Winning votes: the comparative importance of money and time on parliamentary candidates’ electoral performance in Estonia

ABSTRACT

The literature on post-communist democracies has traditionally suggested that expensive media-based campaigns are key to electoral success. Using data from the 2011, 2015, and 2019 Estonian Candidate Study, this article provides an up-to-date evaluation of how important monetary and non-monetary campaign efforts are in shaping candidates' electoral performance in a post-communist democracy. It finds that, while campaign spending continues to influence candidates' electoral fortunes, candidates need to significantly outspend their rivals to enjoy a meaningful increase in their vote share. There is also emerging evidence that candidates are starting to electorally benefit from devoting more time to promoting their candidacy.

KEYWORDS:

• Electoral performance
• campaign effort
• parliamentary candidates
• Estonia

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 See, for example, van Biezen (2003), Chan (2001), Kopecký (1995), and Mair (1997).

2 See, for example, Ibenskas (2014) and Tavits (2012, 2013).

3 The notable exceptions include studies by Koc-Michalska et al. (2014) and Trumm (2016).

4 For background, please note that while there is a natural cap in how many hours it is possible for any candidate to spend on her campaign, the Political Parties Act of 2014 (Riigi Teataja 2015) – and its previous versions – are relatively unrestricted when it comes to regulating campaign spending in Estonia. Parties are limited to how they can raise funds – i.e., allowed sources include membership fees, state funds, donations, and revenue earned from dealings with party assets –, and there are restrictions on the kind of donations parties and candidates can accept – e.g., donations from foreign interests are banned, unless from persons holding the permanent right of residence or the status of a long-term resident in Estonia –, but campaign spending remains unrestricted for parties as well as individual candidates. There is no legally defined campaign period and campaign spending remains uncapped for both in terms of their overall spend as well as spend on any particular type of campaign advertisement. There is a ban on vote buying, but parties and candidates are effectively free to spend how much they want (and can).

5 The studies by Gibson and McAllister (2011, 2015) do not control for campaign spending.

6 There are also a number of other recent studies exploring the electoral performance of individual candidates in post-communist democracies. These studies, while valuable in furthering our understanding of the different factors that contribute to candidates’ electoral success (or failure) in the region, do not tend to capture the effect of campaign effort. For example, Allik (2015) focuses primarily on the role of gender, Górecki and Kukołowicz (2014) and Jankowski and Marcinkiewicz (2019) on gender and gender quotas, Marcinkiewicz (2014) on ballot position effects, and Ragauskas (2021) on gender bias in terms of list placement and partisanship.

7 A study of candidates’ electoral performance in Estonia will also contribute to the broader research agenda that focuses on preferential voting in open list systems.

8 The range of questions included in the different versions of the Estonian Candidate Study provides a valuable opportunity to explore how various factors shape candidates’ electoral performance over time, but it is important to also note that there are some data-related limitations arising from the coverage of the different versions of the Estonian Candidate Study. For example, we know that candidates’ expected likelihood of success tends to shape their campaign choices (Sudulich and Trumm 2019), and it is also likely to relate to their actual electoral success (or failure), but cannot be included in this study as the question was not asked in 2011. The role that candidates’ pre-campaign evaluation of their electoral chances plays in influencing the relationship between campaign effort and electoral performance is something that future research should explore further.

9 The Estonian Candidate Study surveys have received ethics approval from the University of Exeter (2011), the University of Sheffield (2015), and the University of Nottingham (2019).

10 Please see Appendix A for further information about the samples.

11 Please see Appendix B for additional descriptive information about the variables used in the analyses.

12 Campaign spending relates to money spent as part of candidate’s personal election campaign. As such, it does not include party-level spending, national or regional, that might also promote the candidate in question through campaign materials such as billboards featuring the pictures of all party candidates in a constituency. The survey question was phrased “Thinking about your campaign budget, how much money did you spend on your personal campaign?” and candidates were able to provide an open-ended answer. Please see the original questionnaire for the Estonian version of the question at https://www.siimtrumm.com/estonian-candidate-study.

13 Estimates from models that exclude candidates who did not spend anything on their campaign and candidates who spent more than three times the constituency average are robust to those reported here.

14 The relative campaign spending measure also mitigates the endogenous nature of it (Benoit and Marsh 2010).

15 Campaign time relates to hours spent campaigning by the candidate herself. It does not include any campaign support, in the form of campaign time, provided by volunteers through activities such as canvassing on behalf of the candidate. The survey question was phrased “How many hours per week did you spend campaigning during the final month before the election day?” and candidates were able to provide an open-ended answer. Please see the original questionnaire for the Estonian version of the question at https://www.siimtrumm.com/estonian-candidate-study.

16 The two measures tap into slightly different aspects of candidates’ campaign effort not only on theoretical, but also empirical grounds, correlating at less than 0.4 at all three elections. Moreover, variance inflation factors are low for all independent variables in the multivariate models, providing further evidence that including campaign spending as well as campaign time simultaneously to the models does not lead to multicollinearity problems.

17 Cases that fall outside the 0–10 range for vote share and/or 0–6 range for campaign spending and/or campaign time are excluded from the figure.

18 An exploratory analysis of candidates who stand out by performing better or worse than one would anticipate, given their campaign effort, offers mixed insights. There does not appear to be a clear pattern with regards to the latter in terms of what might explain under-performance, but over-performance does seem to be linked (at least to some extent) to candidates’ public profile. Many of the former are either established political figures, at local and/or national level, or hold relatively high-profile and public-facing occupations such as actors and doctors.

19 Estimates from models that operationalise candidates’ list placement as their list position, rather than whether they headed their party’s constituency-level candidate list or not, are robust to those reported here.

20 It is important to note that the campaign time measure does not capture non-monetary campaign support that a candidate may receive from volunteers on her behalf. It is possible that the effect associated with campaign time may be stronger if the measure included time spent by the supporters of the candidate on behalf of the candidate.

21 Predicted values are provided for all effects, other than those associated with partisanship, that are statistically significant in Table 2.

Additional information

Notes on contributors

Siim Trumm

Siim Trumm is an Associate Professor in the School of Politics and International Relations at the University of Nottingham. His research focuses on electoral politics, representation, and political participation. His recent articles have appeared in the British Journal of Political Science, European Journal of Political Research, British Journal of Politics and International Relations, and Political Studies.