When the Senate returned the government’s energy legislation (Bill C-69) to the House of Commons last spring, media attention focused on the unprecedented number of amendments proposed by the Upper House. The only problem: nobody quite agreed on the number.
According to the CBC, “The Senate passed an unprecedented 188 amendments.”[1] For its part, The Hill Times reported there were “nearly 100 amendments from the Senate.”[2] Meanwhile, the National Post stated that, “The Senate passed more than 229 amendments.”[3] They can’t all be right… or can they?
The reality that takes many by surprise is that there is no one way to count amendments. Indeed, “amendment” is an ambiguous term. In the parliamentary world, amendments are made by motion. A motion in amendment in its simplest form is a parliamentarian advancing the proposition “I move that Bill X be amended….” A single motion in amendment, however, might have multiple elements: that is, “I move that Bill X be amended by changing the title and by adding a preamble….” To that end, when speaking of “an amendment,” is one speaking about a motion or the individual elements of a motion?
Even if one has clarity about whether “amendments” refers to the number of motions moved or the number of individual elements in motions, the actual metric of “amendments” made to a bill is unlikely to yield useful information.
To illustrate some of the challenges with counting amendments, let’s take a hypothetical bill establishing a tax credit for which group A is eligible. While being considered before a committee of Parliament, suppose a parliamentarian moves to add group B and another parliamentarian moves to add group C – and both these amendments are adopted. If the legislation were to pass in this form, the list would read that groups A, B and C are eligible. While two separate amendments – corresponding to two motions – were moved, some may suggest that only one amendment was actually made because only one thing changed in the bill: the list of eligible groups.
Imagine now that this same hypothetical bill leaves committee with groups A, B, and C eligible, but group C is struck from the bill by an amendment at a later stage of the legislative process. The motion to remove group C is also an amendment, but does it count towards the total? While the bill now reads that groups A and B are eligible – which may look like one amendment has been made since the bill started – three separate motions in amendment brought the bill to its current form.
Suppose now that a further amendment is made to revise terminology: for example, group B was phrased in terms of “Aboriginal peoples” but legislators now believe it should read “Indigenous peoples.” The bill, as amended, still only has two groups eligible (A and B) but four motions in amendment have passed. However, here there’s an additional twist: this vocabulary change is only necessary in English and so the French version of the bill shows no change even though this additional motion in amendment was passed.
Practitioners, scholars, and pundits all have differing ideas about how amendment counting should occur. The truth is that there is no one universally accepted way, and they all can lead to confusion. Indeed, some scholars propose not attempting to assess any number of amendments but instead to measure legislative change in other ways. One of the more creative contributions to the discussion is a 2019 work entitled “Mapping Mutations in Legislation: A Bioinformatics Approach,”[4] which examines how much text variation there is in a bill from one state to the next, drawing parallels to DNA code mutations.
The fact that an amendment is made reveals nothing about the impact or the extent of its associative change. A single amendment to a bill can correct a small typo or attempt to insert an entirely new enactment within a bill. The single addition of “not” in English or “ne pas” in French can change the entire way in which a provision is to function. Indeed, depending on the procedure of the legislature, there may also be multiple ways to effectuate the same legislative change, which could lead to misleading amendment statistics.
For example, if a bill creates a scheme where something is approved in three phases, each lasting 30 days, it could be that one amendment changes all references from “30 days” to “60 days” in the bill (this is possible, for example, in the Senate at Third Reading). Or, this same change could be accomplished through three individual amendments, one for each phase. Amendments are not always as efficiently packaged as possible, and in some cases, it may be politically advantageous for legislators to divide their amendments. For example, parliamentarians of a caucus may seek to slow proceedings by forcing more votes, or they might simply divide amendments to involve more members in a particular debate.
A slight wrinkle can also come from how one considers clause deletions. In the case of a committee of the Senate or House of Commons defeating a clause of a bill, this is not accomplished through a vote on a proposed amendment; rather, the committee does not agree to carry the clause, and the result is an amendment to the bill deleting the clause. Yet, a motion in amendment to delete a clause is the proper way to accomplish this same modification at a later legislative stage in both the Senate and House. Accordingly, if one counts “motions in amendment” one might not capture committee deletions but could capture deletions later in the process. Again though, the number of clauses deleted might not provide a reflection of anything: In Bill C-69 as received by the Senate, clause one was approximately 90 pages of text whereas clause 122 was two lines in English, three in French. Deleting clause 1 would result in a 90-page difference in the bill, a far cry from a few-line change that would have been observed through a deletion of clause 122.
Further counting discrepancies can arise depending which documents are used to count amendments, particularly if they reflect consolidated amendments. That is, suppose an amendment is made to replace line one on a page of a bill and another amendment is made right after it to replace line two on the same page. A committee report might indicate one amendment – a combined amendment replacing lines “one and two.” Such an amendment could also find expression in the message sent by the House that amends a bill when reporting its amendments back to the House in which a bill originated. Someone counting motions in amendment from the transcript of a proceeding may see that two motions in amendment were adopted, but certain documents might suggest only one amendment was made. This is also an issue when sub-amendments are made; should each proposal for legislative change agreed to by the legislature be counted individually as an amendment?
When amendments are proposed by one House to a bill originating in the other, the messages subsequently exchanged between the Houses may provide for additional amendments – but these may also be hard to count. For example, suppose the Senate adopts amendments and presents them to the House for its concurrence. The House might accept some amendments and reject others – or, it might propose further modifications to Senate amendments or propose amendments in consequence to Senate amendments (or even in consequence to its proposed changes to a Senate amendment). Indeed, the counting can get quite tricky at this stage if one seeks to establish a figure for total modifications made by one Chamber or the other in the end legislative result.
In limited cases, it may also be that a bill is modified from its original form but not through the express adoption of a motion in amendment or a deletion of a clause. For example, some Senate committees have adopted a motion that the “Law Clerk and Parliamentary Counsel be authorized to make technical, numerical and typographical changes and adjustments to the amendments adopted by the committee.” Similarly, House of Commons Standing Order 156 permits the House’s Law Clerk to make non-substantive corrections to bills.
Comparing the text of a bill between stages of the legislative process may reveal that changes have been made but not as the direct result of any identifiable amendment proposed during the legislative process. In some – albeit rare – cases, a change might occur without explicit amendment because the correction of an error is required. As explained by the Law Clerk of the Senate when testifying before the Standing Senate Committee on Fisheries in 1999:
The options for correcting a bill are twofold.
The first option is to correct the bill by amendment. If you wish to correct the bill by amendment, you can correct a small matter or you can make major changes.
The second option is to correct the bill through your officers. However, that is called a “correction of a parchment error,” and obviously there is very limited room to make those sorts of corrections. […] A parchment error cannot be corrected if there is any possibility that we are going against the will of Parliament.[5]
Parchment corrections, modifications under Standing Order 156 in the House or made under an authorization of the Law Clerk in the Senate can result in modifications to the text of a bill though they are not truly amendments in the traditional sense and may not necessarily be evident at a particular moment. As such, a number provided for ‘amendments’ at a particular time during the legislative process does not necessarily reflect the number of modifications made to a bill at that time, let alone how many might be in the end resulting legislation. Moreover, it needs to be recalled that a single amendment can affect anything from one word to multiple pages of a bill.
Depending on what information is being sought, it might be appropriate to speak of motions in amendments moved or adopted, per cent change of legislative text between stages, or even the total length of printed amendment text (for example, to compare the length of one legislative message to another in the Journals for a particular House within a given session). However, it must be kept in mind that changes may not always be evident if looking at only one language version of a bill. In addition, because of the possibility of parchment corrections, any assessment prior to the Royal Assent version being available may be incomplete.
In sum, while legislative metrics such as the number of amendments made to a bill may make for eye-catching headlines, they don’t tell the full story of legislative change. Behind any individual amendment count may be a bigger story – perhaps one of politics, policy, or even Parliament more broadly. Indeed, sometimes the most interesting story is the amendment that is never moved because of behind-the-scenes political negotiations.
Effort spent tabulating amendments to a bill may largely be in vain and possibly at the expense of understanding the effects of changes to legislation. Arguably, understanding how legislation – as modified – will operate and impact Canadians is much more important than having any number for describing how many changes were made to the legislation during its journey in Parliament.
[1] https://www.cbc.ca/news/politics/c69-environmental-assessment-senate-1.5171913
[2] https://www.hilltimes.com/2019/06/17/environmental-assessment-bill-c-69-enters-final-stretch-in-senate-red-chamber-divided-on-tanker-ban-bill-amendments/204419
[3] https://nationalpost.com/news/controversial-bills-c-69-and-c-48-to-become-law-just-one-day-after-senate-enforces-arctic-offshore-oil-ban
[4] Ruth M Dixon, Jonathan A Jones, Mapping Mutations in Legislation: A Bioinformatics Approach, Parliamentary Affairs, Volume 72, Issue 1, January 2019, Pages 21–41, https://doi.org/10.1093/pa/gsy006.
[5] https://sencanada.ca/en/Content/Sen/committee/361/fish/21ev-e
