As of 2021, I have been writing scholarly journal articles and blogging here on Parliamentum for ten years. What I find most interesting is seeing how my earlier works have influenced or been used by other scholars, in some cases now a decade after its original publication.
One such example came to my attention earlier this year: Jonathan S. Gould’s “Codifying Constitutional Norms” in the Georgetown Law Journal. Gould cited the piece that Nick MacDonald and I wrote on officialising constitutional conventions in cabinet manuals; we drafted it in 2011, but it did not appear in the Journal of Parliamentary and Political Law until 2012.
- Gould, Jonathan. “Codifying Constitutional Norms.” Georgetown Law Journal 109, no. 4 (March 2021): 703-765.
The Similarities between American Constitutional Norms and Constitutional Conventions in the Commonwealth Realms
An Assistant Professor of Law at the University of California at Berkeley, Gould, like many American lawyers and political scientists, seems to have become interested in the uncodified constitutional norms which complement the United States Constitution and American statute laws and make the American presidential-congressional system work in practice because of Trump’s chaotic presidency. Trump became violating American norms even during the Republican presidential primaries in 2015-2016 by refusing to disclose his tax returns, and his callow disregard for the norms of American politics most infamously culminated in the riot and half-baked insurrection on 6 January 2021 where his supporters stormed the Capitol whilst chanting “Hang Mike Pence!” and brandishing the traitor’s Confederate battleflag. The Capitol Riot which bookended Trump’s presidency also underscored that peaceful transitions of power under the United States Constitution depend in no small way upon the unwritten norms that until recently most Americans – scholars, politicians, and the people alike – took for granted as unassailable and as obvious and unremarkable as the fact that we breathe air. Gould acknowledges from the outset: “Legal scholarship in the United States has, until recently, mostly neglected the prospect of codifying constitutional norms.” Gould’s observation corroborates my memory from ten years ago; while Nick and I did not really delve into much of the American scholarship on codification when we researched our piece in 2011, I also cannot recall having encountered any American efforts in this field either. We cited some American precedents and the Federalist Papers as contrasting examples to what the Commonwealth Realms had done; we did not encounter anything complementary.
American scholars use the term “constitutional norms” instead of “constitutional conventions” as we tend to do in the Commonwealth Realms, perhaps because a “constitutional convention” in the United States usually refers to the gathering in Philadelphia in 1787 during which the Framers replaced the Articles of Confederation with what became the United States Constitution. Americans tend to use “convention” in a more concrete way, referring to a specific event or meeting, like the quadrennial conventions of the Democratic and Republican Parties at which they officially select their respective presidential candidates. In any case, the content of Gould’s articles makes clear that he and other American scholars use the term “constitutional norms” in precisely the same manner and context as we typically say “constitutional conventions.” Even the Bowden-MacDonald and Gould definitions themselves bear some similarity. We said:
“Constitutional conventions are unwritten, politically enforceable norms. These norms evolve from practices and customs that complement and contextualize laws or the written constitution. Norms imply exceptions, and more broadly allow for exemptions. In practical terms, conventions help decision-makers determine how they should act in any given situation.”
Gould defines constitutional norms as “legally unenforceable principles that govern the conduct of public officials, the structure and function of government, and the operation of campaigns and elections.” He adds:
First, constitutional norms share basic characteristics with other types of norms. Norms “emerge from decentralized processes,” rather than through legislation, regulation, or any other formal process. They need not be written down at all. The content of norms “reflect[s] established practice.” If norms are enforced at all, their enforcement is informal. Constitutional norms, in particular, are “enforced by the threat of political sanctions, such as defeat in reelection, retaliation by other political institutions and actors, or the internalized sanctions of conscience.”
Second, constitutional norms are constitutional by virtue of their subject matter. They concern “the ‘machinery of government,’ that is, the relation between the main branches of government, their prerogatives, and the limitations on their powers.” They typically dictate how public officials, whether elected or appointed, ought to behave. To call a norm a constitutional norm distinguishes it from other kinds of norms, such as social norms, cultural norms, or academic norms.
What Gould calls “informal enforcement,” we in the Commonwealth Realms call “political enforceability.” Once more, Gould and other Americans use slightly different terminology but describe the same phenomenon.
Gould also describes the same risks or potential consequences of codification as we did a decade ago: ossification and justiciability. Codifying norms into the constitution or statute ossifies them in amber so that they can no longer evolve organically; they also become legally enforceable before the courts, empowering the judiciary at the expense of the executive and legislature. Gould takes that possibility seriously, noting that “it can be useful for norms to retain a degree of ambiguity, and codification risks freezing in place norms that ought to be able to evolve to meet shifting or unanticipated circumstances.”
On the first score, he even cited the same example as we had. The Twenty-Second Amendment codified into the United States Constitution the precedent that George Washington set in 1796 that presidents shall serve only two terms in office, because a republic of laws and the rule of law depended upon a constrained executive and a presidency that must not degenerate into a de facto dictatorship. Voluntarily relinquishing executive authority preserved republican virtue in a fledgling country. Not until 1940 did any president successfully break with convention when Franklin D. Roosevelt sought and won an unprecedented third term; he also won a fourth term in 1944. The consecutive exigencies of the Great Depression and Second World War created the exceptional circumstances necessary to break Washington’s precedent, but Roosevelt’s successful election to four terms nevertheless opened up the frightening possibility of a dictatorship and a presidency-for-life.
On the second count, Gould warned that codifying norms presents “a set of practical and physical hurdles”and that “codifying constitutional norms can enhance judicial power.” He argues that Congress should “codify norms through soft law or intra-branch law rather than through ordinary statutes.” What Gould terms “soft law” corresponds very closely to what Nick and I dubbed in 2012 “officialisation” of constitutional conventions through cabinet manuals as opposed to codification of constitutional conventions through statute or constitutional amendments. Gould largely draws the same conclusions and proposes the same arrangement that we did ten years ago. We all recognise the perils of converting politically enforceable constitutional norms into legally enforceable constitutional or statutory provisions – codification, in other words. Nick and I characterised cabinet manuals, official interpretations of constitutional conventions, as “officialisation”, something both authoritative in its own way yet also less drastic than codification. Officialisation retains the political enforceability of constitutional conventions and keeps them out of the grasping, prying hands of lawyers and judges.
In the American context, “soft law” (officialising norms) could take the form of joint congressional resolutions. Gould argues that soft law “would formalize norms without making them judicially enforceable” and give Congress the initiative in clarifying “the contours of an otherwise fuzzy norm.” But as in the Commonwealth Realms, this process of officialising constitutional norms into soft law flows from political considerations and does not represent a disinterested and impartial academic exercise. If Congress officialises norms in joint resolutions, it can then control the political debate around what these norms mean and what behaviour on the part of the president breaches those norms. Furthermore, officialising norms bypasses the numerous veto points within America’s strict separation of powers – one house of Congress against the other, the president against both houses, and the possibility that the courts could strike down some statutes as unconstitutional because they infringe upon the president’s authority under Article II. These same considerations applied and motivated the ministries and the civil services in Canada, the United Kingdom, New Zealand, and Australia when they variously officialised constitutional conventions through cabinet manuals and guidelines on the caretaker convention and the like. However, perhaps Gould could have acknowledged that the House of Representatives passed resolutions more than once in the 19th century that a president should serve only one term, and that subsequent presidents ignored these politically motivated declarations.
Gould distinguishes between “soft law”, a form of officialisation by which Congress could impose its interpretation of constitutional norms upon the president in an attempt to define the terms of any political debate around these norms, versus “intrabranch law”, another form of officialisation by which the executive, legislature, and judiciary set their internal rules of conduct and, in effect, police themselves. This distinction arises in the United States because of the strict separation of powers. I would argue that cabinet manuals in the Commonwealth Realms straddle these two American categories in some respects, because they act both as internal guidelines for how an incumbent ministry behaves but also define the terms of debate amongst politicians, journalists, and the general public when controversies arise over the constitutional advice that prime ministers offer to Her Majesty the Queen or Their Excellencies the Governors General.
Codification means either passing a constitutional amendment or a statute that converts the substance of the norm into positive law. Gould identifies the Twenty-Second Amendment of 1951, which formally limited a president to two terms, and a series of statutes that Congress enacted in response to Nixon’s misdeeds between 1974 and 1978 – namely, the Congressional Budget and Impoundment Control Act of 1974, the Presidential Recordings and Materials Preservation Act of 1974, and the Presidential Records Act of 1978 – as “paradigmatic codified constitutional norms.” In addition, the Faithless Elector Laws that most states have enacted codify the norm that the electors within a state’s Electoral College must cast their votes for the presidential candidate who won the plurality or majority of the popular vote in that state instead of casting their ballots for their individual and personal preference.
In Canada, Peter H. Russell and his acolytes led the charge to officialise constitutional conventions in cabinet manauls for what they saw as the abuses and misdeeds of Conservative Prime Minister Stephen Harper, especially between 2008 and 2011, with respect to the fixed-date election law, prorogation, the confidence convention, and the process of forming governments. It matters not that most of their arguments flowed from fundamental misinterpretations or ignorance of the relevant precedents, as well as the nature of Responsible Government itself; they nevertheless succeeded in capturing Canadian political science in the twenty-tens before abandoning altogether once Justin Trudeau led the Liberals to a parliamentary majority in 2015, even though Trudeau’s prorogation of 2020 mimics Harper’s of 2009 almost perfectly, and even though Trudeau in 2021 obtained a snap election despite not first having lost the confidence of the House of Commons just as Harper had done in 2008. In addition, their arguments depended upon ignoring the Manual of Official Procedure of the Government of Canada, which the Privy Council Office devised in 1968 and which internal memos which I obtained under Access to Information show that it still used as late as 2008, given that this early cabinet manual expressly supported Harper’s controversial actions. They sought to impose their own interpretation of constitutional conventions by proxy through cabinet manuals but failed in that initiative. In any case, the facts of the precedents would not have supported their claims.
In contrast, Trump undoubtedly flouted and exposed the weakness of some constitutional norms in the United States. In both countries, perceived or real instances of flouting constitutional norms prompted calls for codifying these norms in order to prevent future prime ministers or presidents from imitating their predecessors. This impulse amounts to a retroactive legal enforcement. Gould emphasises throughout his article that successful codification almost always takes place shortly after the president who breached the old norms has left office, which explains why Congress and the states passed the Twenty-Second Amendment a few years after Roosevelt died in office during his unprecedented fourth term and why Congress passed and forced Ford and Carter to sign into law several statutes limiting the authority of the presidency after Nixon resigned in disgrace to avoid impeachment in 1974.
Gould cites examples of some other norms within the American presidential-congressional system that Congress could codify today. Norms, not constitutional or statutory provisions, dictate that the President should not interfere with the Department of Justice’s investigations, that the Speaker of the United States House of Representatives invite the President to enter the chamber and deliver the State of the Union, and that the Supreme Court consists of nine justices. (I suspect that this custom that the Speaker of the House must invite the President into the House of Representatives flows from the Commonwealth tradition that the Sovereign or Governor General cannot enter the House of Commons, or House of Representatives in Australia and New Zealand). Gould did not acknowledge that the radical elements of the Democratic Party openly advocate enacting a law to increase the number of justices on the Supreme Court to overcome its conservative majority. Trump’s flagrant breaches of several norms have prompted Democrats in Congress to propose legislation that would codify into statute that “the President not interfere with federal investigations, […] that presidential candidates and sitting Presidents release their income tax returns, [… and] that Presidents not use the pardon power to place themselves, their family members, or their close associates above the law.” The debates over the Twenty-Second Amendment – a thinly veiled rebuke of Franklin Roosevelt – and the statutes enacted in the mid-1970s in response to Nixon and Watergate show that codifying norms as a kind of retroactive enforcement against one president in particular usually do not happen until the offending president has left office (whether by death, defeat, retirement, or resignation) and, furthermore, only succeed when the party opposite commands majorities in the House of Representatives and Senate. These latest proposals to codify the norms that Trump broke or ignored will fail until the Democrats gain large enough majorities in Congress and an audience before either a Democratic president or a Republican president amendable to repudiating his predecessor.
The Dialectical Nature of Constitutional Conventions or Constitutional Norms
Gould highlights the difficulty in codifying norms by pointing out that scholars and politicians sometimes cannot agree on precisely what a norm entails, or sometimes whether a norm even exists at all. This, too, holds lessons for the Commonwealth Realms like Canada. I recently completed a manuscript which I hope will appear in 2022 addressing this very question. Constitutional conventions are, by definition, dialectical and not unilateral or declarative, which means that they always remain subject to open debate and interpretation – not to scholarly fiat, and not even to cabinet manuals. This idea has probably lurked in my mind for the better part of a decade, and I have to thank Gould, because reading his article finally fettered it out into the open and allowed me to articulate what I have long held.
In the Commonwealth Realms, we often refer to the test devised by British constitutional historian Sir Ivor Jennings to ascertain whether a practice or custom qualifies as a binding constitutional convention. It consists of three questions: what are the precedents? Do the politicians consider themselves bound by a rule? Is there a reason for the rule? The very fact that the Jennings Test relies on questions in and of itself highlights the dialectical nature of constitutional conventions: we are supposed to define the scope of conventions through debate and dialogue.
In Canada, every controversy where a Governor General has dismissed a Prime Minister – Head dismissing Brown in 1858, Aberdeen dismissing Tupper in 1896, Byng dismissing King in 1926 – or where a Prime Minister has used prorogation or dissolution tactically (like Chretien’s early elections in 1997 and 2000, Chretien’s prorogation in 2003, Harper’s early election in 2008 and prorogations in 2008 and 2009, and Trudeau’s prorogation in 2020 and early election in 2021) all generated controversy and debate at the time that they occurred. Scholars likewise continued to debate some of these precedents for decades afterwards. But many Canadian scholars do not seem to accept the dialectical nature of constitutional conventions and instead insist upon some kind of bizarre unilateral and declaratory method of imposing their own interpretations and foreclosing the possibility that we could even debate the meaning and scope of conventions in the first place. For instance, this passage from Aucoin, Jarvis, and Turnbull in 2011 probably first prompted this entire thought and always struck me as absurd and wrong in every respect. They lamented that “we have seemed content to allow the confusion and disagreement of the King-Byng Affair to fester for more than 85 years.” What they dismiss as confusing multi-decadal dispute festering like a gangrenous sore in fact constitutes a lively and healthy dialectic amongst Canadian politicians, historians, lawyers, and political scientists over the last century over a crucial precedent. Precedents where governors have dismissed prime ministers usually provoke controversy both at the time amongst politicians themselves and afterwards amongst historians.
Gould’s work shows that the same kinds of debates have happened in the United States. Gould throughout also describes constitutional norms as dialectical – even though he did employ that exact description. However, he provided numerous examples where codifying a norm into law generated political controversy and debate amongst politicians and scholars alike. The very fact that this happened shows that constitutional norms in the United States are just as dialectical as constitutional conventions are here in Canada and the other Realms. Gould outlines the fraught process of identifying, defining, and codifying constitutional norms in the United States:
Even when a norm can be identified, would-be codifiers might not be able to agree on its precise content. “Social norms are hard to describe; they are fuzzy; they drift.” The same holds for constitutional norms, which are “beset with problems of defining their true content.” A norm’s precise scope will often be debatable, subject to different interpretations as different actors read the relevant history differently.
Gould then delves into the paradigmatic precedent of codifying Washington’s Two-Term Rule into the United States Constitution for all time.
Norms surrounding presidential terms are illustrative. It might seem that there was a norm from the birth of the republic: no President may serve more than two terms. But a closer look reveals considerable ambiguity. President Washington did not view his decision to forgo running for a third term as setting a precedent or as resulting from normative obligation. Prior to President Franklin Roosevelt running for and winning a third term, several presidents either considered running for a third term or ran for a third term unsuccessfully. There was disagreement about whether partial terms or nonconsecutive terms should count for purposes of the two-term norm. Although many believed that the norm existed so long as voters would electorally punish any President who ran for a third term, others denied that the norm existed in the first instance.
The two-term norm remained uncodified for well over a century, which allowed questions about its scope, application, and even its existence to linger in the background. Codification forced these issues to the foreground. The congressional debate over the Twenty-Second Amendment featured discussion of how to treat both partial and nonconsecutive terms. That debate shows how codification can force public officials to confront details and edge cases, even if there is broad agreement about the norm’s core application.
Washington’s Two-Term Rule and the subsequent debates over ratifying the Twenty-Second Amendment therefore perfectly illustrate the dialectical nature of constitutional norms: some regarded the norm as absolute, others as ambiguous and flexible (especially if one president inherited the term of another) – and others still denied that it existed at all. Those who believed that the norm existed also held the corollary viewpoint that voters would punish any president who sought a third term and thereby demonstrate the political enforceability (or what Gould calls “informal enforcement”) of this norm. But Franklin Roosevelt proved that hypothesis wrong in 1940 when he won a third term. Gould notes that some American scholars argued in the late 19th and early 20th centuries that Washington’s Rule did not really exist.
Other American scholars have considered this question and therefore, indirectly, the dialectical nature of constitutional norms in the United States. Bruce Peabody and Scott Gant wrote a treatise in 1999 on the ambiguities contained within the Twenty-Second Amendment and the debates between 1796 and 1940 over whether Washington’s Two-Term Norm existed and, if so, under what circumstances it applied. If we applied the Jennings Test to Washington’s Two-Term Rule (prior to 1951 or 1940), it probably does not meet the threshold of a constitutional norm. One could easily articulate a reason for why this rule should exist, namely to prevent a republic from descending into a dictatorship and imperial presidency. But the precedents themselves offer a decidedly mixed view. However, the other two questions – identifying the precedents and whether presidents considered themselves bound by the rule – reveal a decided mixed record.
First of all, only the presidents who finished two terms could have even considered the possibility of running for a third; this would limit the sample up until 1940 to George Washington himself, Thomas Jefferson, James Madison, James Monroe, Andrew Jackson, Ulysses Grant, Grover Cleveland (with his two non-consecutive terms), Teddy Roosevelt, and Woodrow Wilson (though the strokes which struck him dumb in his second term effectively made this question moot) – only 9 out of the 30 men who had served as president up to the point where Franklin Roosevelt won this third term. And of these men, only a few of them considered themselves bound by Washington’s Two-Term Rule.
Curiously, as Peabody and Gant show, Washington’s personal correspondence suggests that he did not consider himself bound by the rule that would later bear his name. They believe that he simply wanted to retire to his estate instead of consciously desiring to become the American Cincinnatus. That hypocritical sphinx Jefferson could not seem to decide whether he considered himself bound by the rule or not. On the one hand, he hedged against running for a third term to Vermont’s legislature in 1807, declaring (ambiguously) that he “should unwillingly be the person who should furnish the first example of prolongation beyond the second term of office”; on the other hand, he had mused ironically in 1805 about running again in 1808 in order to prevent a “monarchist” (i.e., a Federalist or Whig) from winning that year. You can imagine my shock to discover that this crypto-revolutionary who could brazenly wax poetic about all men being created equal whilst himself owning slaves also contradicted himself on this less morally vexing question. Jefferson did, however, in the end limit himself to two terms. Madison and Monroe also likewise both restricted themselves to two terms each, and the Virginia Dynasty ended with the latter. Andrew Johnson wrote a special message to Congress advocating that presidents should serve only one term; the radical Republicans in Congress apparently agreed with him and introduced their own resolutions to this effect.
Grant nearly challenged this norm in 1876. However, Republican conventions (that’s how Americans use the word) in several states publicly declared their opposition to re-nominating Grant for a third term. Grant wrote back to Pennsylvania’s resolution and insisted that he had not considered himself a candidate. However, like Jefferson, Grant left open the possibility that he might seek a third term “under such circumstances as to make it an imperative duty – circumstances not likely to arise.” The House of Representatives passed a resolution in December 1875 describing Washington’s Norm as a “time-honored custom” and admonishing anyone who would attempt to breach as it as “unwise, unpatriotic, and fraught with peril to our free institutions.” Yet Grant did not consider himself strictly bound in principle to adhere to Washington’s precedent. While he did not seek re-election in 1876, he contested the Republican nomination once more in 1880 “and led all other candidates through 35 ballots” before his opponents banded behind James A. Garfield (and perhaps in so doing saved Grant from the assassin’s bullet which felled Garfield). Here Grant exposed the limits to Gould’s approach of officialising norms in soft law because he ignored a congressional resolution and sought a third term anyway.
Teddy Roosevelt’s case also contains a crucial ambiguity alluded to above: he escaped the irrelevant confines of the vice-presidency in September 1901 after an anarchist assassin’s bullet slew William McKinley, but Roosevelt never ran for president in November 1900, which left open the possibility that only his term from March 1905 to March 1909 counted as a “first” term. In 1904, T. Roosevelt rejected that Jesuitical casuistry and declared that he regarded his tenure through succession between 1901 and 1905 as his first term; he pledged, “Under no circumstances will I be a candidate for or accept another nomination.” While Roosevelt later grew to regret that public declaration not to run for a “third” term in 1908, neither his personal pledge nor Washington’s Norm stopped him from attempting to win a non-consecutive term in 1912. Roosevelt attempted to wrest the Republican nomination from Taft in 1912 by dissembling his potential re-election that year as his true second term. The Republicans stuck with Taft, and Roosevelt ran as the leader of the Progressive Party instead, splitting the vote and getting Wilson elected.
I would argue that the Two-Term Norm meets only one of the three questions in the Jennings Test; while we could easily articulate a reason why the norm should exist, the precedents show, if anything, that it did not, in fact, exist because Grant and T. Roosevelt both actively sought to win non-consecutive third terms, and other presidents like Washington and Jefferson regarded the two-term norm as contingent upon the absence of exigent circumstances. The Jennings Test suggests that this two-term norm never really existed except as part of the mythos surrounding George Washington as the American Cinninatus, a hero steeped in republican virtue who carried out his duty to defend his country and then voluntarily retired to his farm, and therefore part of the mythos of the United States of America itself.
- Insurrectionists Have Betrayed the Revolution of 1800 (January 2021)
- A Caretaker Convention in the United States? (February 2016)
- Alexander Hamilton Recommended Responsible Government in 1788 (November 2012)
- Put It in Writing: Making Constitutional Conventions Official in Canada, the United Kingdom, New Zealand, and Australia (September 2012)
 Jonathan Gould, “Codifying Constitutional Norms,” Georgetown Law Review 109, no. 4 (March 2021): 708.
 James W.J. Bowden & Nicholas A. MacDonald, “Writing the Unwritten: the Officialisation of Constitutional Conventions in Canada, the United Kingdom, New Zealand, and Australia” Journal of Parliamentary and Political Law 6, no. 2 (August 2012): 367.
 Gould, “Codifying Constitutional Norms,” 711.
 Gould, “Codifying Constitutional Norms,” 712.
 Gould, “Codifying Constitutional Norms,” 708.
 Bowden & MacDonald, “Writing the Unwritten,” 400; Gould, “Codifying Constitutional Norms,” 706.
 Gould, “Codifying Constitutional Norms,” 707.
 Gould, “Codifying Constitutional Norms,” 711.
 Gould, “Codifying Constitutional Norms,” 711.
 Gould, “Codifying Constitutional Norms,” 749.
 Gould, “Codifying Constitutional Norms,” 750.
 Gould, “Codifying Constitutional Norms,” 750-751.
 Peabody & Grant, “The Twice and Future President,” 579-580. Congress expressed his dissatisfaction in Martin Van Buren by resolving that no president should serve more than one term. In any event, Van Buren did only serve one term after losing re-election. Congress also resolved during Andrew Johnson’s accidental presidency that presidents should only remain eligible for one term.
 Gould, “Codifying Constitutional Norms,” 751.
 Gould, “Codifying Constitutional Norms,” 718-720.
 Gould, “Codifying Constitutional Norms,” 729-730.
 Gould, “Codifying Constitutional Norms,” 715-716.
 Gould, “Codifying Constitutional Norms,” 722.
 Peter Aucoin, Mark Javis, and Lori Turnbull, Democratizing the Constitution: Reforming Responsible Government (Toronto: Emond-Montgomery Publishing, 2011), x.
 Gould, “Codifying Constitutional Norms,” 724.
 Gould, “Codifying Constitutional Norms,” 724-725.
 Bruce G. Peabody and Scott E. Gant, “The Twice and Future President: Constitutional Interstices and the Twenty-Second Amendment,” Minnesota Law Review 83 (1999): 577.
 Peabody & Grant, “The Twice and Future President,” 578-579.
 Gould, “Codifying Constitutional Norms,” 724.
 Peabody & Grant, “The Twice and Future President,” 579.
 Peabody & Grant, “The Twice and Future President,” 580.
 Peabody & Grant, “The Twice and Future President,” 581.
 Peabody & Grant, “The Twice and Future President,” 581.
 Peabody & Grant, “The Twice and Future President,” 583.
 Peabody & Grant, “The Twice and Future President,” 583.