Political Donations and Democratic Equality in Canada
Equality is a key tenet of democracy. With respect to the financing of federal political parties, one issue relevant to equality concerns has received surprisingly little attention: the phasing out of political parties’ annual per-vote subsidy, set to occur in 2015. Donations from individuals will henceforth become the parties’ primary source of funding; but not all Canadians donate equally. By examining a sample of disclosed donors from Elections Canada databases, combined with census data on neighbourhood income levels, this study establishes that donors are substantially more likely to come from wealthier sections of Canadian society. Despite a relatively low cap on donations – individuals can currently give no more than $1,200 annually – wealthier Canadians carry disproportionate weight in the total aggregate of donation dollars. The study concludes by briefly comparing federal rules to regulations at the provincial level and suggesting methods of mitigating inequalities in the political finance system.
The very essence of democracy is equality.1 Fairness in politics is undeniably something Canadians desire. Since the 2011 federal election, these principles have been front and centre, as a number of issues relating to the conduct of elections have been making headlines and have been actively debated by Canadians concerned about the quality of democracy in this country.
Through this period, one important issue has received surprisingly little attention: the phasing out of the per-vote annual subsidies for political parties.2
These subsidies were first introduced in 2004 as part of a larger package of reforms to party finance regulations, which included a ban on corporate and union donations and caps on political contributions by individuals. In part, the annual subsidies were designed to offset the revenue that would be lost from the new limits placed on donations. However, they also were seen by many as a means of creating a more egalitarian system of party finance, since each and every voter would have some control over the distribution of public funds to the various parties.
In 2015, these subsidies will be fully eliminated and political parties’ main source of financing will be donations by individual Canadians (currently capped at $1,200 per year).3 The parties will continue to benefit from very substantial public funding: not only are there generous tax credits for donations, but they are also reimbursed for a significant portion of their electoral expenses. Nevertheless, rather than public funds being partly directed by the voting preferences of all citizens, after 2015 they will be fully contingent on the support of individual donors.
One potential reason for the lack of debate on the changes may simply be that the prospective system based on donations is believed to have merit. Rather than parties receiving money automatically from the public purse, as occurs with per-vote subsidies, they will have to engage with their supporters to earn their money. The system also appears quite egalitarian since parties must rely on many small donations to secure substantial funds rather than a small number of large donations.
The presumption of equality, however, warrants closer investigation. While $1,200 can appear to be a relatively low cap, it does represent a substantial amount of money for many Canadians who might consider donating to a political party. It is reasonable to assume that contributions of this size are out of reach for many would-be donors in lower income groups. At the same time, larger donations have the potential to count for much more than smaller donations: receiving $1,000 from one person is the same as receiving $50 from 20 people. The concern, therefore, is that there may be significant inequalities in the system in the form of a general skew to donation patterns across income levels. This research seeks to assess whether these concerns about potential inequalities, in particular income inequalities, are merited.
Donations to Canadian Political Parties: What We Know
Prior research into political financing in Canada is quite limited. Despite substantial changes in political finance regulations over the past 15 years, including the banning of corporate and union donations, the introduction of caps on individual donations, as well as the introduction (and now elimination) of per-vote subsidies, there has not been much investigation of patterns of political giving.
The most relevant recent study, by political scientists Harold J. Jansen, Melanee Thomas, and Lisa Young, is entitled ‘Who Donates to Canada’s Political Parties?’.4 Jansen and his colleagues do find that those with higher incomes are more likely to give to political parties, but this does not emerge as a dominant factor in their analysis. Age is a more powerful demographic variable – older Canadians are considerably more likely to donate to parties – while other factors such as membership in a political party and political interest have the strongest effects on political giving. The main limitation of the study, however, is that it is based on surveys that only probe whether respondents have given money to a political party. The surveys do not ask about the size of their donation(s) and therefore this important dimension cannot be considered in the analysis.
Research in the United States that has taken the size of political donations into account has found income to be a more significant part of the story. In their 1995 book Voice and Equality: Civic Voluntarism in American Politics, Sidney Verba, Kay Lehman Schlozman, and Henry Brady examined various forms of political participation in the United States with a particular emphasis on the way in which tangible resources influence participation rates across different forms of political and civic engagement. With respect to political donations, they found that income had a very powerful impact, especially when the size of donations was taken into consideration. Those with deeper pockets contributed much more, on average, than those of lesser means. Of course, the principal limitation of this study for our purposes is that political finance law in the United States differs greatly from Canada, with few effective upper limits on the amount an individual can donate. It does, however, demonstrate the potential importance of examining donation amounts in order to assess the influence of income on political giving.
Methods
The current research sought to fill the gaps in our current knowledge by drawing upon available data sources to analyze potential inequalities in Canada’s system of political donations. Two main steps were involved in the data collection process (further details can be found in the Methodology Appendix in the electronic version of this article on the Canadian Parliamentary Review’s web site). The first was to gather information on a randomly selected sample of donors to the five parties with a representative in the House of Commons, by using publicly available data on donors giving $200 or more. Just under one thousand such donors (out of a total of 92,470) from the years 2010 and 2011 were sampled from databases downloaded from the website of Elections Canada. For each sampled donor, two key pieces of information were obtained: the size of their donation and their postal code.
The second step in the data collection process was to use the postal codes of donors in combination with 2006 census data to determine the median household income of the dissemination areas where donors lived (dissemination areas, or DAs, are the smallest geographical units used by Statistics Canada, each containing 400 to 700 individuals). Census data were also used to gather information on income levels for all Canadian DAs for comparison purposes.
As an additional follow-up step, donations by the same 1,000 individuals were tracked over a longer period (2007 through 2011) using the Elections Canada donor databases. The same step was carried out for individuals in the databases sharing the same surname and postal code as the initial donor – individuals assumed to members of the same family. This additional step provided information about total donations over a longer period for both individuals and families.
Finally, aggregate data from Elections Canada on total donation amounts for each of the parties was assembled and analyzed. This is the only component of our analysis which includes information on donors making contributions of $200 or less. Using this data allows us to establish this important point: while those giving more than $200 to a party in the years 2010 and 2011 made up just 24.7 per cent of all donors, their donations accounted for approximately 63.4 per cent of total donation dollars. While this skew in favour of larger donors would almost certainly be even greater if there were no cap on political donations, it is still very substantial.5 Therefore, while our analysis, by necessity, is limited to a minority of donors (since only individual donations over $200 are publicly disclosed), it does encompass a majority of the donation dollars. Focussing on this group is then a reasonable way of gaining some basic insights into potential inequalities in Canada’s system of political donations.
Findings: General Patterns of Donations
Table 1 provides information on donation amounts among our sample of donors (those giving more than $200 to a party in the years 2010 and 2011). Fifty-nine per cent gave an amount between $200 and $400. A further 23 per cent gave an amount between $400 and $750. The remaining 18 per cent donated between $750 and $1100 (the donation cap in 2010 and 2011).6
Table 1 also indicates mean donation amounts, for all donors combined ($495) and within each of the three categories of donors ($314, $556 and $1026). Using this information, Table 1 provides estimates of the share of the total donations coming from these three groups: 37.5 per cent, 25.9 per cent and 36.6 per cent, respectively. In other words, while there are more than three times as many donors in the ‘small’ donor category ($200 to $400) compared to the ‘large’ donor category ($750 to $1100), donations from the two groups account for roughly equal dollar amounts.
Donation Size |
Donors |
Mean |
Total donation dollars (%) |
---|---|---|---|
$200-$400 |
59.2 |
$314 |
37.5 |
$400-$750 |
23.1 |
$556 |
25.9 |
$750-$1100 |
17.7 |
$1026 |
36.6 |
Total |
100.0 |
$495 |
100.0 |
Using additional information gathered in a second phase of data collection, Table 2 shows patterns of donations over multiple years from 2007 to 2011, as well as family member donations over the same period. The majority of donors in the total sample are multi-year donors (nearly 70 per cent donated in more than one year). Furthermore, those who give larger amounts are particularly apt to be multi-year donors. Almost all in the top donor group (86 per cent) made a contribution in more than one year, with 48 per cent giving in four or five years. By contrast, in the small donor group 63 per cent gave in more than one year and only 22 per cent gave in four or five years.
Donation Size |
||||
---|---|---|---|---|
No. of years donating |
$200-$400 |
$400-$750 |
$750-$1100 |
Total |
1 |
37.3% |
27.0% |
14.1% |
30.8% |
2-3 |
40.9% |
34.3% |
37.9% |
38.8% |
4-5 |
21.8% |
38.7% |
48.0% |
30.3% |
Total |
100.0% |
100.0% |
100.0% |
100.0% |
Family member donating |
10.1% |
16.3% |
19.5% |
13.4% |
Smaller donors are also less likely to have family members making donations over the 2007-2011 period. We were able to identify donations from family members for just over 10 per cent in the small donor category, compared to nearly 20 per cent for the large donor group (Table 2).7 These points further underline the concentrated nature of political donations: the small minority who give the largest amounts tend to give more frequently and are more likely to have a family member giving as well.
Political Donations and Income
To further probe potential inequalities in the system of political donations, information was collected on income levels, both for donors and Canada-wide. Census data provided median household income8 for the dissemination areas corresponding to the postal codes provided by donors, as well as median household income for all DAs in Canada.
The Canada-wide data were used to calculate income quintiles: five income categories each containing 20 per cent of all Canadian DAs.9Our donor sample was then divided into the same five categories. If income were unrelated to political donations, we would expect to see approximately 20 per cent of the sample in both the bottom and top quintiles (and in all intermediate quintiles as well). Instead, as Figure 1 shows, 37 per cent of the donor sample were found to be living in DAs in the top income quintile, while 10 per cent were in DAs in the lowest quintile. Instead of a ratio of 1:1, the ratio of top to bottom income quintiles in the donor sample is nearly 4:1.
This inequality in political donations across income groups is even more severe for the top donor category. Among those giving amounts over $750, 51.7 per cent are in the top income quintile, compared to 8.0 per cent in the bottom quintile, a ratio of 6.5:1 (for the middle donor category the ratio is 3.8:1, and for the small donor group it is 3.1:1). There is also a connection between income and size of donation apparent in mean donation amounts: $485 for the lowest income quintile, compared to $543 for the top income quintile.
Unequal patterns of political giving across income categories are evident as well in the results for multi-year donations and family donations. In the top income quintile, 33 per cent had given in 4 or 5 years between 2007 and 2011, compared to 24 per cent of those in the bottom quintile. Similarly, we were able to identify donations from family members for 13 per cent of those in the top income quintile compared to just under 2 per cent (1.5 per cent) of those in the bottom quintile. The mean value of the total amount donated by individuals and their family members over the five year period also differed significantly: $1440 for the bottom quintile versus $1907 for the top quintile, a difference of 30 per cent.
In sum, the principal source of inequality in political donations across income groups is the simple incidence of giving: relatively few Canadians living in low-income areas give more than $200 to a political party in a given year (a group which accounts for nearly two-thirds of all donation dollars). This basic inequality is exacerbated by the fact that those in lower income areas who do make donations over $200 tend to give smaller amounts than those in high income areas, to give less frequently, and are less likely to have family members making donations.
As for those who donate amounts of $200 or less, we do not have individual-level data available to undertake similarly precise calculations. However, we do know, from the analysis of Harold Jansen and his colleagues reported above, that survey data asking about political donations in general reveal that those with higher incomes are more likely to be donors. Since those donating $200 or less account for about three-quarters of all donors, it is likely that these general results would hold for this group – in other words, that even among the smallest donors, there would be a greater incidence of individuals from high-income areas (though we would anticipate that income disparities would likely be less pronounced in this group, therefore having some mitigating effect). Any equalizing potential of donations of $200 or less is also limited by the fact that donations of this size only represent 37 per cent of all donation dollars.
Political Donations and Parties
One further inequality often noted about the pattern of political donations in Canada is that one party, the Conservative Party, has been much more successful in raising money than the others. In fact, the Conservatives up until recently have raised more money from their supporters than all the other parties combined.
It does not, however, appear that this fundraising edge derives from a special advantage for the Conservatives among high-income groups. The ratio of donors in the top and bottom income quintiles for the Conservatives is 3.6:1, which is just about the same as the overall result for all parties combined. For the Liberals it is substantially higher at 6.4:1, while for the NDP, it is 2:1. The mean donation amount to these three parties among our sample reflects these differences: $509 for Conservative donors, $523 for Liberal donors and $444 for NDP donors. Our samples sizes for the Greens and the Bloc are somewhat smaller, but the general patterns for these two parties look more like the NDP than the other two parties. If there is a difference between parties, it is mainly between the two traditionally dominant parties in Canadian electoral politics and the other three. But at the same time, the issue of income inequality cuts across parties: all are more dependent for their donation dollars upon Canadians with higher incomes.
Conclusion
Equality is a core principle of democracy. Clearly, the federal system of political donations has some noteworthy inequalities. Despite a seemingly low cap on donations (now set at $1,200 but soon to rise to $1,500), it remains the case that ‘big’ donors carry considerable weight, accounting for a substantial proportion of total donation dollars. Furthermore, these donors are disproportionately from the wealthier sections of Canadian society. Arguably, the fact that these donation dollars are multiplied by the provision of substantial public funding (in the form of tax credits and reimbursement of electoral expenses) only aggravates these inequalities.
The subject of political donations, and our system of political finance more generally, would benefit from further research and public debate. These issues are also pertinent at the provincial level, where rules governing political finance vary widely. Regulations in many provinces are more lax than at the federal level. A majority still allow political contributions from sources other than individuals. Caps on donations are generally higher or non-existent. In many cases, tax credits for donations are the principal source of public funding; the more egalitarian method of providing annual subsidies to parties based on votes in the previous election is less common.
From the standpoint of democratic equality, regulations in Quebec are the most rigorous. In 1977, Quebec was the first Canadian jurisdiction to ban corporate and union donations and to set a relatively stringent cap on individual donations ($3,000 annually). Recent debates in the province around issues of political finance led the Parti Québécois government to reduce the annual donation cap to just $100 in 2013. The province also uses a system of matching donations from government in lieu of tax credits, which presumably facilitates giving by those for whom money is tight and a tax credit at the end of the year is insufficient incentive to make a larger donation. In addition, Quebec has increased the annual subsides provided to parties based on votes at the previous election.
Assuming greater equality would be desired by Canadians, there are clearly a number of policy changes to consider. In our view, the most simple and pressing at the federal level would be the re-introduction of the per-vote subsidy to offset the inequalities evident in the system of political donations. Certainly these issues should be thoroughly examined and discussed in a way they have not been to date.
Notes
- Acknowledgements: Preliminary work on this paper was carried out as part of a course in political science research methods at the University of New Brunswick. Students in the class carried out the first phase of data collection, as well as aiding in the methodological design and examination of preliminary results. The assistance of the following students is gratefully acknowledged: Jeffery Barrieau, Tia Beaudoin, Jake Doucet, Megan Henwood, Sean Hutchins, Adetunji Lawal, Adam Melanson, Joshua Michel and Justine Salam. We would also like to thank Siobhan Hanratty, Government Documents and Data Librarian at UNB, for her assistance in accessing and working with the Statistics Canada databases used in this study.
- It is surprising for various reasons, not least because this was a key issue in the prorogation crisis of 2008, as the Liberal Party and the New Democratic Party moved to form a coalition, supported by the Bloc Québécois, in response to the Conservative Party’s attempt to eliminate the per-vote subsidies in their November budget update.
- The cap is increased from time to time to allow for inflation, It was increased from $1100 to $1200 in 2012 and will increase again to $1500 under provisions of the new Fair Elections Act.
- The paper was presented at the 2012 conference of the Canadian Political Science Association. See http://www.cpsa-acsp.ca/papers-2012/Jansen-Thomas-Young.pdf
- A relevant comparison can be drawn here with charitable donations, for which there is no cap. A 2012 report from Statistics Canada (based on the 2010 Canada Survey of Giving, Volunteering and Participating) found that donors in the top 25 per cent accounted for 83 per cent of all donation dollars to Canadian charities. This suggests that the concentration of giving is more pronounced for charitable donations than for political donations and that the cap on political donations does have some effect in mitigating potential inequalities. (Charitable donation results are reported in Martin Turcotte, ‘Charitable Giving by Canadians,’ Canadian Social Trends, April 16, 2012, p. 26).
- Donations of exactly $400 (16.4 per cent of the sample) are included in the $200-$400 category. Donations of exactly $750 (9.7 per cent) are included in the $400-$750 category. Donations of exactly $1100 represent 8.0 per cent of the sample. Our choice of donor categories corresponds to the tax credit system: donations up to $400 receive a tax credit of 75 per cent, amounts between $400 and $750 earn a credit of 50 per cent, and amounts over $750 receive a credit of 33.33 per cent.
- These should be considered lower end estimates, since the method of identifying family members was a conservative one. See Methodology Appendix for details.
- The precise variable used is median family income for ‘economic families.’
- The precise income quintiles are as follows: $47,969 or less; $47,969.10 to $59,481; $59,481.10 to $71,189; $71,189.10 to $87,012.80; and $87,012.90 or above. DAs for which median income data are suppressed by Statistics Canada due to small sample sizes (2.8% of all DAs) were excluded from these calculations.
Methodology Appendix
Sampling
The population in the main portion of our study is those donating more than $200 to one of the five political parties during the years 2010 and 2011. Simple random sampling would have produced a sample with more than 50% Conservative donors. In order to ensure reasonable sample sizes for the other parties (for purposes of analyzing donation patterns across parties), disproportionate stratified sampling was used instead. This involved under-sampling Conservative donors and over-sampling the other parties (while still using random selection procedures for each party). The final sample sizes are shown in Table A-1. Statistical weights were applied in all analysis to correct for this disproportionate sampling and ensure that the results accurately reflected the population from the sample was drawn.
The margin of error for percentage measures based on the full sample of 999 donors is plus or minus 3.1%, 19 times out of 20.
Year |
|||
---|---|---|---|
Party |
2010 |
2011 |
Total |
Conservative |
199 |
199 |
398 |
Liberal |
100 |
100 |
200 |
NDP |
100 |
100 |
200 |
Green |
50 |
51 |
101 |
Bloc Québécois |
49 |
51 |
100 |
Total |
498 |
501 |
999 |
Data collection
Data on political donations were gathered from spreadsheets (for each party and year) downloaded from the Elections Canada web site (these can be found in the Contributions and Expenses database – under Registered Party Financial Transactions Return, Statement of Contributions Received). Often donors had made more than one contribution in a given year, in which case the total donation amount for the year was recorded. In a handful of cases, donations were to leadership candidates rather than the party. These donations were excluded from our calculations.
Postal codes of donors were also recorded. In order to link these postal codes to dissemination areas, the Postal Codes Conversion File from Statistics Canada was used. Census data from the 20 per cent long-form sample were then used to gather information on median family incomes (for ‘economic families’) in these dissemination areas. The 2006 census was used rather than the 2011 census due to concerns about data quality for the latter, following the controversial switch to a voluntary long-form census questionnaire. The 2006 data were felt to be sufficiently recent to provide accurate income data at the DA level, since average income levels within neighbourhoods are unlikely to change rapidly in the space of a few years.
Census data at the DA level are sometimes suppressed, when sample sizes are small, in order to ensure the confidentiality of respondents. For 25 donors (2.5 per cent), median income data were not available for this reason. Therefore, results reporting donation patterns across income groups are based on 974 cases.
The income data in our study reflect incomes in the areas where donors live, not the incomes of individuals. This is an unavoidable limitation of our study. There could be some individuals misclassified by this method (richer people living in relatively poor DAs or poorer people living in relatively wealthy DAs). But these errors, to at least some degree, would presumably cancel one another out. Further research using other methods that could elicit individual income levels (e.g. surveys) might be beneficial to corroborate our results.
Tracking individuals, 2007-2011
As a follow-up step, we also used Elections Canada donor data from 2007-11 to determine five-year donation patterns. In doing so, it was assumed that where there was a match of both first and last name for different years, it was the same individual donating, even if the person’s postal code had changed (which was the case for 12 per cent of multi-year donors). In some cases, the use of middle initials provided further confirmation of identity.
Identifying family members
Collection of data on family members donating to the same political party as the donor was also undertaken. Family, for our purposes, was defined as any person sharing a surname and postal code with the donor. Only donors whose last names occurred five times or fewer were included in this analysis. The concern, for more common surnames, was that there could be two donors sharing the same surname and postal code who were not, in fact, members of the same family. In cases where the surname appeared more than five times, the family donation variables were coded as missing (which results in a smaller sample size of 660 for family donation results).
By basing our analysis of family donations on shared surname, we do not take into account the many different family arrangements that exist which involve different surnames being used by family members. For this reason, our estimates of the incidence of family donations are likely conservative.
Verifying residential status
In the course of doing this research, it came to light (from media accounts of alleged infringements of donation laws in the 2011 federal election) that people do not necessarily use their home addresses when declaring their donations, but may instead use a work address. If this were a common practice, it could undermine the methodology of our study, as we would be collecting information on income levels in the neighbourhoods where people work, not where they live.
To address this concern, a verification procedure was carried out. Using a randomly drawn sample of 30 donors from our database, their names and places of residence(town or city) were entered on the White Pages website to see if a residential listing appeared for that individual. In cases where this method was unsuccessful, Google Maps was used to examine the satellite image of the postal code area in question to determine if it appeared to be a residential area. In all but two cases we were able to confirm that the postal code in question did correspond to a residential neigbourhood.
Other issues
In perusing donations data on the Elections Canada website, we became aware of an exception to the law concerning donation limits: there is no limit on political donations coming from a person’s estate. For example, in 2010, a bequest of over $210,000 was made to the NDP. This would be a worthwhile area to examine for additional information on the equality of the political finance system. Dedicated and wealthy donors, the same individuals who already make significant donations on an annual basis, might leave large sums of money to the political party of their choice.