Equality via bottom-up referendums? The role of initiators

Direct democratic instruments are discussed by politicians and academia as a means to bring citizens “back in” and thereby to increase support for democracy. Especially bottom-up referenda are expected to offer disadvantaged groups an opportunity to put their issues on the agenda. Yet, the question remains open whether they are really successful in this regard. Are disadvantaged groups able to pursue their interests via direct democratic instruments? Or are (successful) bottom-up referenda mostly initiated by groups that are already influential to further improve their status? Encompassing, comparative research is missing. In order to close these research gaps, this paper addresses the following questions empirically: (1) Which groups initiate bottom-up votes? (2) How do the bills proposed by different groups relate to equality? And finally, (3) which bills succeed at the ballot? To answer these questions, we employ a descriptive analysis of all national direct democratic votes in European democracies between 1990 and 2015. We detect that bottom-up direct democratic instruments can give disadvantaged groups an opportunity to put their issues on the agenda. Our findings can serve as a basis for further, in-depth statistical analyses.

Introduction—direct democracy and equality

Direct democracy and especially bottom-up referenda¹ are increasingly popular these days, with its supporters arguing that it can provide the means to bring the citizens “back in” and increase support for democracy and democratic values (e.g., Dalton, 2004; Bowler et al., 2017). At the same time, the process of rising social and political equality, which has been essential for legitimizing democracy, is claimed to stagnate or even to be on reverse in Western democracies (Bartels, 2008; Kronauer, 2010; Castel, 2011; Neckel, 2013; Rosanvallon, 2013; Piketty, 2015).

Several scholars expect that especially bottom-up referenda can have a positive impact on equality, since they would offer disadvantaged groups an opportunity to get their issues on the agenda (popular initiative) or prevent parliamentary legislation that further disadvantages them (veto referendum) (Eder and Magin, 2008; Serdült and Welp, 2012). In contrast, there is also a large strand of literature arguing that in particular bottom-up referenda tend to be “hijacked” by established, well-off actors, providing an additional instrument to channel their political interests and positions into the system (Kriesi, 2007).

Convincing arguments have been made on both sides of the aisle, which are, however, mostly based on anecdotal and limited empirical evidence (Gamble, 1997; Kriesi, 2007; Vatter and Danaci, 2010; Hornig, 2011b; Serdült and Welp, 2012; Leemann, 2015; Welp and Ruth, 2017). What is missing so far is encompassing, comparative research on the questions who initiates bottom-up referenda, whether bottom-up referenda aim at increasing equality or at improving the positions of the well-off, and whether equality-increasing or -decreasing bills are supported in the popular vote. In other words: Does it make an equality-related difference whether bottom-up referenda are initiated by parties (left, right, in parliament or not), civil society (long-term, issue group, single person), others (e.g., companies) or multiple actors? In order to answer these questions, we draw on a comprehensive dataset containing all national direct democratic votes in European democracies between 1990 and 2015. Since bottom-up referenda differ between popular initiatives and veto referenda, we apply this differentiation. In contrast to popular initiatives, which suggest new legislation, veto referenda refer to existing parliamentarian legislation.

We answer our research questions by firstly offering an overview on (the initiators of) bottom-up votes,² secondly analyzing which bottom-up bills aim at increasing or decreasing equality, and thirdly comparing the success rates of votes initiated by different actors.

Our findings provide several instructive findings. They indicate that many votes are initiated by civil society groups, parties or coalitions of both. While more pro-equality bills are put up to vote, contra-equality bills, which are initiated less often, seem to have a slightly higher change to actually succeed. Popular initiatives mainly aim at increasing equality, while veto referenda more often include measures that hamper equality. Finally, a direct democratic bill is most likely to succeed at the ballot if the vote is initiated by a coalition of different actors.

Who initiates, who contributes to equality, and who succeeds? Theory, state of the art and assumptions

The following lines summarize the theory as well as the state of the art and conclude assumption. The lines are structured along our three guideline questions,³ namely: (1) Which actors initiate bottom-up votes? (2) Do the bottom-up bill proposals aim at increasing equality or improving the positions of those who are already rich and powerful? And finally, (3) which bills succeed at the ballot?

Regarding the first question on which actors initiate bottom-up votes, we have to start with clarifications of our definitions. Following Altman's (2019) definition of “citizen-initiated mechanisms of direct democracy” (Altman, 2019, p. 7) and Geißel et al. (2019a), we define bottom-up referenda as direct democratic votes initiated “by the people” through the collection of signatures. This includes votes on popular initiatives, where citizens draft a bill and gather enough signatures to put it to a direct democratic vote, and veto referenda, i.e., citizen-initiated direct democratic votes on bills that have already been adopted by parliament. We further differentiate between the bills, which include all direct democratic bills coming to a vote, and outputs, i.e., adopted bills that won a majority of votes and passed a possible quorum. Beyond the scope of our analysis is an investigation of the actual outcomes of these votes (see also Krämling et al., 2022).

Some scholars consider bottom-up votes as an opportunity for citizens, especially for disadvantaged groups, to put their issues on the agenda and to act as potential veto players (Serdült and Welp, 2012; Leemann, 2015). Other scholars reject the idea of bottom-up referenda as a tool of empowerment for disadvantaged groups. Empirical findings are mixed. Results from Switzerland reveal that direct democratic procedures are likely to be controlled by political parties—not by disadvantaged groups (Kriesi, 2007; Leemann, 2015). An analysis of all national bottom-up direct democratic experiences worldwide 1874–2009 shows that civil society actors are the most active initiators (Serdült and Welp, 2012). However, these civil society groups, as well, seem to have close links to political parties—a point also emphasized by Eike-Christian Hornig in his work on party dominance in direct democratic decision making in Western Europe (Hornig, 2011a; Serdült and Welp, 2012).

The further argument goes that groups using direct democratic tools need to have knowledge on the process, to provide certain organizational and financial resources, and to believe that they might win sufficient votes. Accordingly, parties and civil-society organizations can be expected to play a major role when it comes to initiating direct democratic votes. Regarding party characteristics, one can expect differences between inter- and extra-parliamentary parties. Extra-parliamentary opposition parties do not have the opportunity to introduce legislative acts via parliament and might therefore use the “detour” of direct democratic instruments. Considering civil-society groups we might differentiate between long term civil society groups such as unions and so-called “one-topic”-groups that are founded explicitly to initiate a referendum on a certain topic and campaign for it. The former can be assumed to have higher organizational resources and political networks than the latter. Thus, we distinguish in our analysis not only between civil society groups and political parties but also between inter- and extra-parliamentary parties as well as between long-term civil-society groups and so-called “one-topic”-groups. This distinction does not necessarily relate to equality but is considered as a major differentiation in the civil-society literature.

We now turn to the second question: do the bottom-up bill proposals aim at increasing equality or improving the positions of those who are already rich and powerful? To answer this question, we first of all have to explain our understanding of equality. We define equality as narrowing the gap between disadvantaged (A) and well-off groups (B) with regard to a certain benchmark (X) (see also Rawls, 1971; Alexy, 1986; Altwicker, 2011; Westen, 2016; Geißel et al., 2019a,b). In order to close this gap, groups need to be treated unequally sometimes—those who are worse off might get benefits at the expense of those who are better off (Sartori, 1992; Altwicker, 2011). As a result, we consider those direct democratic bills as pro-equality, which aim at narrowing the socioeconomic, political or legal gap between disadvantaged and well-off groups and thereby making society more equal. This might include proposals of equal or unequal treatment of certain groups.⁴

Regarding the outputs of bottom-up bills in terms of equality, it seems rather difficult to give a general projection. Respective findings are mixed. For example, welfare spending is found to be higher in countries with options for bottom-up votes, and their use increases redistribution in the long run in Swiss cantons (Blume et al., 2009; Blume and Voigt, 2012; Morger and Schaltegger, 2018). Comparative descriptive analyses of national bottom-up votes in Europe confirm these results by revealing that the outputs of these votes more often aim at increasing than at decreasing socioeconomic equality (Geißel et al., 2019a). In contrast, a negative effect of bottom-up votes compared to mandatory ones on minority protection in Switzerland is confirmed by Vatter and Danaci (2010). Also analyses of national votes in Europe by Geißel et al. (2019a) seem to point in a similar direction. Similarly, analyses of direct democratic votes in US states find that discrimination of minorities mainly happens in bottom-up votes (Gamble, 1997).

Generally, it is reasonable to assume that the intended impact of a proposed bill on equality depends on the initiator. If, for example, a disadvantaged group initiates a bill, we can expect the bill to aim at promoting the interests of this group, i.e., to enhance equality. If political parties initiate referenda, the equality-related goal probably depends on their ideological position. Conservative or rather liberal parties might put bills to a vote that benefit large economic players which—one might argue—could overall lead into a contra-equality direction. Social-democratic or rather left parties might push for bills that promote, for example, workers and union rights which have the potential to promote equality.

Research considering the relation between initiating group and the outputs of referenda are missing. None of these studies investigate whether the outputs of bottom-up votes in terms of equality differ depending on the group that initiates the vote. Our analyses will reveal if the assumptions mentioned above hold for the votes under investigation and lay the ground for further hypothesis-building and –testing on the topic.

Additionally, we will investigate whether there are differences in the equality-reference between popular initiatives and veto referenda. Proponents might use popular initiatives to target minorities with policies that would not make it through the more progressive legislatures (Lewis, 2013) and employ veto referenda to vote against parliamentary bills that aim at expanding the rights of minorities. In contrast, both tools might offer opportunities to protect minorities from oppressive parliamentarian legislatures and put their interest on the agenda. A differentiated analysis allows to detect whether one tool is more popular than the other in order to promote (or hamper) equality.

Finally, let us turn to the last question, i.e., which bills succeed at the ballot. One strand of literature deals with the question of what determines the success of direct democratic bills in general and (bottom-up) referenda specifically.⁵ However, very few look at relation between initiating group and output. Thus, we do not know whether bottom-up referenda initiated by powerful actors, i.e., actors with a lot of resources who have the capacity to run successful campaigns, are more likely to succeed at the ballot. We might assume that disadvantaged groups seldom fulfill these requirements. In addition, we can assume that parties also have a higher incentive to initiate referenda with a good chance of winning because they gain benefit for the next election. We can therefore expect that those referenda initiated by parties are more likely to succeed at the ballot than those initiated by other groups.

To sum up, the literature on the topic so far seems to suggest that the main actors in initiating bottom-up referenda are civil society groups and political parties. Building on our theoretical considerations, it makes sense to distinguish parties according to their ideological positions and representation in parliament. In addition, we differentiate between long term civil society groups such as unions and so-called “one-topic”-groups. The next section lays out our methods in more detail.

Data and methods

For our descriptive analysis we draw on a dataset of all direct democratic bills at the national level in European democracies in the period from 1990 to 2015. The dataset includes information on—among other things—the initiators of a given direct democratic bill, the results, voter-turnout, different context factors, and if a bill aims at increasing equality within society (pro-equality) or hinders equality (contra-equality).

For this distinction between pro- and contra-equality bills, we conduct an extensive coding based on the question if a bill proposes measures that fostered or hampered equality: every bill that proposes measures which would give the greatest benefit to disadvantaged groups, i.e., groups disadvantaged according to our equality dimensions, will potentially contribute to “narrowing the gap,” in the sense that it would improve the position of those groups relatively to better-off groups. Therefore, these bills are considered as equality-promoting (or pro-equality). Vice versa, bills that propose measures which primarily benefited better-off groups—and thereby further increased the gap to disadvantaged groups—are considered as equality-hindering (or contra-equality). The assessment is based on multiple data sources at hand such as the bill proposal itself, NGO reports, newspaper articles, political science articles and legal text. In order to increase intercoder reliability, multiple researchers did the coding independently from each other. Intercoder-Reliability is high (Cohen's Kappa of 0.9269). Table 1 gives examples of pro- and contra-equality bottom-up votes included in our dataset.

Table 1

Table 1. Examples of pro- and contra-equality bills.

For the purpose of this paper, we only look at bottom-up referenda, differentiating between popular initiatives and veto referenda. For all these votes we investigate who initiates the votes, if they are pro- or contra-equality votes, and if they ultimately succeed at the ballot box or not. We group the initiators in different categories in order to get a better overview on the data at hand. Table 2 lists these categories.

Table 2

Table 2. Categories of initiators.

As mentioned earlier, our analysis considers all national-level direct democratic votes that took place between 1990 and 2015 in European democracies. As democratic count all those European countries that—in the respective year—are considered free, according to the well-established Freedom House index. In total, our dataset includes 515 cases—since we only look at the bottom-up referenda in this paper, we end up with a total of 321 cases. Direct democratic bills that did not make it to the ballot (either because they did not fulfill the requirements, have been withdrawn or adopted by parliament) are not included in our analysis.

Most of the direct democratic votes in our dataset—and therefore also bottom-up votes—take place in Switzerland. Table 3 provides an overview of the distribution of votes per country.

Table 3

Table 3. Bottom-up votes per country, 1990–2015.

Switzerland does account for more than half of our total of 321 bottom-up direct democratic votes on national level between 1990 and 2015. Italy, Liechtenstein, and San Marino follow far behind. This has implications for our theoretical assumptions, as Switzerland is an exceptional case in many regards. For example, the governing parties are defined in advance to be the four largest parliamentary ones, so there is no coalition-building in the traditional sense. Members of the governing coalition often have to compromise and therefore face an incentive to distinguish themselves from the other parties in government. As a result, especially the governing parties at the more extreme end of the political spectrum might turn to direct democratic options in order to achieve electoral advantages. This somewhat modifies our assumption that above all extra-parliamentary parties will initiate bottom-up votes for the Swiss cases.

In addition, Table 3 reveals that a narrow majority of the bottom-up votes included are popular initiatives. This stems mainly from their popularity in Switzerland, whereas for example in Italy, veto referenda are more common.

The following section contains the results of our descriptive analysis, presenting first, the initiators of bottom-up votes, second, their outputs in terms of equality, and third, which votes succeed at the ballot.

Who initiates, who contributes to equality, and who succeeds? Empirical findings

Which groups initiate bottom-up votes? The second column of Table 4 shows that half of the bottom-up votes in our dataset are initiated by civil society groups, initiative committees founded to initiate this vote, or single persons, i.e., politicians, entrepreneurs. Figure 1 illustrates these findings for key actors. Parties (in and out of parliament) make up roughly 25% of all bottom-up votes, while the remaining 25% are initiated by a conglomeration of different actors to a large extent. These are mostly one or several parties, together with interest groups such as environmental groups, unions or religious organizations.

Table 4

Table 4. Initiated and adopted bottom-up votes per type of initiator.

Figure 1

Figure 1. Initiated and adopted bottom-up votes per type of initiator.

Investigating column 2 in more detail reveals that almost all of the initiating parties hold seats in parliament at the time of the vote. This contradicts the theoretical assumption that bottom-up votes are used above all by extra-parliamentary parties. Parties who already have legislative mandates seem to use this tool, either because they do not possess a majority in parliament to adopt a policy, because they want additional legitimation for it, or because they want to take the issue off the electoral agenda. Switzerland, with its consensual political system, does not seem to be the main driver of this phenomenon—some of the Swiss votes are initiated by parties represented in parliament, but their number is disproportionately low given the high share of Swiss votes in the total dataset.

Another surprising finding is that it is center-right or right parties who initiate a direct democratic vote more often than left parties. This speaks against the intuitive assumption of left parties being the main actor in this arena because they embrace a participatory view on how democracies should work. While we can assume this to be one of the motivations behind the 25 votes initiated by left parties, right parties obviously also know how to use direct democratic votes to achieve their goals. These might either be policy-oriented, e.g., in votes on asylum policy or rights of same-sex couples. And especially populist right parties use these votes to embrace their ideology of fulfilling the “true will of the people.”

In addition, civil society groups are indeed pretty active in initiating bottom-up votes—especially in Switzerland with its long direct democratic tradition. To Swiss civil society groups, direct democratic instruments seem to be important tools to put their interests on the political agenda. Given the importance and history of direct democratic votes in the Swiss political system, the groups probably have established strategies that help them to succeed in doing so. Most often, these are groups that existed for quite some time, providing the organizational and financial resources necessary to collect enough signatures and organize a successful campaign. However, there is also a considerable number of issue groups, i.e., groups that were founded for the single purpose of initiating a vote on a certain policy. We are not able to investigate who is behind all of these groups, but many of them are sub-sections of established civil society groups or a coalition of representatives from different groups. Finally, 10 votes are initiated by single persons—always ones with the necessary fame and fortune, such as politicians or entrepreneurs.

Are there differences between the types of initiators regarding the bottom-up tool they use most? Table 5 reveals that especially right parties in parliament initiate popular initiatives far more often than veto referenda. Whereas left parties use the popular initiative more often as well, center parties in parliament prefer veto referenda. Civil society and issue groups also tend toward popular initiatives, while coalitions of different actors show a clear preference for veto referenda. Unfortunately, we have little information whether civil society actors endorse certain bottom-up bills officially initiated by parties. In total, there are slightly more votes on popular initiatives than on veto referenda.⁶

Table 5

Table 5. Initiated and adopted popular initiatives and veto referenda per type of initiator.

How do the bills proposed by different actors relate to equality? Less than half of the bottom-up bills (and respective outputs) affect equality, i.e., 124 out of 321 referenda refer to an equality related issue. Differentiating them according to the types of initiators, the numbers in each slot get quite small, as Table 6 and Figure 2 show. Nevertheless, there are some interesting differences when we compare the initiators.

Table 6

Table 6. Pro- and contra-equality bills and outputs according to types of initiators.

Figure 2

Figure 2. Pro- and contra-equality bills according to types of initiators.

First, regarding the total numbers, we see more bills aiming at increasing equality than at decreasing it, whereas there are more policies adopted that have a contra- than those that have a pro-equality output. In the European context with mainly Swiss examples, bills potentially increasing the gap between disadvantaged groups and the better-off majority seem to be slightly more likely to succeed than bills potentially narrowing this gap.

Second, when looking at the initiators, the fact that the majority of bills have pro-equality intentions mainly stems from bills initiated by parties. There, the number of pro-equality bills is almost twice as high as the number of contra-equality ones. It is interesting to note that this applies to left as well as right parties, contradicting our expectation that left parties initiate pro-equality bills more often than right ones do. Numbers for civil society initiators and multiple actors are (almost) balanced. This also comes as a surprise, as one could have expected civil society to be more supportive of the interests of disadvantaged groups than already powerful parties are. But, as noted already above, many of these civil society groups are in fact well-established and powerful actors. And the issue groups often consist of members of different privileged groups. As a result, the finding that political parties often initiate bottom-up votes does not necessarily mean bad news for equality. On the contrary, these parties (regardless of their ideological position) often seem to push the interests of disadvantaged groups.

One last point that has to be stressed here is the higher number of contra-equality bills succeeding at the ballot compared to pro-equality ones. This concerns different kinds of topics: pro-equality bills aiming at de-monopolizing the Italian TV system, introducing a minimum wage, or preventing harsh measures in the asylum system fail at the ballot, whereas initiatives for stricter asylum policies or against welfare programs succeed. The higher success rate of contra-equality measures is especially obvious when we look at votes initiated by multiple actors: here, the number of pro- and contra-equality bills is the same, while contra-equality outputs clearly extend pro-equality ones. This finding suggests that the organizational and financial resources of the initiators are an important factor, with resourceful groups more often proposing contra-equality bills and being more successful in doing so. However, more detailed case studies would be necessary to examine in detail the specific topics. With the low number of actually adopted referenda (30) it is difficult to provide generalizable answers.

Table 7 reveals that this higher proportion of contra-equality outputs mainly stems from veto referenda. Although there are more popular initiatives adopted that include measures potentially decreasing equality than those that include pro-equality measures, the difference is smaller than for the veto referenda. In addition, veto referenda in general are already at the bill-level slightly more often contra- than pro-equality. These findings suggest on the one hand that a veto referendum is a popular tool to put parliamentary bills to a vote that benefit disadvantaged groups. On the other hand, they support the assumption that popular initiatives can be useful in order to promote interests of disadvantaged groups and put their issues on the agenda.

Table 7

Table 7. Pro- and contra-equality bills and outputs: popular initiatives and veto referenda.

Turning to the types of initiators, we see that especially parties initiate more often pro-equality popular initiatives than contra-equality ones. But parties as well as by civil society actors initiate a higher number of contra-equality veto referenda. Yet, due to the low numbers, this has to be interpreted with caution.

To sum up, bottom-up direct democratic bills seem to fulfill the promise of giving voice to the interests of disadvantaged groups in terms of narrowing the gap between them and a (better-off) majority. Especially popular initiatives provide a higher change to increase equality. In terms of different initiators, parties seem to do more for the interests of disadvantaged groups than civil society groups.

Are bills in general more likely to succeed if they are initiated by a certain type of actor compared to another one? Column 4 of Table 4 reveals that the main difference is between the success rates of votes initiated by one actor and those initiated by a coalition of multiple actors: the latter have by far the highest probability to succeed (more than 40% do), whereas only half as much are successful that are initiated by a party or members of civil society (15.9% resp. 19.4%). Obviously, the advantage of parties in terms of financial and organizational resources and access to the media compared to civil society groups or single persons is not as distinguished as one could presume. The recipe for success of bottom-up votes seems to be collective action of parties and civil society groups, combining efforts and their resources and influence to succeed at the ballot. Such collective actions also combine a multitude of constituencies, broadening the range of supporters and signaling to the voter that the issue is important enough to collaborate.

Conclusion

Direct democratic instruments and specifically bottom-up referenda are currently discussed vividly. But little is known about the initiators of bottom-up votes and their outputs considering equality. Do bottom-up referenda help disadvantaged groups to bring their issues on the agenda and improve their position? Or are respective instruments mainly used to serve the already well-off? This paper closes these research gaps by analyzing all national votes in European democracies between 1990 and 2015. It investigates the following questions: (1) Which groups initiate bottom-up votes? (2) Do the bottom-up bill proposals aim at increasing equality or improving the positions of those who are already rich and powerful? And finally, (3) Which bills are succeeding at the ballot?

Regarding the first question, it is mainly civil society actors who initiate bottom-up votes, but also (parliamentary) parties both from the left and from the right as well as coalitions of different actors. Turning to the second question, more bottom-up bills aim at increasing equality than at decreasing it. This finding is mainly due to votes initiated by parties, while the numbers of pro- and contra-equality bills initiated by civil society or multiple actors are more balanced. We find interesting differences between bottom-up tools—popular initiatives more often aim at fostering equality than veto referenda.

Summing up our results, bottom-up direct democratic instruments seem to give disadvantaged groups an opportunity to put their issues on the agenda—especially popular initiatives. Many of the votes in our dataset aim at increasing equality. But when it comes to which bills succeed at the ballot, the picture darkens. At least in the cases and timeframe we investigate, contra-equality bills are more likely to be adopted.

However, since the number of cases is rather small, the findings must be interpreted with care. One more limit of our study is due to the fact that more than half of the votes included in our sample come from Switzerland. Switzerland is special due to its long tradition of direct democracy and its consensual political system. More research is necessary. A point for further analysis should be the investigation of the initiators of bottom-up votes and their campaign strategies in more detail. Which civil society groups choose the direct democratic path? Which direct democratic bills do not make it to the ballot, but are successful in that they are included in government's policy? Results could yield ways how disadvantaged groups might benefit more from bottom-up votes, and what could prevent their interests from being discriminated against in these votes. In addition, future research might investigate factors influencing the initiation and the outputs of bottom-up referenda. For example, does the option binding results have an influence? Our finding that center-right or right parties initiate a direct democratic vote more often than left parties needs to be scrutinized in more detail. Also, a much more detailed analyses of the parties might be instructive: where there relatively new in parliament, when they initiated a referendum, have they been part of the opposition for a very long time, etc.? Finally, differentiating between bills that aim at increasing the rights and benefits of disadvantaged groups and those that aim at preventing a further deterioration of their situation can be an interesting topic for further analysis. There remains a lot to investigate on bottom-up direct democratic bills, their initiators and their connection to equality; this paper offers an important first overview and point of departure.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://search.gesis.org/research_data/SDN-10.7802-2128?doi=10.7802/2128.

Author contributions

BG: Writing–original draft. AK: Writing–original draft.

Funding

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by the German Research Foundation (DFG) under Grant GE 1083/23-1.

Acknowledgments

We thank the German Research Foundation (DFG) for their generous funding of the project Inequality and Direct Democracy in Europe. We also thank Lars Paulus, former team member, and our student assistant Amelie Buck for her efforts in data collection. Finally, we thank participants of the Workshop Innovation and Power: Exploring Democratic Innovation and Its Main Actors at the ECPR Joint Sessions 2020 and panelists of the panel Direct Democracy in Different Contexts at the IPSA World Congress 2021 for their feedback on an earlier version of this paper.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

1. ^“Bottom-up referenda” are defined as direct democratic votes initiated “by the people,” not initiated by government, parliament, or required by constitution or law. For a more detailed definition see Data and methods.

2. ^We include only bottom-up initiatives that actually came to a vote, not those that did not fulfill signature requirements, were withdrawn by the initiators or were accepted by the respective parliaments.

3. ^When looking at the current literature that deals with the initiators of (bottom-up) referenda, it becomes clear that a lot of studies focus on motivations for certain actors to initiate direct democratic procedures. Among the reasons mentioned are mediation, agenda, legislative, legitimacy, and power-reinforcement motives (Qvortrup, 2006; Morel, 2007; Rahat, 2009; Welp and Ruth, 2017; Gherghina, 2019). Since we, in this article, are not so much interested in the motivations of actors, we acknowledge this strand of literature, but do not put a particular focus on it.

4. ^Our Codebook in the Appendix provides further information on the coding procedure.

5. ^The literature names a large number of potential factors for the success or failure of direct democratic bills, among them institutional hurdles, the existence of multiple direct-democratic instruments (Moser, 1987), experience with direct democratic procedures, party support for a bill, and the existence of counterproposals (cf. Eder, 2010).

6. ^Note that bottom-up votes with missing information on the type of initiator are not listed here.

References

Alexy, R. (1986). Theorie der Grundrechte. Frankfurt: Suhrkamp.

Google Scholar

Altman, D. (2019). Citizenship and Contemporary Direct Democracy. Cambridge: Cambridge University Press. doi: 10.1017/9781108634397

Crossref Full Text | Google Scholar

Altwicker, T. (2011). Menschenrechtlicher Gleichheitsschutz. Springer VS: Berlin/Heidelberg. doi: 10.1007/978-3-642-18200-6

Crossref Full Text | Google Scholar

Bartels, L. M. (2008). Unequal Democracy. The Political Economy of the New Gilded Age. Princeton, NJ: Princeton University Press.

Google Scholar

Blume, L., Müller, J., and Voigt, S. (2009). The economic effects of direct democracy: a first global assessment. Public Choice 140, 431–461. doi: 10.1007/s11127-009-9429-8

Crossref Full Text | Google Scholar

Blume, L., and Voigt, S. (2012). Institutional details matter: more economic effects of direct democracy. Econ. Gov. 13, 287–310. doi: 10.1007/s10101-012-0115-9

Crossref Full Text | Google Scholar

Bowler, S., Denemark, D., Donovan, T., and McDonnell, D. (2017). Right-wing populist party supporters: dissatisfied but not direct democrats. Eur. J. Polit. Res. 56, 70–91. doi: 10.1111/1475-6765.12166

Crossref Full Text | Google Scholar

Castel, R. (2011). Die Krise der Arbeit. Neue Unsicherheiten und die Zukunft des Individu-ums. Hamburg: Hamburger Edition.

Google Scholar

Dalton, R. J. (2004). Democratic Challenges, Democratic Choices. The Erosion of Political Support in Advanced Industrial Democracies. Oxford: Oxford University Press. doi: 10.1093/acprof:oso/9780199268436.001.0001

Crossref Full Text | Google Scholar

Eder, C. (2010). Ein schlüssel zum erfolg? Gibt es ein patenrezept für volksentscheide in den deutschen bundesländern? Polit. Vierteljahresschr. 51, 43–67. doi: 10.1007/s11615-010-0003-2

Crossref Full Text | Google Scholar

Eder, C., and Magin, R. (2008). “Direkte demokratie,” in Die Demokratien Der Deutschen Bundesländer. Politische Institutionen im Vergleich, eds M. Freitag, and A. Vatter (Opladen: Barbara Budrich), 257–308.

Google Scholar

Gamble, B. S. (1997). Putting civil rights to a popular vote. Am. J. Polit. Sci. 41, 245–269. doi: 10.2307/2111715

Crossref Full Text | Google Scholar

Geißel, B., Krämling, A., and Paulus, L. (2019a). It depends...different direct democratic instruments and equality in Europe from 1990 to 2015. Politics Gov. 7, 365–379. doi: 10.17645/pag.v7i2.1881

Crossref Full Text | Google Scholar

Geißel, B., Krämling, A., and Paulus, L. (2019b). More or less equality? Direct democracy in Europe from 1990 to 2015. Z. Vgl. Polit. 13, 491–525. doi: 10.1007/s12286-019-00435-3

Crossref Full Text | Google Scholar

Gherghina, S. (2019). How political parties use referendums: an analytical framework. East Eur. Polit. Soc. Cult.33, 677–690. doi: 10.1177/0888325418803164

Crossref Full Text | Google Scholar

Hornig, E. C. (2011a). Die Parteidominanz Direkter Demokratie in Westeuropa. Baden-Baden: Nomos. doi: 10.5771/9783845228808

Crossref Full Text | Google Scholar

Hornig, E. C. (2011b). Direkte demokratie und parteienwettbewerb: überlegungen zu einem obligatorischen referendum als blockadelöser auf bundesebene. Z. Parlam. 2011, 475–792. doi: 10.5771/0340-1758-2011-3-475

Crossref Full Text | Google Scholar

Krämling, A., Geißel, B., Rinne, J., and Paulus, L. (2022). Direct democracy and equality—a global perspective. Int. Political Sci. Rev. 44. doi: 10.1177/01925121211058660

Crossref Full Text | Google Scholar

Kriesi, H. (2007). “The role of the political elite in swiss direct-democratic votes,” in Direct Democracy in Europe. Developments and Prospects, eds Z. T. Pállinger, B. Kaufmann, W. Marxer, and T. Schiller (Wiesbaden: VS Verlag für Sozialwissenschaften), 82–93. doi: 10.1007/978-3-531-90579-2_7

Crossref Full Text | Google Scholar

Kronauer, M. (2010). Exklusion. Die Gefährdung des Sozialen im Hoch Entwickelten Kapita-Lismus, 2. Aufl. New York, NY: Frankfurt a.M.

Google Scholar

Leemann, L. (2015). Political conflict and direct democracy: explaining initiative use 1920-2011. Swiss Polit. Sci. Rev. 21, 596–616. doi: 10.1111/spsr.12190

Crossref Full Text | Google Scholar

Lewis, D. C. (2013). Direct Democracy and Minority Rights: A Critical Assessment of the Tyranny of the Majority in the American States. New York, NY: Routledge. 111–117. doi: 10.4324/9780203109342

Crossref Full Text | Google Scholar

Morel, L. (2007). The rise of ‘politically obligatory' referendums: the 2005 French referendum in comparative perspective. West Eur. Polit. 30, 1041–1067. doi: 10.1080/01402380701617449

Crossref Full Text | Google Scholar

Morger, M., and Schaltegger, C. A. (2018). Income tax schedule and redistribution in direct democracies: the Swiss case. J. Econ. Inequal. 16, 413–438. doi: 10.1007/s10888-018-9376-z

Crossref Full Text | Google Scholar

Moser, C. (1987). Erfolge kantonaler volksinitiativen nach formalen und inhaltlichen gesichtspunkten. Schweiz. Jahrb. Polit. Wiss. 27, 159–206.

Google Scholar

Neckel, S. (2013). “Refeudalisierung” – systematik und aktualität eines begriffs der ha-bermas'schen gesellschaftsanalyse. Leviathan 41, S39–S56. doi: 10.5771/0340-0425-2013-1-39

Crossref Full Text | Google Scholar

Piketty, T. (2015). Das Kapital im 21. Beck: Jahrhundethomrt, München. doi: 10.17104/9783406671326

Crossref Full Text | Google Scholar

Qvortrup, M. (2006). Democracy by delegation: the decision to hold referendums in the United Kingdom. Representation 42, 59–72. doi: 10.1080/00344890600583792

Crossref Full Text | Google Scholar

Rahat, G. (2009). “Elite motives for initiating referendums: avoidance, addition and contradiction,” in Referendums and Representative Democracy. Responsiveness, Accountability and Deliberation, eds M. Setälä, and T. Schiller (London: Routledge), 98–116.

Google Scholar

Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Belknap Press of Harvard University Press. doi: 10.4159/9780674042605

Crossref Full Text | Google Scholar

Rosanvallon, P. (2013). Die Gesellschaft der Gleichen. Hamburg: Hamburger Edition.

Google Scholar

Sartori, G. (1992). “Gleichheit,” in Demokratietheorie, ed S. Giovanni (Wissenschaftliche Buchgesellschaft), 326–354.

Google Scholar

Serdült, U., and Welp, Y. (2012). Direct democracy upside down. Taiwan J. Democr. 8, 69–92.

Google Scholar

Vatter, A., and Danaci, D. (2010). Mehrheitstyrannei Durch volksentscheide? Zum spannungsverhältnis zwischen direkter demokratie und minderheitenschutz. Polit. Vierteljahresschr. 51, 205–222. doi: 10.1007/s11615-010-0019-7

Crossref Full Text | Google Scholar

Welp, Y., and Ruth, S. P. (2017). “The motivations behind the use of mechanisms of direct democracy,” in Let the People Rule? Direct Democracy in the Twenty-First Century, eds S. P. Ruth, Y. Welp, and L. Whitehead (Colchester: ECPR Press), 99–119.

Google Scholar

Westen, P. (2016). Speaking of Equality: An Analysis of the Rhetorical Force of “Equality” in Moral and Legal Discourse. Princeton, NJ: Princeton University Press.

Google Scholar

Appendix

Table A1

Table A1. Codebook.