research on tax complexity. Second, she offers concrete advice on how policy makers might simplify tax policies and processes. Third, she offers some conceptual tools to help taxpayers distinguish between different types of tax costs/complexity, and she offers guidance on the relative complexity of different tax systems. K.B. 1184 n canadian tax journal / revue fiscale canadienne (2020) 68:4

Jonathan B. Forman and Roberta F. Mann, “Borrowing from Millennials To Pay Boomers: Can Tax Policy Create Sustainable Intergenerational Equity?” (2020) 36:3 Georgia State University Law Review 799-843. Tax policy analysts might be fairly criticized for “in the momentism.” Put another way, most analytical work on tax policy—assuming we agree that equity, efficiency, and administrability remain hallmark evaluative criteria for the design of tax systems— takes “now” as the window of analysis. With this work, Forman and Mann join a group of scholars who strongly advocate taking a longer view. In this paper, they develop the concept of “sustainable intergenerational justice.” Their argument is that tax policy is an effective tool for advancing intergenerational justice and that it ought to be deployed in a way that maximizes the resources available for future generations. Forman and Mann apply their analysis to taxes of all kinds—income, payroll, consumption, wealth, and property. Part 4 of their paper explores some tax policies that could be pursued to improve the intergenerational table and to promote an efficient economy. These proposals, which are tentative, include revising tax -ex penditures for oil, gas, and coal; increasing excise taxes on motor fuels; establishing a carbon tax; increasing subsidies for education; and curbing tax breaks for home ownership. K.B.

Anna Binder and Viktoria Wöhrer, Special Features of the UN Model Convention (Vienna: Linde Verlag, 2019) This collection of essays offers a wonderful glimpse into the future of international tax scholarship, policy making, and practice. Each of the 27 authors featured in the collection was a participant in the 2018-2019 LL.M program at Vienna University of Economics and Business. Each contribution explores a different aspect of theUN model convention. After two framing essays—one on the UN tax committee’s role as a player in the international tax policy discussion, and a second on the source-country leanings of the UN model—the essays focus on many of the UN model convention’s articles. There are explorations of articles 5, 7, 9, 12, 12A, 13, 14, 21, 23, 25, and the article on pensions (often 18). A concluding section offers six essays on aspects of the influence of the base erosion and profit shifting BEPS( ) project. A review of these essays is a worthwhile project for anyone interested in the UN model and its evolution. More than that, this collection is a magnificent reminder, for those of us who teach graduate students, of what our own programs can accom- plish when we work collaboratively with students on a common project. K.B. current tax reading n 1185

Philip Alston and Nikki Reisch, Tax, Inequality, and Human Rights (New York: Oxford University Press, 2019) Every now and then you come across a “who’s who?” kind of publication. This is one of those, in the context of tax and human rights. Work undertaken at the inter- section of tax and human rights has proliferated over the past decade. What was once a rare contribution has become a field. This collection comprises 25 chapters, pulled together from the proceedings of a 2016 conference and workshop at the Center for Human Rights and Global Justice at New York University. The contributors were primarily faculty members from law schools around the world, but several contributions are from UN staff, lawyers, and individuals associated with non-profit-sector bodies such as Christian Aid. Three Canadians contributed chapters: Allison Christians, Art Cockfield, and Kathy Lahey. Christians’s chapter explores state responsibility for human rights harms arising from tax law, policy, and practice. She articulates three possible approaches to raising claims about human rights in the context of taxation, and she canvasses the challenges that each approach faces in execution. Claims may be raised between an individual and their state; between the individual and a foreign state; and between the states themselves. The case studies (Glencore and Afrimex) in part II of the paper offer particularly useful examples of the difficulty of holding states accountable. Art Cockfield expands on his work on tax transparency. The primary objective of his chapter is to advocate for fairer and more efficient exchange-of-information policies. Cockfield opens with a compelling example (South Sudan) of the consequences of inadequate data transfer. More generally, he flags the vital importance of quality-of-information issues, capacity building, the use of data analytics, and the development of privacy safeguards. Simply put, Kathy Lahey’s paper is an argument for a change in the fundamental aim of taxation: stop taxing for growth and start taxing for (gender) equality. She highlights the growing global reliance on consumption taxes and the conflict between taxing consumption and human rights. Lahey has proven herself adept at both empirical and historical work, and much of this work highlights the ways in which tax laws, policies, and practices disadvantage women and other marginalized communities. Lahey invokes the role of the Convention on the Elimination of All Forms of Discrimination Against Women as a legal tool compelling more progressive tax strategies. K.B.

Tsilly Dagan, “Re-Imagining Tax Justice in a Globalized World” (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3602678) Tsilly Dagan offers another thoughtful instalment in her series of articles that grapple with the tax-justice justifications for the design of the tax system. In this paper, she groups theorists of the social contract into two categories: those who 1186 n canadian tax journal / revue fiscale canadienne (2020) 68:4 believe that the state has coercive power to promote the joint interests of its constituents (a market-inspired version of the state) and those who believe that the state is a vehicle to advance the collective interests of equal participants (a “community of equals” version of the state). Using this stylized approach to the rich literature on social contract theory, she tests different ways of designing the increasingly elective concept of tax residence. Ultimately, she determines that whatever model of the state one adheres to, the alignment of that view with the design of residence tests will have costs. K.B.

Shu-Yi Oei and Diane M. Ring, “When Data Comes Home: Next Steps in International Taxation’s Information Revolution,” McGill Law Journal (forthcoming) (https://papers.ssrn.com/sol3/papers.cfm ?abstract_id=3671133) The data exchanged between tax authorities, along with the tools available for analyzing the data, have ballooned. Oei and Ring’s careful article offers another of their typically thoughtful analyses—in this case, an analysis of the ways in which governments might use or fail to use data. The authors’ novel argument is that previous analyses have failed to account for domestic politics. The article outlines the needs that modern tax systems have for global information, and it (1) details the mechanisms that governments use to obtain such data, (2) connects data needs to technology and politics, (3) reminds readers of the various exchange agreements that exist around the world, and (4) speculates about what happens when these factors come together in the context of a particular country. The third part of the article, on the subject of “what is likely to happen as the revolution turns inward and data ‘comes home,’ ”2 is likely to be the most inter esting part for people who have been following the exchange-of-information developments. Here, Oei and Ring raise two concerns: How will the country process data, and how will it safeguard data? Countries require linkages between the data and their audit processes; they require enforcement resources; and they need to know how to protect data. The authors also look at some political risks involved in the use of data. Will some countries, for example, be reluctant to use the data? Will the use (or lack of use) of data become politicized? And will the data be used in ways that are unfair or inappropriately targeted? After exploring the uses of data, the authors set out eight variables that will likely affect domestic outcomes: the country’s level of interest in tax revenues and the tax base; the strength of the country’s tax administration; the country’s ability to navigate data-based regulation; the country’s commitment to privacy; the country’s previous experience with data integration; whether the country primarily imports or exports

2 At 30. current tax reading n 1187 data; the strength of the rule of law in the country; and the “wildcard” factors of COVID-19 and the OECD’s active tax projects. K.B.

European Economic Advisory Group, “Fair Taxation in a Mobile World” (2020) 19 EEAG Report on the European Economy 1-113

The European Economic Advisory Group (EEAG) is an independent advisory group of seven international economists affiliated with theCES ifo research network at the University of Munich. It analyzes policy issues from a European perspective, drawing on research-based insights. Although this report on tax policy emphasizes European issues, it will resonate in Canada because of its concern with issues that arise in economies with internationally mobile capital and skilled labour. The report provides a critical summary of tax policy reforms in such economies, focusing on the taxation of firms, individuals, and immobile bases, including wealth. The authors emphasize the issues of tax fairness that result from globalization and digitalization. The established consensus that firms and individuals should be taxed on the basis of residence and that consumption should be taxed by destination is no longer tenable. Firms with digital business models can readily operate with limited physical presence, relying heavily on intangible assets that are highly mobile. Likewise, the mobility of skilled and high-income workers compromises the ability to sustain fair tax policies. This report addresses the consequences of this digitalization and mobility for tax policy. The report was written before the COVID-19 pandemic disrupted global production and supply chains. Nonetheless, the conditions informing the discussion of tax policy will persist into the post-pandemic period. The report is policy-oriented and readable, and it is written in non-technical terms. It begins by analyzing the economic situation of the European Union and other, non-EU countries, and it provides a forecast for the coming year. It emphasizes the growth slowdown caused by international trade disputes and the uncertainty caused by BREXIT. The report then turns to a brief examination of the extent of the digital and technical transformation of business models in European industries. It argues that Europe should do more to compete with the United States and China in developing digital business platforms that combine data privacy with innovation and entrepreneurship. In the next three chapters, the report turns to tax policy challenges, beginning with the challenges of corporate taxation. After reviewing general trends in the evolution of corporate tax rates and revenues, the report summarizes the two main challenges that globalization and digitalization pose to the taxation of multinational businesses. The first challenge is tax competition that exploits the mobility of both real economic activity and the location of profits. The second challenge is that multinationals find it increasingly lucrative to engage in tax planning in order to avoid paying their fair share of taxes. This is made possible by the fact that firms with digitalized business models can, with little physical presence, generate large revenues in a country. The report considers various policy options for addressing both 1188 n canadian tax journal / revue fiscale canadienne (2020) 68:4 tax competition and profit shifting. At one extreme are the unilateral actions of individual countries, such as the digital services tax imposed by France and threatened by other European countries. Other options require international coordination. One option is to allocate international profits by using a system of formulaic appor- tionment analogous to the way that profits are allocated among provinces in Canada. This could effectively deter profit shifting, but it requires a high degree of coordination. Alternatively, countries could follow the OECD’s suggestion: reduce profit shifting by distinguishing between (1) routine, or normal, profits and (2) residual profits, or rents. Normal profits would be taxed on the basis of residence, and resid ual profits on the basis of destination. Countries could mitigate tax competition by agreeing to a minimum tax rate. The report emphasizes the importance of international data sharing and transparency. The second area of tax policy considered in this report is the taxation of mobile workers and jobs. The direct taxation of earnings is critical for addressing inequality, but the mobility of labour compromises fair income taxation. The mobility of high-skilled persons constrains the choice of tax rates on higher incomes, and the mobility of low-skilled persons, particularly within Europe, constrains transfers to low-income persons. The report argues that progressivity can be achieved by broadening the tax base and closing loopholes rather than by changing tax rates. At the same time, countries could use targeted tax expenditures to reduce the outward migration of particular types of high-income persons without compromising the progressivity of the tax system as a whole. The report also argues that some public expenditure reforms can enhance the fairness of the public budget as a whole. For example, education subsidies and grants currently benefit mainly those who go on to earn high incomes. These subsidies could be replaced by state-guaranteed loans whose repayment is conditional on income. The final area of tax policy involves the taxation of immobile bases, especially consumption, land, wealth, and inheritances. The immobility of these bases makes them natural targets for efficient revenue raising, but fairness is a concern. Con- sumption taxes are broad-based and relatively hard to avoid, but they are regressive in themselves. Their revenue-raising potential can be exploited without fairness being compromised if they are combined with other measures such as low-income tax credits. Consumption taxes should also be integrated with environmental tax regimes to ensure that pollution pricing treats domestic and foreign production uniformly. Taxes on land or property values are also relatively efficient because of their immobility, but they are not widely used in Europe. Correctly designed and based on market values, such taxes could be relatively fair while raising local revenue and enhancing accountability. Wealth taxes have recently been touted as progressive, but they suffer from avoidance problems. The report considers the option of a double wealth tax, which combines (1) a high rate on immovable assets with (2) a low rate on movable ones. The report favours a tax on inheritances as a way to reduce income and wealth differences across generations. It advocates simplicity in the design of an inheritance tax as a way to prevent tax-planning opportunities. The authors of the report also argue against exemptions, such as the family current tax reading n 1189 business exemption, as inherently arbitrary and unfair. Finally, they argue for designing the inheritance tax in such a way that transfers to younger heirs are encouraged. Generally speaking, the report contains no surprises. Although it is based on the European case, the principles that it identifies could apply to Canada. R.B.

Tarcisio Magalhães and Ludmila Oliveira, “Transnational Tax Law-Making in Brazil” (2020) 48:8/9 Intertax 708-18. The literature on the concrete experiences of countries undergoing tax reform is much more limited than the literature that makes bolder, more general claims about tax reform. This article by Oliveira and Magalhães offers a first glimpse into the process of tax-law making in Brazil as it has unfolded since the end of the Second World War. The story the authors tell—of the various government and private entities that were engaged in giving the Brazilians tax advice—is fascinating. From the Inter- national Monetary Fund (IMF) to the United States, to the Getulio Vargas Founda- tion, expert tax advice from outside Brazil was never in short supply. The result, in the first couple of decades after the Second World War, was a tax system that offered many incentives for the export sector and for industrialization while exacer bating social inequities. Enter the 1970s and a new (and short-lived) era of economic growth. Oliveira and Magalhães document another round of IMF engagement. Then we are on to the 1980s, when the IMF is replaced by the World Bank. But then the IMF returns along with the Getulio Vargas Foundation. It is a story of recurring actors and of, perhaps, less success than might have been expected. The story ends with the arrival of the OECD, and a call to Brazilian experts to bear greater responsibility for the country’s tax future. K.B.