The Senate of Canada: Coming of a New Age?

Matt Williams of Jesus College, Oxford provides a fascinating overview of the effects of Trudeau’s reforms in the Canadian Senate.

In 2014, Justin Trudeau disbanded the Liberal caucus of Senators in Canada’s upper house. On becoming Prime Minister, in 2015, he appointed independent Senators on recommendations of a non-partisan body. More than half of Senators (58/103 in 2019) are now independents. In this blog, I will assess what observable effects, if any, reform has wrought on the Senate’s representativeness (Pitkin 1967), independent-mindedness (Russell 2001) and “redundancy” (Patterson and Mughan 1999). Statistical analyses of all 16,629 senatorial votes recorded in the 42nd Parliament are presented, along with machine reading data from all 1,611,817 words of enacted legislation. Preliminary evidence suggests that Senators are independent-minded but not transforming legislation, so a new age of Canadian bicameralism is yet to come.

Introduction

In 2014, Justin Trudeau disbanded his party in the Canadian Senate. With the caucus dissolved, incumbent Liberal Senators chose to become independents, or continue as nominal but untethered Liberals. Upon election of the Trudeau Government, an Independent Advisory Board for Senate Appointments was established in 2016. The Board is responsible for receiving applications from Canadians wishing to fill vacant seats, and makes non-binding recommendations to the Prime Minister. By the close of the 42nd Parliament, in October 2019, more than half of Senators (58/103) were independents. There have been no changes to functions and powers, nor to the public accountability of Senators. The Conservative Party opposed these reforms and campaigned in the 2019 election to revert to partisan appointments.

Has a new age come for the Canadian Senate? There is little basis for concluding as much. The positioning of Liberals and Conservatives relative to the reforms – that it improves or deteriorates respectively – are both overstated. Statistical analyses of all 16,629 senatorial votes recorded in the 42nd Parliament are presented, along with machine reading data from all 1,611,817 words of enacted legislation. Preliminary evidence suggests that Senators are to an extent independent-minded, but are not transforming legislation.

A New Age?

What would constitute a new age? This is conceptually slippery. We would need a baseline “old age” relative to which there is demonstrative novelty. Meg Russell (2001) identified four bicameral functions, in which we can fruitfully hunt for novelties – representing interests not accounted for by the lower house, thinking independently of government, acting on that independent-mindedness, and chipping in to parliamentary work. These are abridged here as – representativeness (Pitkin 1967), independent-mindedness, and a capacity to replicate the legislative work of the House (“redundancy” as per Patterson and Mughan 1999).

The Canadian Senate in the preceding 41st Parliament was marginalised by the Harper Government, who latterly refused to fill vacancies. As such, if the 42nd Parliament were compared to its immediate predecessor, there would be evidence for invigoration, but it would be inconclusive. It is nonetheless possible to tackle the question. Tentative comparisons to preceding Senates may be investigated. Senators appointed by different governments can be compared within the 42nd Parliament. And, the legislative output of the Senate can be investigated.

Fundamentally, there have been no changes in powers or functions, just personnel. In the absence of independent accountability for an upper chamber (as per the US Senate), bicameralism depends on coordinative capacity between chambers. Parties are a resource efficient mechanism for this. With a majority of independents, the Senate can be expected to have a reduced capacity to coordinate amongst themselves, and externally with policy partners. This can be expected to generate a transactional rather than transformative approach to Senate business. In that, individual Senators can seek policy goals individually, but face barriers in coordinating for transformational change.

On Representativeness

In demographic terms, the Senate under Trudeau has become more diverse. Where 70% of Harper appointees were male and near 90% were white, the equivalent proportions since 2015 are 43% and 80%. But there is no obvious link to dissolution of the Liberal caucus, nor creation of an independent appointments board. Conservative reform critics seem on safe ground in contending that those appointed under new arrangements could just have likely been appointed under the old rules by Trudeau. Where Conservative criticism is less credible is with regards to the voting behaviour of new Senators, which they contend is as pro-government amongst “independents” as if they were caucusing Liberals.

On Independent-Mindedness

How reliably have different groups of Senators supported the government? The following graph summaries all votes cast in the 42nd Senate. From this we see that members of the Independent Senators Group supported the government 69% of the time. These Senators were more predictably loyal to the government than even residual Liberals.

It appears independent Senators are hardly worthy of the name. However, with correlation analysis the picture obscures. The following figure shows the correlation coefficient estimates of each party in supporting a government whip. A dotted vertical line denotes a coefficient of 0, and 95% confidence intervals are described by horizontal red lines emanating from the point estimates. It is notable that the coefficient of independence for Conservative Senators is three times greater in magnitude (to the negative) than that for the ISG. The Conservatives are more predictably partisan than the ISG, although both groups’ behaviour can be predicted with 95% confidence. Only Non-Affiliated Senators were unpredictable in their voting behaviour.

There are several possible explanations for government support amongst independents. ISG members may be by disposition liberal, even if not officially Liberal. Or, as neophyte legislators, they can be expected to vote with a mandate backed by strong Commons majorities. That the Liberal Senators seem more independent than the Independent Senators is simply because there were only nine such Liberals, they were more experienced, and they were more likely to be absent from votes whipped by the government.

On Redundancy

How much did Senators contribute to legislation? Starting with a simple metric, how many pages of legislation were introduced and primarily legislated for by each chamber? The following figure provides descriptive statistics. House-originated legislation accounted for <1.6m words, where the Senate initiated <0.025% of that amount, at 40,000 words. In terms of legislative spadework, the House is unambiguously dominant.

But what about transforming legislation? An effective metric is to assess changes to the language of legislation wrought by each chamber. This is achieved using sensitive machine reading technologies (Williams various). Linguistic analysis allows us to assess how far qualifying parts of speech (such as adjectives, conjunctions and auxiliary modal verbs) changed per section of legislation. The metric of noun/verb qualifiers per section (NVQPS) gives an indication of change through the parliamentary process. The following table summarises these changes in all House Government Bills with four published bill versions on the LEGISinfo website (>5.6m words analysed at four bill stages).

The table shows that most substantial changes to legislative language were obtained by House Committee (column 1). The Senate (column 3) only contributed on average 1/5th of overall changes to a bill from first reading to Royal Assent. The most marked changes to legislation in the Senate – to C-6 and C-29 – started in 2016 and were finished by 2017. This was before the number of independent Senators had reached a majority in that chamber.

Conclusion

There is evidence that independent Senators are more demographically representative and independent-minded than their Conservative counterparts. The contribution by Senators to parliamentary business is not, however, obviously transformative. The Senate was, like the House of Lords, based on a pre-democratic model of governance. In the transition to mass democracy, it was parties that gave some transformative capacity. Who are masters of the Senate now, if not the electorate nor party executives? Masterless disaffiliates are beholden to their individual morality and interests. But that leaves them without institutional resources to make the most of their individual talents. Without reform to accountability, the delegation chains from Senators to the public is fuzzy, and any veto power they held will have been delegitimised (Tsebelis 2002). There is novelty in the Senate, but to conclude a new age has dawned is as yet unsustainable.