RE-UPDATED: How to spot a “gerrymander” in Canada’s independent redistricting process
February 7th, 2013
[Welcome, National Newswatch readers!]
Sometimes I think the word "gerrymander" is the most over-used and under-understood word in politics. The fact is that Canada has one of the most enviable, non-partisan systems of establishing electoral boundaries in the world — one that was designed and agreed to by every political party of the day.
And during all the back and forth on the Saskatchewan boundaries issue we're about to endure, it will be important to remember that fact.
UPDATE: See addition in the Saskatchewan section regarding the minority report issued to the Saskatchewan boundary commission report, and thanks to a sharp-eyed reader for reminding me to include it.
FURTHER UPDATE: I typed University of Saskatoon instead of University of Saskatchewan at 6 in the morning. Sorry about that.
MORE IMPORTANT UPDATE: I got the Edmonton situation back-to-front the way I wrote it. There used to be 3 all-urban ridings there and 5 rurban ridings. And now the Commission is thinking 7 all-urban ones plus 2 rurban ones plus 2 adjacent ones.
ALSO: It's Michael Starr, not Richard Starr, who I meant.
History of the Electoral Boundaries Readjustment Act (EBRA)
Like so many Canadian innovations, our independent redistricting process owes its birth to the existence of a minority government, but also in this case to the widespread cross-party agreement at the time on the need for change.
While the Manitoba Liberal-Progressive government of D.L. Campbell had already adopted an independent system of provincial boundary setting a decade earlier following on legislation first adopted in Australia, it was then-Prime Minister John Diefenbaker who got the ball rolling federally, when his 1962 motion to support independent boundary commissions every decade was adopted by the Commons late in the Progressive Conservatives’ majority mandate. But Dief never got around to introducing follow-up legislation in his subsequent minority term, and the initiative fell to Liberal PM Lester Pearson to implement together with the NDP’s Tommy Douglas and the now-opposition PCs, after 1963.
It wasn’t just gerrymandering that had bothered Diefenbaker personally as much as what he called "Jimmy-mandering" – the heavy-handed and partisan way in which Diefenbaker felt that then-Saskatchewan Liberal party boss Jimmy Gardiner had exercised his influence over the redrawing of the Saskatchewan federal boundaries in 1947 and 1952. To that point, electoral boundaries were drawn by the government of the day; but the with the 1961 census complete and the notorious acrimony of the Pearson-Diefenbaker era threatening to make any one-sided redistribution impossible to pass through the Commons, a new process seemed to be the only way to ensure the constitutionally-required redistribution could actually take place.
So with the flag flap already largely out of the way, the House of Commons devoted 51 days of at-times stormy parliamentary debate, spanning two sessions of Parliament, where MPs from all four parties sorted through the various options for the new electoral boundary system. They were responding to a Bill tabled by Pearson’s government, but one to which its sponsor Jack Pickersgill said it had an "open mind" to accept any and all amendments. [And to anyone who claims that the House of Commons today is no less acrimonious than in the days of Pearson and Diefenbaker, I defy them to find a single bill on which either of our two last federal governments were prepared to proceed in such an open and cooperative manner.]
Key to the eventual consensus were former Secretary of State Jack Pickersgill, Progressive Conservative M.P. Gordon Churchill and of course the NDP’s estimable House Leader Stanley Knowles who had the most experience with the provincial Manitoba model. The result of their work became the Electoral Boundaries Adjustment Act of 1964 (EBRA), which was adopted unanimously, and whose key principles remain largely intact to this day.
Now, while the parties agreed that electoral boundaries should be set by independent commissions, who would have the last word and final say on them (not even Australia had gone that far, though they later relented after two proposed redistributions were defeated), the Canadian parties did disagree on a number of other elements of the process, as did MPs across party lines from urban versus rural parts of the country.
- Having one national commission or ten provincial commissions – It was eventually agreed that provincial commissions were more likely to grasp local communities of interest than one national commission, though Conservative MPs
RichardMichael Starr and Erik Nielsen wanted a single national commission. While independent provincial commissions can take advantage of local knowledge to better represent communities of interest and local circumstances in their own provinces, the quid pro quo is often inconsistent application of redistricting principles across the country. But that’s the trade-off you have to accept if Central Canada is not to set the boundaries for the whole country, and you want to get the commissions’ work done within the year.
- Who would the commissioners be, and who would appoint them – The Liberals’ bill outlined a four-person commission, with the chair appointed by the provincial chief court justice, one nominee each from the prime minister and opposition leader, and the "representation commissioner" (a position already created in a companion piece of legislation) as the fourth member of every provincial commission. Stanley Knowles didn’t want any partisan appointments at all, and moved a lengthy amendment outlining who ought to be appointed as commissioners (for example, provincial chief electoral officers and university presidents or other academic experts). The Liberals indicated support for Knowles’ amendment, thereby raising suspicions amongst Conservatives about a possible deal between the other two. Eventually, Knowles’ amendment went down to defeat, but the parties agreed that each provincial Chief Justice would choose a member from his or her bench as the chair, and then the Commons Speaker would appoint two other members. In practice the Speaker’s appointments have always followed the spirit of the Knowles amendment, including academic experts on boundaries and redistribution, and public servants with knowledge of the region’s demographics and communities. Boundary commissions are now three-person affairs, the position of "Representation Commissioner" having been eliminated by Parliament in 1979.
- How much variance from the provincial average seat size should be permitted – MPs representing urban seats, along with the NDP caucus, wanted to keep the maximum variance to 20% (the then-all-urban NDP’s opening position was 5% in keeping with the New Zealand model), while MPs representing rural seats, along with the Conservative caucus, wanted to allow as much as 33.3% variance on either side of the provincial quotient. The final bill sawed off the difference at 25%. Later on in the 1980s under the Mulroney Tories, Parliament amended the EBRA to allow for abrogation of the variance in "exceptional circumstances", a provision that has been used to justify keeping Labrador as a separate seat, for example.
- What, if any, other principles should the legislation set out for the commissions to consider – Not only did the 1964 EBRA legislate a maximum variance of ± 25%, it went further than Australia or New Zealand had by mandating boundary commissions to draw ridings that corresponded "as nearly as may be" to the provincial average seat population (the "electoral quotient"). Canadian parliamentarians also agreed on a set of criteria that the independent commissions could consider alongside mathematical equality, including:
- sparsity or density of population, and other geographic considerations
- relative population growth rates (contentious, and so later deleted in 1975), and
- community or diversity of interests of the residents
1986 Mulroney era amendments to the EBRA softened up the phrase "as nearly as may be" to "as close as reasonably possible", and added the "historical background of a particular district" to the list of allowable criteria.
- The desirability of public hearings and public input, and when it should feed into the process – Unlike the Australian system it borrowed from, Canada added the requirement for a round of public hearings after the boundary commissions submitted their first set of proposals. After the 2003 redistribution process, and following the across-the-board panning of the 2003 Saskatchewan boundary commission’s initial proposals, the Chief Electoral Officer recommended that the Act be amended to include an opportunity for input before the commission proposals were drafted. This point becomes important shortly.
- The legitimate role for elected representatives in the process – A role was preserved for Members of Parliament in the process, both to give evidence before the public hearings and later to have the right to file objections to the boundary commission reports. As local representatives, Members of Parliament often have more detailed knowledge about the area, which can be very valuable to the commissions. MPs can also try to influence the commissions to decide in the their own interest, though this can sometimes go horribly wrong, as in the story of Cambridge Liberal M.P. Janko Peric who was adamant that rural Dumfries could not be separated from the city of Cambridge, and then proceeded to lose his seat when Dumfries voted overwhelmingly Conservative in the next election. Once the province’s boundary commission reports to the Speaker, MPs have 30 days to file an objection if at least 10 of them sign on to it. Those objections were originally debated by the Commons as a whole, but since the 1986 amendments to the EBRA now go to a Commons committee instead. The independent boundary commissions have to consider and rule on any MPs’ objection referred to them by the Committee, but do not have to accept it at all when making their final reports.
- Timelines for the commissions’ work – Commissions are obligated to complete all their work within one calendar year. Technology has enabled some aspects of the work to go more quickly, and a post-2003 representation order report by the Chief Electoral Officer contained a number of suggestions for shortening timelines, some of which were enacted by the Harper government’s bill to change the Commons seat count apportionment last year.
- Who should get the final say – This aspect of the 1964 Act has not changed – the independent boundary commissions still get the final say. Australia did not go so far originally, but after failing to get two redistributions through their own house, subsequently followed Canada’s lead. In our case, the boundary commissions submit their final reports to the Chief Electoral Officer, who prepares the "representation order" and sends it to the "governor-in-council" (basically the cabinet), which MUST proclaim it within five days. Then Elections Canada and the political parties have a year to re-organize themselves along the lines of the new boundaries, and after a year any general election will be run on them.
How could you "gerrymander" an independent process?
So, given this very independent process by international standards, why does the term "gerrymander" pop up so quickly in the vernacular? How could you gerrymander a system like this?
Well, if power is defined as "the ability to influence an outcome", then politicians by definition will always look for a way to do just that. The best way to influence an independently-governed process is to do your research, know what’s in your interest, and then marshall your best non-partisan arguments and best non-partisan allies in those arguments, and hope to do so better than your opponent. Some people think a thumb may have been put on the scales in terms of appointments to the boundary commissions as well, but there’s no evidence Speaker Scheer engaged in that to any great extent.
The one thing not to do, though, is to make partisan arguments before the public hearings of a boundary commission, because you will get cut off. Also, you have to make arguments that demonstrate an interest in the impact of your proposed changes on the surrounding area and even province. Presenters who omitted doing so in the Ottawa hearings were called to order quickly, for example.
The way the boundary commissions actually work is to sit down with the electoral geographer they are assigned, take the province’s population, divide it by the number of seats allocated to that province to come up with the "quotient" or target population for every seat, and then start drawing.
Now, because a riding can’t cross a provincial border, and obviously it can’t go into the United States or into one of the oceans, if you were in their shoes you would quickly realize that the work has to make some big decisions, and then start at the corner of a province. So, for example, the Ontario commission considered how to handle the northern Ontario seats, and then knowing that it was going to keep a province-wide quotient, would have started at the border between Vaudreuil, QC and Glengarry-Prescott-Russell, ON. They would have asked if GPR was now too big or too small population-wise, and then deleted a few census tracts or blocks of land that could fit better with Ottawa-Orléans, and/or added a bit of territory from Stormont-Dundas-Glengarry. And then they’d just keep moving east, through Ottawa, eastern Ontario, Kingston, and on through Northumberland county into Durham.
Or the Alberta commission would have taken the extra seats it was given to work with, decided how to allocate them within the province, noticing that Edmonton now had sufficient population for 7.5 seats, but enough surrounding suburbs to make 7 city seats and 2 hybrid urban-suburban-rural seats, where they were previously
5 city seats and 3 hybrid rurbans 3 city seats and 5 hybrid rurbans..
Or the BC commission would have realized that Vancouver city proper had enough population for 6 seats, while the north shore had enough for 2.5 seats, while other parts of the GVRD were due fractional numbers of seats; and then decide to make a riding that crossed the bridge from Burnaby to the eastern part of the north shore that pleases none of the political parties or local municipal interests on either side of the Burrard inlet.
Or the New Brunswick commission would go so far in its initial report to try to avoid another court-case over Miramichi and Acadie-Bathurst, that an extraordinarily under-populated Miramichi would wind up seeing Dieppe hived off from Moncton and stuck together with Fundy Royal instead.
Any big change creates so-called "ripple effects". So a decision in BC to keep Richmond together in two ridings and Delta together in one, has a ripple effect on Surrey to the east. Deciding whether to protect the under-populated but geographically large ridings on the Gaspé in Québec, for another example, could have ripple effects down as far as Lévis and beyond.
Which is how we get to Saskatchewan, and what happened there over the past year.
The case of Saskatchewan and the rurban vs urban boundaries
Speaker Andrew Scheer had more at stake with his appointments to the Saskatchewan Boundary Commission than any other province, since he represents one of the "rurban" seats in Regina himself, Regina—Qu’Appelle. Justice Ronald Mills of Prince Albert was assigned from the provincial bench, while the Scheer appointed Canada’s leading expert on electoral boundaries and redistricting, Professor John C. Courtney of the University of
Saskatoon Saskatchewan, and the president of the Saskatchewan Association of Rural Municipalities (SARM), David Marit.
Once all the boundary commissions were appointed, they had a surprise for us, because they announced a round of pre-consultations before releasing their initial proposals, and people did not have much time after the 2011 election to meet those deadlines. There was more or less take-up in the different provinces, but in Saskatchewan the opponents of the current boundaries had not been snoozing, and they filed numerous academic and other briefs, whose impact was noted in this paragraph from the commission’s initial proposal last August:
"The population shifts of the past decade called into question the continued suitability of the composite urban-rural electoral districts. This issue is central to the Commission's proposal for the province's 14 electoral districts. It was also central to the concerns expressed by a considerable number of Saskatchewanians who contacted the Commission with communications ranging from brief, one-sentence or one-paragraph notes to formal documents presented on behalf of a group or association. These communications almost unanimously voiced opposition to the continued use of hybrid urban-rural districts in Saskatchewan."
Far from the earshot or awareness of Ottawa, the last highlighted sentence set off a prairie firestorm, because of what it meant. To local Conservative Party activists in Saskatchewan it meant that party headquarters had dropped the ball on the pre-submission phase, and that from that point forward they would be fighting a rear-guard action. Fingers were pointed during a behind-closed-doors meeting for over an hour, with Ottawa bearing the brunt of the blame and resentment. Party Operations Director Jenni Byrne is said to have demanded in return that she wanted to see 8,000 submissions in the public hearing phase against the ending of the rurban seat boundaries.
The Conservatives didn’t manage that number of in-person presentations, but when the Saskatchewan commission announced extra public hearing dates for Regina and Saskatoon, the province’s New Democrats and Liberals realized the Conservatives were not going down without a fight, and indeed the Leader-Post’s longstanding political columnist Murray Mandryk declared the Conservatives to have been the "winner" on Twitter, in terms of turnout at the first day of commission hearings in Regina. By the end of the hearings, the pro and con presentations had evened out somewhat, and both sides were being bolstered with petitions and postcards galore. The Prime Minister asserted yesterday that 75% of interventions in the public hearing phase had opposed the rurban boundaries, but CBC Saskatchewan reporter Stefani Langenegger questioned on Twitter after her interview with Justice Mills whether those numbers were a bit of a stretch. They certainly didn’t include any of the pre-submissions from the first round of public input.
The Prime Minister and party robo-calls also claimed that the old rurban boundaries represented the history and traditions of Saskatchewan representation. This is a little ironic, as when the 1966 redistribution (the first conducted under the newly independent EBRA) recommended two rurban seats each in Regina and Saskatoon, then-Conservative leader John Diefenbaker complained that they did not follow the "historical" precedent until 1966 of all-urban Regina and Saskatoon seats, proving that where you stand depends so often on where you sit.
But the NDP historically made a grand miscalculation about its own interests during the last redistribution, one that was being prepared while the party’s own future was under question, and which came into effect just as newly-minted leader Jack Layton was coming onto the scene. Rural Saskatchewan New Democrat activists found an ally in Palliser M.P. Dick Proctor, who would have wound up living just outside a very awkwardly-redrawn version of his riding, and Proctor and former Moose Jaw mayor Don Mitchell led the fight against urban ridings in the 2003 representation order. The provincial section of the party had its own fires to put out, and Wascana Liberal Ralph Goodale’s seat was not affected, so the urban riding proposal had no defenders. and was scrapped by the Commission in favour of keeping essentially the status quo from the 1996 map. Ironically, Don Mitchell won the NDP nomination to try and win his preferred version of Palliser riding in 2008. It didn’t work.
This time around however, the NDP acceded to the opposite consensus emerging from academia and the Saskatchewan Urban Municipalities association, and in spite of Tom Lukiwski’s attempts to re-open a rift between the Palliser and Regina NDPs, the party maintained a common front throughout the entire boundary commission process.
Meantime, media opinion came out against the rurban boundaries early and hard, and has maintained that posture during the remainder of the commission’s work. Last week’s boundary commission report has been endorsed in editorials and columns in both major provincial dailies.
UPDATE: I am reminded by Colby Cosh to point out that the Saskatchewan report included a dissenting opinion by David Marit. Marit's arguments are said to recapitulate a lot of the interventions he made during the hearings, but which did not find support from either Justice Mills nor Professor Courtney. The existence of a minority boundary commission report has been called "unprecedented", but I did uncover a minority report to a New Brunswick federal boundary commission in the past — again dealing with the continued under-population of certain "historic" seats in the province.
Were the out-going rurban ridings "gerrymandered"? You could argue it both ways. They weren’t designed to meet the political interest of just one political party in an unfair process; they were supported by all three parties, in one case of which the party supported them in spite of their best interest. Maybe it was an inept gerrymander on their part to try and influence the outcome. But the commission retained the final say, and if it had decided to proceed with urban boundaries in 2003, no-one could have stopped the commission from doing so.
Is the current Conservative effort to influence the final report through the Committee process at Procedure and House Affairs (led by Parliamentary Secretary and affected Regina-Lumsden-Lake Centre M.P. Tom Lukiwski) and via the push-poll robocalls a "gerrymander"? Well, it is an attempt to influence the outcome, but the Prime Minister also understands that under the Act there is nothing his government can do if the independent Saskatchewan Boundary Commission rules against his Saskatchewan caucus, short of introducing an Act to suspend the entire Redistribution process. But that would mean running the next election without the 30 extra seats in Ontario, Alberta and BC, which is unlikely to be an acceptable trade-off to him.
Conservative Party operations director Jenni Byrne may end up taking the fall, but if so, it would not simply be over the robo-calls, but the perception that party headquarters dropped the ball with the pre-submission phase of the Saskatchewan boundary commission process last fall. Whether she deserves that more than, say, the former provincial Conservative and Sask Party organizer Tom Lukiwski, only the Saskatchewan Conservative caucus knows for sure.
The upshot will probably see at least one retirement from the Conservative backbench in the province, and maybe an incumbent or two running against each other for the nomination. At least one Saskatoon Conservative won’t be back after 2015, and perhaps more. And the Regina-Wascana and Regina-Lewvan seats look like opposition enclaves. Virtually untouched in the entire process has been the seat of Regina-Qu’Appelle, and who holds that seat again? Oh yes … the Speaker of the House of Commons.
I will update this post a bit later with a bibliography and clickable references, but I've stayed up so late writing this that it got light outside somehow.