The Problem

Direct democratic practices have internationally gained in importance in recent years. While the direct par-ticipation of citizens has a long tradition in Switzerland and in more than half of the US States (1),  direct de-mocratic decision practices have become widely accepted in other countries only in the most recent past (2).  Given the current political situation in Germany, the introduction of direct democratic elements on the fed-eral level is hardly to be expected (3).  However, on the level of the federal states one may very well speak of a “triumph of direct democracy” (4).

While the majority of the population invariably argues in favour of a direct participation in political deci-sion-making processes in the form of popular initiatives and referenda, academic and public discourses of-ten voice a fundamental scepticism (5).  In Germany, critics argue against the introduction of direct democracy on the federal level by invoking above all the – allegedly – negative experiences in the Weimar Republic (6).  They also fear that interest groups with a high potential of resources might use the decision mechanisms of direct democracy for their own purposes and enforce a policy at the expense of the majority (7).  Last but not least, critics fundamentally doubt that electors in direct democracy have the competence to make (complex) collective decisions on factual issues (8).  They assume that direct democratic practices would entail high costs of information for all citizens; that citizens would not inform themselves sufficiently or, at worst, not prop-erly at all. Presupposing this kind of information behaviour, one would have to dismiss all referenda as arbitrary. Assuming this degree of uninformed-ness, the results of direct democratic votes would hardly repre-sent the citizens’ preferences.

Recent studies have, however, shown that the fundamental scepticism vis-à-vis direct democratic decision mechanisms is misplaced. For one, new theoretical models have become available which have drawn atten-tion not just to the costs of direct democracy, but also to the costs of practices of representation (9).  For another, empirical studies and laboratory experiments indicate that the presupposition of a fundamental incompetence of voters in votes on factual issues is wrong (10).  On the basis of theoretical considerations, the present article aims to show that voters do not only make rational decisions when choosing parties and poli-ticians, but they also – under certain conditions – decide in terms of their own preferences when factual is-sues are concerned.

The article is structured as follows: I will first categorize the costs of direct democratic and representative practices from a constitutional economic perspective (section 2) and then examine more closely the cost of information in direct democracy relevant to the study (section 3). On the basis of these theoretical consid-erations I will make some recommendations for the implementation of direct democratic practices.

The Optimal Degree of Delegation in Democracy

Should individuals in a democracy vote themselves on collective interests or should they elect representa-tives who decide on their behalf? In the overwhelming number of real and existing democracies, citizens delegate their decision competence to intermediaries, who – for a legislative period – are to represent the interests of their voters to the best of their knowledge. However, as the institution of representative democ-racy is not an exogenic given of a collection of data of a society, we must ask ourselves whether and to what extent handing over decision competence to delegates is efficient.
I will therefore try to develop criteria according to which an individual having to decide upon an optimal constitution in view of delegating collective decisions will act (11).  The constitutional problem can be reduced to the question of choosing the optimal degree of delegation (12). 
To analyse the individual constitutional calculus in view of delegating collective decisions, let us first of all consider the case of a pure direct democracy. What are the costs arising for an individual in such a constitu-tional arrangement? To decide in favour of any one of the available options, citizens will acquaint them-selves with the upcoming ballot topic as well as the preferences of other voters. However, the individual’s search for information involves expenses: citizens have to acquire, take in and store information. Furthermore, as the search for information squeezes out other profitable activities, temporal opportunity costs arise. The process of finding intra-personal consensus – depending on the kind of collective decision – obviously involves profit penalties for the individual. These costs will in the following be referred to as the costs of finding intrapersonal consensus. As democratic societies must manage a large number of collective issues, each individual would obviously have to accept high costs of finding intrapersonal consensus for a variety of collective decisions. Therefore, having representatives take over the decision competence on behalf of individuals may in many cases well represent a welfare increase in democratic society. Graph K1 in figure 1 shows the cost of finding intrapersonal consensus as a function of the degree of delegating in a democracy. If the degree of delegation is 0, that is, in the case of a pure direct democracy, the costs of finding intraper-sonal consensus are highest.

Along with the costs of finding intrapersonal consensus, each individual will further judge the optimal constitution according to the degree to which collective action comes close to his or her own preferences. As it is advantageous for the members of the collective – as far as the costs of finding intrapersonal consensus are concerned – to not make all decisions themselves, but to cede their decision competence to representatives, the members of the collective (principals), by delegating their power of decision, simultaneously run the risk of having representatives (agents) who do not always act on their behalf, but who act opportunistically (13).  According to the assumptions of economic behavioural theory, agents act corresponding to their own goals, which may differ from the goals of their principals. Principals, moreover, face the problem of being only imperfectly informed about their agents and their agents’ behaviour. The assumption of oppor-tunistic behaviour combination with the asymmetrical distribution of information results in the agency problem: Collective decisions made by delegates do not in all cases correspond to the preferences of the majority of individuals (14).  The individual facing the constitutional problem will to a certain degree of probability confront this problem and will incorporate corresponding profit penalties in their individual calculus. Profit penalties that an individual delegating his or her decision competence to a delegate may have to accept will be referred to as (probable) external costs of delegation.

If an individual considered a constitution exclusively from the point of view of the (probable) external costs of delegation, those costs would reach their maximum with a degree of delegation equal to 1. Those costs can evidently be lowered by choosing lower degrees of delegation. The larger the extent of collective decisions that are directly made by individuals, the lower the (probable) external costs. In the extreme case, all members of society always vote directly on collective issues. In the case of a pure direct democracy, the (probable) external cost of delegation by definition reaches its minimum. Graph K2  in chart 1 illustrates this relation. The sum of these two cost factors results in the total cost GK. The minimum of the graph of total costs GK represents the optimal degree of delegation in a democracy. The optimal constitution in terms of delegating collective decisions therefore lies in the principle of a mixed system, in other words in a semidirect democracy (15). 

Fig. 1: Optimal degree of delegation in democracy   

<!--[if !vml]--><!--[endif]-->

Source: Author.

The Market for Information in Direct Democracy

In direct democracy, citizens need more information more often because in addition to casting their ballot in parliamentary elections at regular intervals, they vote on specific factual issues. The question is whether citizens in such an institutional arrangement are offered incentives to search for the information required for making a decision so that they may eventually decide “correctly” in terms of their own preferences (16).  A “correct” decision in this context is meant to refer to a preference-based decision for the voting individual. An individual will evidently make the right decision for himself if he not only knows his own preferences, but if he also gathers further factual information on the subject of the vote. However, citizens in a direct democracy accept information costs only if the benefit of information gathering exceeds the costs. In the following, I will therefore analyse to what extent citizens in a democracy are subject to incentives in order to request information on the upcoming ballot drafts and whether institutions emerge in a direct democracy which provide an efficient offer of information.

Demand Conditions

The following demand conditions for information must be considered significant for a direct democracy:

• Information gathering and selection: Voters would incur substantial costs of information gathering if he or she commissioned a third party to provide this information for him/her. In modern democracies characterised by mass media, information is made available to citizens almost free of charge so that the pro-curement cost of information can be considered negligible (17).  Yet, precisely because voters have such an abundance of information at their disposal almost free of charge, selecting information involves costs for the individual (18). 
• Taking in and processing information: Individuals taking in and processing information on the subject to be voted on have temporal expenditures. They will, however, only accept those expenditures if the marginal cost of informing themselves is smaller than or equal to the marginal utility (19).  For one thing, we can assume that the cost of information gathering in view of collective concerns is higher for the in-dividual than the cost of a private good (20).  As “the factual, temporal and spatial range of variation of consequences resulting from collective decisions is very wide, generally wider than the one of private decisions” (21), it is relatively costly for an elector in direct democracy to make decisions that are in the in-terest of his own preferences.
For another, electors in direct democracy will only benefit from their commitment, if the result of the vote is being changed and an additional benefit is thus being realised indirectly through the process of taking in and processing information and through casting a corresponding vote. As the probability of having a decisive influence on the final result of a vote by making one’s own decision is extremely low, the yield of informing oneself must be considered marginal from this perspective.
Last but not least, well-informed collective decisions represent a collective good in themselves. As excludability cannot be taken for granted, members of society who do not participate in a vote and who do not inform themselves benefit from a well-informed collective decision. The costs of keeping oneself informed accumulate for the respective individual, while profits are spread out over the entirety of members of the society. In this logic, an individual in the context of general votes in direct democracy will hardly endeavour to obtain decision-relevant information. Seen from an information economics viewpoint, demand conditions for information in (direct) democracy can therefore be characterised by the phenomenon of rational ignorance (22).
• Although electors in direct democracy have but little influence on the result of the vote and although the informative commitment has the character of a collective good, one can nevertheless empirically ob-serve that many members of society inform themselves (23).  There are several reasons for this behaviour (24):
1. In the run-up to the decision of the vote, certain citizens outweigh others in private and/or public discourses. Journalists, representatives of associations, intellectuals or politicians may possibly use this weight to influence the decision of the vote in their favour. Because they benefit from a higher marginal utility, these individuals will obviously be ready to accept information costs.
2. Along with the investment use of information considered up to now – a use resulting from a proper and thus welfare-increasing collective decision – one may also ascribe a consumptive value to informative commitment. Regardless of whether other members of society benefit from the increased level of information, voters keep themselves informed because the subject of the vote draws their interest and because information as such represents a benefit. Taking in and processing information becomes a private good.
3. As the run-up of general votes involves a process of direct discussion on the subject of the vote among citizens, the personal level of information may become a “matter of prestige”(25).  The social environment of a person may build up high expectations regarding the level of information to which the person concerned will live up to.
4. As citizens in direct democracy are directly and visibly concerned, they will evidently make greater efforts at obtaining, taking in and processing the information that is relevant to them (26).

Terms of the bid

The following terms can be considered essential for the information offer in direct democracies:

• In direct democracy, similar to representative democracy, institutions that can significantly reduce information costs will arise. In the pre-decision stage, opponents and advocates of the voting draft will form and aim to influence voters. Slogans and watchwords will enter public debates. During this decision process, standardised statements and comments of parties and stakeholders make it easier for citi-zens to take in information (27).  It may very well be rational for citizens to transfer a part of their information sovereignty and follow the recommendations of lobbying groups (28).  If citizens should lack the respective resources, they can save a large amount of information costs by trusting the accuracy of respective recommendations.
In the course of voting discussions, competition between advocates and opponents of the voting draft results in an indepth exchange of arguments and in criticism of the opposite party. Similar to private market economies, competition in the run-up to votes in direct democracy can be described as a “dis-covery process” following v. Hayek. Without the competition between voting parties, much information, individual preferences and new attempts at solving a problem would remain undiscovered (29).
Electors in direct democracy may benefit from the information offer resulting from this competition process at a low price and make an informed choice with respect to their personal interest (30).  Citizens in direct democracy are therefore offered a wider range of information compared to citizens in representative democracy.
• As certain citizens will have a higher weight in private and/or public discussions than others, the former will accept information costs and try to influence the decision of the vote in their interest. Along with lobbyists and politicians, specialists in particular will intervene in discussions and introduce their expertise on the subject of the vote.
• Apart from the information offer arising spontaneously due to the competing voting parties, it is customary in many direct democracies to provide electors with official information on the subject to be voted on before the final vote (31).  In Switzerland, each elector receives an information brochure prepared by the government before casting his or her vote; this brochure gives reasons for the ballot draft and lists the arguments of both proponents and opponents. Similarly, it is common practice in the US states with direct democracy governance components to distribute so called elector handbooks among citizens (32).  This kind of institutionalised information offer may substantially reduce information costs for each citizen and contribute to a certain objectification in the information competition.

In direct democracy, both the demand for information and the information offer differ from those in a repre-sentational system. For one, the institution of direct democracy offers more incentives for citizens to be well-informed. For another, the providers of political information are forced to offer information not only on the occasion of periodically upcoming parliamentary elections, but also in the case of concrete factual decisions. Through the interplay of supply and demand, mechanisms that may substantially reduce the information costs of individuals and increase the benefits of informative commitment will develop in direct democracy (33). 

Conclusion

As has been shown by a constitutional economic analysis of the optimal degree of delegation in direct democracy, neither a pure direct democracy nor a pure representative democracy can in principle be judged efficient in terms of transaction costs. Depending on its concrete set-up, a mixed system combining elements from direct and representative democracy will most likely minimise the sum of the (probable) exter-nal delegation costs and the intra-personal costs of finding consensus.

Based on the mentioned theoretical considerations, the following recommendations for the practical design of direct democratic decision procedures can be made:

1. An objectified basis of information raises the rationality of direct democratic decision making to a significant degree. As is common practice in Switzerland and in many US states, electors should receive information handbooks or brochures well before they cast their vote; moreover, handbooks and brochures should present the arguments of opponents and proponents in a balanced manner.
2. Competition among the providers of information in the run-up to the vote is to prevent a systematic distortion of the information offer. The danger of a selective information offer is most likely to oc-cur if the proponents and opponents of the subject to be voted on have an information monopoly. In order for information in direct democracy to be communicated appropriately, it is essential that electors have free access to information relevant to the decision.
3. Discussions in the run-up to general votes bring about an exchange about the pros and cons of the subject to be voted on. As citizens engage in political discussions on specific factual issues, at least a part of the electorate shows a learning effect. Extended timelines in the run-up to votes can pre-vent the major negative effects of “Stimmungsdemokratie” or spin doctor democracy.

Direct democratic decision mechanisms should be given greater attention in the political system as they contribute to managing highly differentiated abstract societies. In analogy to economic markets, managing political processes in modern democracies may happen in a much more differentiated manner in the context of direct democratic procedures. The periodic casting of votes for a party no longer adequately represents the wide range of individual preferences in highly differentiated societies. Direct democratic decision processes may in contrast be used by the citizens of a democracy on the basis of specific cases, and according to pattern prediction, they are most likely to lead to preference-based and thus improved political results.

Notes 

 (1) On this topic, see the overview by Butler/Ranney (1994).

(3) See Blankart (2000), p. 607ff.; Hartleb/Jesse (2005).

(4) Heußner/Jung (1999), p. 13; Kost (2005).

(5) See for example Bosbach (2002).

(6) See Schiffers (2002), p. 65ff.

(7) See Zimmermann/Just (2000); Kotte (2004), p. 141ff.

(8) See Wittmann (2001), p. 55ff.

(9) See Lupia/Matsusaka (2004), p. 464; Kotte (2004), p. 71ff.

(10) On this topic, see Lupia (1994a) and Lupia (1994b).

(11) See also Buchanan/Tullock (1965/1999).

(12) See Kotte (2004), p. 71ff.

(13) See Williamson (1990), p. 54.

(14) See Jensen/Meckling (1976), p. 308.

(15) See also Feld/Savioz (1998), p. 42ff.

(16) See Bohnet (1997), p. 146.

(17) See Lehner (1981), p. 26.

(18) See Kirsch (1997), p. 219.

(19) See Downs (1968), p. 210.

(20) See Wittmann (2001), p. 55f.

(21) Kirsch (1997), p. 215.

(22) See Downs (1968).

(23) On empirical evidence, see Benz/Stutzer (2004).

(24) On this topic, see Kirsch (1997), p. 220ff.

(25) See Kirchgässner/Feld/Savioz (1999), p. 54f.

(26) See Kirchgässner/Feld/Savioz (1999), p. 54.

(27) See Eichenberger/Frey (1994), p. 783.

(28) Lupia (1994a) und Lupia (1994b) support this claim by experimental and empirical evidence.

(29) See Adamovich/Wohlgemuth (1999), P. 131.

(30) See Frey/Kirchgässner (1993), p. 140.

(31) See Kleinewefers (1997), p. 68.

(32) See Glaser (1997), p. 122ff.

(33) See Eichenberger/Frey (1994), p. 783.

References

Adamovich, I.B., M. Wohlgemuth (1999): "Exit" und "Voice" im Systemwettbewerb: Das Zu¬sam¬men¬wirken von Föderalismus und direkter Demokratie in der Schweiz, in: Streit, M.E., M. Wohl¬gemuth (Hrsg.): Systemwettbewerb als He¬¬rausforderung an Politik und Theorie, Baden-Baden, S. 123-151.

Benz, M., A. Stutzer (2004): “Are Voters Better Informed When They Have a Larger Say in Politics? - Evi-dence for the European Union and Switzerland”, in: Public Choice 1-2, Vol. 119, S. 31-59.

Blankart, C.B. (2000): “Wie könnten Abstimmungen auf Bundesebene geregelt werden?“, in: Wirtschafts-dienst 10, S. 607-611.

Bohnet, I. (1997): Kooperation und Kommunikation; eine ökonomische Analyse in¬di¬vi¬dueller Entscheidun-gen, Tübingen.

Bosbach, W. (2002): Mit der Staatsform spielt man nicht - Gegen eine plebiszitäre Um¬gestaltung unserer Demokratie, in: Wirtschaftsdienst 9, S. 518-521.

Buchanan, J.M., G. Tullock (1965/1999): The Calculus of Consent, Logical Foun¬da¬tions of Constitutional Democracy, Ann Arbor.

Butler, D., A. Ranney (1994): Referendums around the World. The Growing Use of Direct Democracy, Washington.

Downs, A. (1968): Ökonomische Theorie der Demokratie, Tübingen.

Eichenberger, R., B.S. Frey (1994): Bessere Politik durch Föderalismus und di¬rek¬te De¬mokratie, in: C. Hermann-Pillath, O. Schlecht, H.F. Wünsche (Hrsg.): Markt¬wirtschaft als Aufgabe: Wirtschaft und Gesellschaft im Über¬gang vom Plan zum Markt, Stuttgart, S. 773-787.

Feld, L.P., M.R. Savioz (1998): Vox Populi, Vox Bovi? Ökonomische Aus¬wir¬kung¬en di¬rekter Demokratie, in: Grözinger, G., S. Panther (Hrsg.): Kon¬sti¬tu¬tio¬nelle Öko¬nomie - Sind unsere gesellschaftli-chen Regelsysteme in Form und gu¬ter Verfassung?, Marburg, S. 29-81.

Frey, B.S., G. Kirchgässner (1993): Diskursethik, Politische Ökonomie und Volks¬ab¬stim¬¬mungen, in: Ana-lyse und Kritik, Nr. 15, S. 129-149.

Glaser, U. (1997): Direkte Demokratie als politisches Routineverfahren: Volks¬ab¬stim¬mung¬en in den USA und in Kalifornien, Erlangen/Jena.

Hartleb, F., E. Jesse (2005): Direkte Demokratie in der Bundesrepublik Deutschland: Positionen und Kon-troversen“, in: Massing, P. (Hrsg.): Direkte Demokratie – eine Einführung, Schwalbach/Ts., S. 13-39.

Heußner, H.K., O. Jung (1999): Einleitung, in: Heußner, H.K., O. Jung (Hrsg.): Mehr di¬rekte Demokratie wagen, Volksbegehren und Volksentscheid: Geschichte - Praxis - Vorschläge, München, S. 11-23.

Jensen, M.C., W.H. Meckling (1976): Theory of the Firm: Managerial Behavior, Agen¬cy Costs and Owner-ship Structure, in: Journal of Financial Economics 3, S. 305-360.

Kirchgässner, G., L.P. Feld, M.R. Savioz (1999): Die direkte Demokratie - Modern, er¬folgreich, ent-wicklungs- und exportfähig, Basel/Genf/München.

Kirsch, G. (1997): Neue Politische Ökonomie, Düsseldorf.

Kleinewefers, H. (1997): Die direkten Volksrechte in der Schweiz aus öko¬nomischer Sicht - ein Überblick, in: Borner, S., H. Rentsch (Hrsg.): Wieviel direkte De¬mo¬kratie verträgt die Schweiz?, Chur, S. 61-93.

Kost, A. (2005): Direkte Demokratie in den deutschen Ländern: eine Einführung, Wiesbaden.

Kotte, S. (2004): „Politische Transaktionskosten in der Demokratie“, in: Folkers, C. (Hrsg.): Kollektive Entscheidungen, Wirtschaftspolitik und Öffentliche Finanzen, Band 13, Frankfurt a.M. u.a.

Lehner, F. (1981): Einführung in die neue politische Ökonomie, Königstein/Ts.

Lupia, A. (1994a): “The Effect of Information on Voting Behavior and Electoral Outcomes: An Experimen-tal Study of Direct Legislation”, in: Public Choice 78, S. 65-86.

Lupia, A. (1994b): “Shortcuts versus Encyclopedias: Information and Voting Behavior in California Insur-ance Reform Elections”, in: American Political Science Review 88, S. 63-76.

Lupia, A., J.G. Matsusaka (2004): “Direct Democracy: New Approaches to Old Questions”, in: The Annual Review of Political Science 7, S. 463-482.

Schiffers, R. (2002): “´Weimarer Erfahrungen`: Heute noch Orientierungshilfe?“, in: Schiller, T., V. Mit-tendorf (Hrsg.): Direkte Demokratie. Forschung und Perspektiven, Wiesbaden, S. 65-76.

Williamson, O.E. (1990): Die ökonomischen Institutionen des Kapitalismus, Tü¬bing¬en.

Wittmann, W. (2001): „Direkte Demokratie – Bremsklotz der Revitalisierung“, Frauenfeld, Stuttgart, Wien.

Zimmermann, K.W., T. Just (2000): Interest Groups, Referenda, and the Political Pro¬cess: On the Effi-ciency of Direct Democracy, in: Constitutional Political Eco¬no¬my 11, S. 147-163.