Saturday, January 28, 2012

Why the Canadian Senate will never be Reformed

For almost as long as the Canadian Senate has been around, there have been calls for its reform. Minor adjustments have been made. For example, Senators used to be appointed for life, whereas now, they must retire at the age of 75. Considering that Senators can be appointed as young as 30, while life expectancy is around 80, the maximum possible tenure as is reduced from an expected 50 to an exact 45. So a typical Senator will have to sit through one less House of Commons election. A lot can happen in 5 years, of course, but it's fairly small in comparison to a potential 45 or 50 year career.

Another change happened in 1982, when the Senate was given the right to veto constitutional amendments under certain circumstances. But there is little appetite in Canada for constitutional amendments in the first place, so it is unlikely that they will ever be able to use this power if even they feel the need to exercise it. The practical effect of this privilege is close to nothing.

Though these reforms have been made, their effect is small, and besides, they are not the reforms that need to be made, and the changes that do need to be made will probably require constitutional amendment. This is where the problem lies.

There are many who feel that the Senate should be elected rather than appointed (I'm ambivalent, but lean more toward elected for the simple reason that the potential advantages to an appointed Senate rarely seem to be realized). The legitimacy of an elected Senate, however, is severely undermined by the current seat distribution.

The number of Senate seats each province gets is the sort of thing that we might learn about in a grade school or high school history class, or perhaps on a class trip to the Parliament Buildings in Ottawa. I think my grade 10 history teacher tried to teach me about it. He was something of a bore, though, and his efforts were for naught. Those who did pay attention to their history teachers probably put in just enough effort to be able report the facts on a test and never gave it much thought after that. In that light, I probably shouldn't blame my teacher. Though he was truly a bore, even the most engaging teacher would have a difficult time making this topic interesting to most people (indeed, I'm wondering whether I should continue writing at this point, or just assume that you've all stopped reading out of boredom).

At some point, I can't recall when or why, I became interested in Senate on my own. After examining the numbers, I realized just how out of whack the distribution of seats is. One would expect that if Province A has more Senate seats than Province B, then Province A also has a larger population than Province B. There are many ways in which this fails to be true with the current distribution.

For example, every western province, from Manitoba to British Columbia has 6 Senate seats and all have populations of greater than 1 million. Compare this with Nova Scotia and New Brunswick, both of which have populations of less than 1 million and 10 Senators each. All 4 western provinces have a larger population but fewer Senators than these two eastern provinces.

We can also look at the way the seats are distributed across regions, rather than provinces. For the purposes of distributing Senate seats, the provinces are divided into four regions, Atlantic Canada, Quebec, Ontario, and Western Canada. Originally, Senate seats were distributed equally to these four regions, 24 for each. At that time, though, Newfoundland was not yet part of Canada, and thus did not have any of the 24 seats for the Atlantic region.

When Newfoundland did join confederation, 6 seats were added to the Senate for the province, rather than taking seats from the other 3 Atlantic provinces and giving them to Newfoundland. This results in a total of 30 seats for Atlantic Canada, compared to the 24 for any of the other three regions. But the total population of Atlantic provinces is smaller than the population of any of the other regions (in fact, the population is less then half that of any other region). So the region with by far the lowest population has the highest representation in the Senate.

In fact, the total population of the 4 provinces in the region is less than the populations of four out of the other six provinces. Atlantic Canada has a population of around 2.35 million, representing 6.75% of Canada's population, while holding around 28.5% of the seats in the Senate (30 out of a total of 105). The extent of the imbalance should be obvious. If it's not, it should become so when you consider the fact that Atlantic Canada has a smaller population than either of Alberta and BC, and yet has 5 times as many Senators as each do.

It's not unreasonable to over-represent provinces or regions with lower populations. Indeed, this may be considered a feature. The number of House of Commons seats allotted to each province is more reflective of the population (though the distribution has some significant flaws of its own). Since the region has such a small population, it is easy to ignore the interests of this region for the sake of other regions. Over-representing smaller provinces in the Senate makes it harder to do this.

With smaller provinces having more seats than some larger provinces and one region having 5 times as many seats as provinces with larger provinces, though, the over-representation of smaller provinces has become absurd. Because of this, a fully elected Senate with the current seat distributions would arguably be less democratic than the current appointed Senate.

Thus, in order for the Senate to be meaningfully reformed, providing a mechanism for electing Senators is not enough, and might even worsen matters. The seat distribution must also be changed. And this is where we run into problems. While some changes are possible, the seat distribution cannot be changed without changes to the constitution.

There is little appetite for this, but suppose it could be done. One way to address the imbalance is to add seats. The Senate is already large enough, however. At 105 seats, it is larger than the US Senate, despite the fact that Canada's population is only about a tenth of the US's. Adding more seats is unnecessary and should be out of the question. This means that certain provinces are would lose seats.

There are a few different options for adjusting the distribution. The two simplest are representation-by-population (or rep-by-pop, for short) and equal-representation. In rep-by-pop, the seats are assigned so that the percentage of seats that a province gets is roughly equal to that province's percentage of the total population. For equal representation, every province gets the same number of seats. Typical proposals for the Canadian Senate give 10 seats to each province (and 1 for each of the three territories, for a total of 103 seats, down two from the current 105). There are other possibilities that represent a sort of middle ground between rep-by-pop and equal, but they are harder to explain (and thus, are more likely to be rejected, since fewer people will understand them).

Regardless of the proposal, though, in order for it to pass, it needs to have the approval of 7 provinces representing 50% or more of the population. Although this might seem like a reasonable condition, the population distribution makes this a big hurdle, especially in the case of Senate reform.

Here is why.

If rep-by-pop is chosen, six of the provinces stand to lose seats. Nova Scotia and New Brunswick would go from 10 to 3 and 2, losing 7 and 8 seats respectively. Saskatchewan, Manitoba, Newfoundland and Labrador, and Prince Edward Island would all lose 3 or 4 seats. Since Senate reform is more popular in Western Canada, it might be possible to convince Saskatchewan and Manitoba, despite the fact that they would lose seats. If they were on board, though, the four Atlantic provinces would likely object, meaning that at most 6 provinces would support the proposal, not enough for it to pass. (Ontario would also get 39 seats, up 15 from the current count. This massive increase in power for Ontario might make other provinces more reluctant to support the proposal).

If equal representation is chosen, then Ontario and Quebec both lose 14 seats, western provinces gain 4, as does Newfoundland and Labrador, while PEI gains 6. Nova Scotia's and New Brunswick's allotments are unaffected. Since 6 provinces gain seats, and 2 lose none, it might be easier to gain the support of these 8 provinces, but Quebec and Ontario would surely object. Ontario's population is the largest, representing nearly 40% of the population, but they only have 22% of the seats in the current distribution. It is unlikely they would be willing to let their share of the vote shrink by more than half to less than 10% (only a little more than one quarter of their population share). Quebec, due to its unique linguistic situation within Canada, is not likely to accept a share of seats in either the House of Commons or the Senate that is significantly lower than their share of population (around 24%). But Quebec and Ontario make up 62% of the population. Thus, although it might be possible to get the required 7 provinces on board, those 7 provinces will probably not represent than 50% of the population.

The above paragraphs take the negative view that, rather than act in a principled manner, the provinces would defend an obviously broken system out of self-interest. Perhaps I could be accused of not having enough faith in our elected officials. Perhaps the accuser would be right.

I also think I am right. Although the seat distribution in the House of Commons is better than in the Senate, it is not without its problems. Due to a number of rules that determine how many seats each province gets, Ontario, Alberta, and British Columbia, the fastest growing provinces, are severely underrepresented in the House.

This underrepresentation has been a problem for a while, and it's been growing. Legislation has been tabled that will rectify the situation, but when the Prime Minister announced that the three provinces would receive more seats, the smaller provinces grumbled about their loss of influence, despite the fact that their seat counts remain the same, and even under the new distribution, they will still all be over-represented in the House of Commons. Quebec also objected, and some were proposing that Quebec have a permanent 24-25% of the seats, regardless of their population share (their objections are somewhat justifiable, but the proposed solution is not). A few years ago, Ontario's premier criticised the Senate because he thought Ontario has too few seats (thus suggesting he might not understand the purpose of the Senate, but that's a matter for a different day). If he thinks Ontario should have more seats than currently, it's unlikely that he would accept a proposal that takes seats away from his province. Whenever a loss is threatened, even if it's just a relative loss, no matter how justifiable, there is opposition. Based on these examples, then, I would find it hard to accept that premiers would take a reasonable approach to Senate reform if their seat numbers could be affected.

As I wrote above, there are other ways to decide how to distribute seats among the 10 provinces and 3 territories, but they are not as natural as representation by population, nor as simple as equal representation. But the justification might not be easily understood or accepted, so I wouldn't expect them to gain much traction. Thus, it is likely a choice between the two I described. No matter which proposal is chosen, though, it is likely to fail, since either there will not be 7 or more provinces to support it or the seven or more provinces that do support it will not make up 50% or more of proportion. Therefore, meaningful Senate reform is likely impossible.