Canadian Parliamentary Review

Current Issue
Canadian Region CPA
Upcoming Issue
Editorial and Stylistic Guidelines

HomeContact UsFrançais

Promises and Policy: A Study of Two Parliaments
Christopher Garner

At the time this article was written Christopher Garner was  a doctoral candidate at Nuffield College, University of Oxford.

Academic studies and public opinion polls show that a high percentage of Canadians believe that politicians have no intention of keeping their promises. A recent recent (unsuccessful) recall campaign in British Columbia made failure to keep electoral promises the basis for a recall petition. This article looks at the experience of two national Parliaments 1974-1979 and 1988-1993; and considers whether governments do keep their promises. If not what are the consequences for the democratic process.

A number of academics have pursued the question of what is at the root of the decline in trust and respect for political institutions.  Neil Nevitte has argued quite elegantly that Canada is experiencing a trend similar to that found in all other post-industrial states: a shift in values leading to a decline in deference.1

At the centre of this decline in deference by the general public are structural changes which have been taking place in society since 1945.  These changes in the nature of work, levels of education, and economic well-being and security in turn affect value shifts among newer generations. The result is a paradox of increasing levels of interest in politics by the public with a concurrent decline in the support for, and identification with, traditional forms of political representation.  Canada’s once deferential public now demands a greater degree of input into how things are run.

Michael M. Atkinson approaches the question of decline in trust and efficacy from a different angle.2  He argues that at the root of Canada’s dissatisfaction are two competing visions of what democracy ought to be.  On the one hand Canadians subscribe to an integrative ideal of democracy, based on the Burkean model of representation.  Here deference is a product of the trust we have in our institutions; institutions that govern for us, not with us.  

On the other hand is the aggregative ideal, wherein leaders and institutions are encouraged to be responsive to popular opinions.  In this theory, politics is a means of satisfying the interests and preferences of citizens, in the same way in which producers are expected to satisfy the demands of consumers.  It is this ideal to which Canadians are increasingly subscribing.

Furthermore, Canada’s institutions were designed along the lines of the ‘integrative’ ideal in the form of a Westminster-style parliament, wherein political parties compete openly for public support in forming a government based on an election manifesto. Upon gaining popular support the winner is expected to carry out their manifesto, mustering all of parliament’s powers to do so.  This is, after all, what is commonly implied by ‘giving a government a mandate’.

The constitutional and theoretical basis of government in Canada, then, is by its very nature integrative, paternalistic and Burkean in theory and practice.

What we are presented with in these explanations is an image of great discord between what we have as our form of representation and that which we want.  If this is true, it only seems logical that dissatisfaction and disdain will follow.

Do governments do what they say they will do? The answer disgruntled Canadians give is, simply put, no they do not. This feeling lies behind the negative characterizations of politicians.  But if the proponents of this view are correct then we have a fatal flaw in our democratic system.  After all, elections are about giving governments a mandate for action, and if they do not achieve this, something is seriously wrong.

A study of this issue noted that:

To the extent that parties discuss policies during campaigns, they typically emphasize quick fixes selected on the advice of professional pollsters and party strategists.  They eschew coherent programs... Canadians’ inability to use elections to choose among policy options is further exacerbated when victorious parties ‘forget’ about their campaign promises or enact policies contrary to them.3

Surely this diagnosis is too severe.  But how can we know if governments act on their promises?  This is especially problematic if, as the study suggests, parties eschew coherent programmatic commitments, thus denying us the specifics from which to compare.

One way to get around this problem is found in the study of issues saliency.  That is, to know what parties and subsequent governments intend to do one should look at the general themes which they present at specific points in time.  The inherent assumption is that parties will only emphasize those issue that it feels are salient for the electorate, and are vote-gaining issues.

This approach has the advantage of looking beyond the specifics stressing instead statements that outline a party’s personality, differentiating it from other parties. Moreover, it assumes that party platforms are tools for directing votes and for representing the party’s particular policy space via general emphasis of issues.  To paraphrase Kim Campbell, ‘elections are no place to be talking about specific policies’.

Two fundamental hypotheses can be set out in order to test the strength of mandates in Canada.  The first can be stated as follows: The salient issues in the party manifesto should be transferred to the Throne Speech when that party takes office.

But as the phrase goes, ‘talk is cheap’.  That is, we can take a look at party manifestos and the subsequent government’s general intentions without actually seeing whether the government acts upon those intentions.  Thus, our second hypothesis is: The salient issues found in the Throne Speech should be transferred to government bills.


It is important to outline briefly the method used for tapping into issue saliency, as it can seem to be a nebulous topic.

First, due to the scope of this article only two parliaments and their preceding elections will be examined: the 30th Parliament (1974-78), which saw a Liberal government under Mr.Trudeau, and the 34th Parliament of Mr. Mulroney’s Progressive Conservatives (1988-92).  Both of these parliaments represent the second majority governments for their respective leaders, and as such we can assume that the leaders have learned the ropes of government and can get their business achieved effectively.

For the sake of clarity seven salient issue domains have been delineated for coding purposes:

  • Foreign Relations: including references to international cooperation, support for the UN, aid to developing countries, and the need to maintain a military.
  • Democracy: which covers such issues as human rights, involvement of citizens in decision making, and a gambit of constitutional issues including reform.
  • Political System: entails issues relating to political ethics, governmental efficiency, and federal-provincial relations (outside of the constitution).
  • Economic issues: from market regulation, infrastructure and technology programmes, to demand-side economic policies.
  • Quality of Life: takes account of those aspects of governance relating to social welfare programmes (often referred to as the ‘sacred trusts’) and educational provisions.
  • Social Fabric: covers issues ranging from cultural diversity and  multiculturalism to law and order.
  • Social Groups: is reserved primarily for ‘favourable mentions’ of specific groups, such as labour unions, agriculturists, and underprivileged minorities (immigrants, handicapped, etc.)

These issues are the basis for coding each party’s election literature and each governments’ throne speech and legislative packages.  The procedure used in coding both election literature and throne speeches was to take every sentence of the document and place it into the relevant domain.  The legislative packages are approached in the same manner, however their place was determined by the subject matter and directive which each bill contained.  Any sentences that did not fall within the gambit of these seven domains was coded as ‘miscellaneous’.

Although there are recognizable limitations to this approach one must recall that the objective here is not to test specifics.  We are looking for trends in issue saliency in order to determine whether governments maintain their respective positions in a given policy space or, have they ‘forgotten their promises’.

Results from Two Parliaments

The test results of the hypotheses are presented in Charts 1 and 2 as percentages of the total manifestos, throne speeches, and legislative packages for each governing party.

Clearly, economic concerns led the way in all areas of Trudeau’s government, followed by quality of life issues.  This is not surprising as this period was plagued by economic uncertainty in the form of rampant inflation and unemployment.  Secondly, ‘social justice’ issues were at the heart of Trudeau’s own philosophy, and that of the Liberal Party, and so are not unsurprising by the strength of their presence.  As the fourth session’s throne speech notes: “social services are essential to closing the gap of rich and poor, and of ensuring a standard of living for those in adversity.”4

The greater emphasis in the ‘social fabric’ domain in both throne speeches and legislation is noteworthy.  In the manifesto little emphasis is placed on issues such as multiculturalism and bilingualism, and appeals to national unity.  However, there is a distinct increase in such issues during the course of Parliament.

The most obvious explanation for this rise is found in the later part of the Liberal’s mandate, wherein a reading of the fourth session’s throne speech is steeped in the rhetoric of ‘crisis’.  This is not surprising given the events:  the introduction of Bill 101 in 1977 by the Parti Québeçois, which followed on the heals of the Air Traffic Controllers’ Dispute and rise of national concerns for the future of Canada’s federation, cumulating in the Pépin-Robarts report of 1978.

This explanation is also applicable to the rise observed in the ‘democracy’ domain.  Here such issues take the form of expressed concern for constitutional reform, and the debate over adopting a Bill of Rights for Canada.

An explanation behind the disparity between manifesto emphasis and that of the throne speeches for the ‘social group’ domain is more difficult to find.  The throne speeches for this period simply do not address any particular groupings.  One possible explanation for this will be addressed below.

Turning to the Mulroney government’s record, outlined in chart 2, we find a similar trend to that of the Trudeau government.  First, we can note the degree of convergence between party manifesto, throne speeches and legislative initiatives in the field of economics.  In fact, the convergence between throne speech and legislation in three other domains –external relations, political system and social fabric– are noteworthy.  The degree to which this convergence takes place would be enough to turn the head of the most die-hard disgruntled Canadian.

Again, the two domains that buck-the-trend are ‘democracy’ and ‘social groups’.  The former domain rises in emphasis over the course of this parliament due to prime minister Mulroney’s commitment to constitutional reforms aimed at “bringing Québec back in to the constitution”.  In fact, it is worth noting that the emphasis placed on this issue more than doubles between the second and third throne speeches.  This increase in emphasis has its roots in the debate and failure of the Meech Lake Accord which led to a plethora of constitutional committees, cumulating in the Charlottetown Accord and referendum in 1992.

The social group domain requires a different explanation for the noted divergence of emphasis.  The explanation is two-fold: first, as the rationale behind issue saliency goes, parties are adverse to specific commitments.  To single-out certain social groups during elections would be incongruent with the idea of presenting vague, thematic, broad-based appeals to the electorate.  Rather, parties are more likely to emphasize issues which will attract those social groups, without actually having to commit to policies that are directed at them.  For example, a party’s commitment to social programmes and income support would attract such social groups as the working poor, and women.  Whereas an emphasis on tarrif protection on agricultural produce would appeal to farmers.

Having emphasized these issues, the party-in-government is then open to act in a manner which benefits these groups, through the introduction of policies directly affecting them.

For the most part this seems to be the case here.  However,  the throne speech hardly mentions groups, whereas manifestos and legislative packages have a much higher percentage reference.  Why would this be the case, given the above explanation?

Second, the nature of the throne speech sheds some light on this question.  The throne speech is an instrument for setting out a general framework for upcoming policy.  It can be seen as an abbreviated version of the manifesto and, for the party in office, as a ‘state of the nation’ address.  Upon examination one finds that it is more thematic in nature than much of the manifesto:  it divides the government party’s priorities into broadly defined topics, and commits them to being ‘proactive’ in the resolution of problems, for the most part without mentioning how it is going to do so.  

Where Promises and Politics Meet

The above discussion of the trends illustrated in charts one and two introduced briefly some of the substantive reasons for convergence and divergence in the emphases place on specific issue domains at specific points in time, by both parties.  In particular, these explanations point to a volatile political environment to which the governments had to respond -e.g. the constitutional issues that reared their head during both periods.

However, behind these substantive reasons lie the institutional constraints of parliamentary government.  In particular, certain issues must be dealt with in the House whereas others need not.  For example, the provision of social and economic services, financial expenditures and appropriations must be dealt with in parliament.  Whereas other decisions can be taken away from the floor of the House, being dealt with by orders-in-council, which are often far more plentiful than government legislation.  This too may contribute to the degree of disparity between legislation and its antecedent manifestos and throne speeches.

Moreover, as noted earlier, talk is cheap; the flip-side of which is, action is costly.  It is through such a lens that the disparity between talk (embodied in the rhetoric of manifestos and throne speeches) and legislative action in a domain such as ‘democracy’ can be understood.  Talk of constitutional change does not necessitate the introduction of government legislation until such time as that change is to be affected.  Furthermore, discussions of constitutional affairs have a significant public audience, as there is a perception by the public that such affairs are central to their own values.  Governments therefore tend to spend more energy discussing these issues, and mobilizing popular support for their proposals than in the actual, procedural, implementation of these proposals.  This was definitely true of the two periods under study here.

This leads us to what can be called the ‘expectation approach’.  The idea here is that governments, and political parties generally, act in an environment of uncertainty.  The contours of the political canvas are such that the diverse, regional, nature of Canada leaves government with one option: paint with a broad brush.  Vague and thematic manifestos, and to a lesser degree throne speeches, leave parties room to maneuver in order to respond to what may be around the next proverbial corner.

As such, governments find it necessary to emphasize and de-emphasize issues depending on the weight they give to three factors:

  • the saliency the issue has for the public writ large;
  • the expected value of pursuing a given issue; and
  • the anticipated nature of extraneous events with which a government may have to contend during the course of its mandate.

These factors have their greatest impact during elections, but their relevence is felt at all times for practitioners of politics.  The first factor refers specifically to the function of parties during elections, as they seek to mobilize their supporters, both actual and potential.  For example, Mr. Trudeau’s Liberals fought the 1974 election on the opposition to wage and price controls, and a general theme of economic growth.  Mr. Mulroney’s Progressive Conservatives emphasized free trade, national reconciliation and the preservation of Canada’s social programmes.  As opinion polls show, these issues were foremost in the minds of the electorate at the time of their respective elections.

Therefore, it can be suggested that ceteris paribus, parties emphasize certain issues only if they are salient in the minds of the electorate.

Factors two and three imply an amount of cost-benefit analyses mixed with an element of crystal ball gazing, and have greater weight for a party-in-government.  The expected value of pursuing a given policy is based upon the idea that all political action entails a cost, especially in terms of time and other resources as the party seeks to mobilize support for its policy, both inside parliament and within the electorate.  

This calculation is especially important for parties in Canada due to the omnipresent competing interests of regional and linguistic nature.  As such, it is necessary for parties-in-government to spend a lot of their resources in the brokering of interests in order to gain the needed support.  This in turn leads to the desire for flexible governments; governments that government from the centre of the political spectrum.

The third factor –crystal ball gazing– is part of any government, and political party, activity.  The idea here is that parties understand that during the course of their mandate they will inevitably face events that were not fully planned for.  It is the truism that ‘events will foil even the best laid plans’.  However unpredictable such events are the party-in-government will be expected to respond effectively.  Thus, parties anticipating extraneous events will design their platforms with what they perceive as being ample room for maneuver.  The result is an emphasis placed on those themes that the party believes are possible ‘unexpected events’. The rationale here is that such action helps to prepare the electorate for such events, and should the case arise, it is easier for the government to then raise support for its policy response.

This vague, thematic brokering of interests is one explanation behind the characterization of the  Liberal and Conservative parties as the ‘tweedledum and tweedledee’ of Canadian politics, as they converge in their approaches to governance and the issues they emphasize.  The two periods under study here would seem to be no exception. Both parties place considerable emphasis on the importance of a ‘strong economy, social justice, and national unity’.  For both a commitment to parsimony in economic affairs is essential for good governance; the ‘sacred trusts’ of Canada’s social safety net must be maintained; and the unity of the often unstable federation must be supported, through proactive policies and constitutional reforms.  

In summary, the party manifesto and throne speech are designed to be vague and thematic.  This allows the party-in-government room to maneuver and respond to political events that are beyond its control, and to attempt to shape the preferences of the public with their platforms and proposals for action. The Trudeau and Mulroney governments are no exception.

What the above analysis presents is a view of the trends in issue saliency emphasis at particular points in time. What the analysis does not do is reveal the changes in policy direction that may occur within these issue domains. But looking at such specifics was not, after all, the purpose of examining issue saliency transfer.  However, shifts in policy directions did occur during these two periods to some degree.  

One example of such a shift occurs during the 30th Parliament in the economic domain.  Over the course of this parliament the Liberal government’s throne speeches moved from emphasis on support for market regulation and intervention, to that of increasing economic parsimony.  In fact, the fourth throne speech outlines explicitly that “the primary objective [of this government] is to establish an economic climate that is conducive to private sector growth” by means of limiting government expenditure while contracting bureaucratic structures.5

The clearest example of a change in the direction of policy during Mulroney’s governments came during the 33rd parliament.  The Progressive Conservatives entered government in 1984 with an economic trade policy which ruled out a Canada-US free trade accord.  However, by 1987 this policy had been clearly reversed, as Prime Minister Mulroney and President Reagan announced the first stage toward such an agreement.


It has been said that in Westminster systems of government “constitutional theory is simple: a government can do anything it wishes.”  The disgruntled Canadian might respond by saying that they do what they want, but that what they want is never what they promise.  The trend illustrated for the two Parliaments studied here suggests that there is cause to side with governments, and not the majority of public opinion, on this issue.  

We have seen that parties do tend to emphasize the same issue domains when in government as when seeking the support of the electorate.  

We have also seen that where the emphasis in the salient issues diverge from the manifesto it can be traced to three possible reasons: First, the government is responding to meta-political issues such those raised in constitutional affairs.  Second, the nature of the throne speeches excludes references to specifics.  And third, governments face certain institutional constraints, for example in the form of procedures outlining which actions must be introduced into the House of Commons as bills, and those which can be taken without legislation.

Canadians do have cause to voice dissatisfaction where governments change policy direction completely during the course of their mandate.  But as long as governments remain vague in their election promises commiting little to paper, Canadians must remember that ‘talk is cheap’, and it is the actions of governing parties once in office that provide the best ground for judgement.


1. Neil Nevitte. The Decline of Deference, Toronto: Broadview Press, 1996.

2. Michael M. Atkinson. “What Kind of Democracy Do Canadian Want?” Canadian Journal of Political Science, vol. 27 no. 4, 1994, pp. 717-745.

3. Harold D. Clarke, Jane Jenson, Lawrence LeDuc, and Jon Pammett. Absent Mandate: Interpreting change in Canadian Election (2nd. Ed.) Toronto: Gage Educational Publishing, 1991, p. 25.

4. Canada, House of Commons, Debates, 30th Parliament, 4th Session, 1978, p. 1.

5. Ibid., p. 3. This speech further directs government departments to reduce their budgets by $1B, and projects a reduction in government spending for 1978-79 by $500M to $2B.

Canadian Parliamentary Review Cover
Vol 21 no 1

Last Updated: 2020-09-14