Markets in votes: Alienability,strict secrecy,and political clientelism

Introduction

There are two standard rationales supporting bans on markets in votes. According to the first one, markets in votes would introduce distortions in the political representation of citizens’ views due to their differences in wealth. In electoral terms, markets in votes would make the voice of the rich louder at the expense of the poor’s. According to the second one, the commodification of votes would weaken individuals’ commitment to promoting the common good by distorting their sense of civic responsibility. Due to the liberty to sell one’s vote, voting itself would come to be understood as a right granted to the individual for his own gain. Discussions on these matters have tended to ignore the experiences of societies where informal forms of vote buying are widespread. Those lessons suggest that the fundamental concern with their legalization would be necessarily related neither to the wealth-based distortions of the ideal of equal voice nor to the psychological impact of the introduction of monetary considerations. The fundamental concern would rather arise from the array of undesirable consequences that a sustained reliance on clientelistic exchanges is likely to have on the quality of democratic governments. Those clientelistic exchanges would result from the mere granting to individuals of a legal liberty to alienate their votes.

The interest in markets in votes might be thought to be largely driven by theoretical considerations. The interest would lie in providing a clarification of the normative basis upon which bans on markets in votes rest. Along the inherent value in providing such an account, there is the value of integrating our understanding of other related issues.

First, there is the relationship between the legal prohibition on vote selling and the presumption of voter liberty behind the wide legal discretion assigned to individuals in the use of their own votes. Brettschneider (2007: 24) argues that markets in votes ‘would not allow individuals to make their own autonomous decision about politics’. Freiman suggests the opposite (2014: 762). If we believe that voting is not purely justified on instrumentalist concerns, and that it would be undemocratic to force a citizen to exercise his right to vote in any particular way, Mill’s question seems at least to be pertinent, ‘…on what ground can we blame him for selling it, or using it to recommend himself to any one whom it is his interest to please?’ (1861: 489).

Second, there is the normative relation between bans on markets in votes and other common practices of democracies. Two particular matters are worth highlighting. There is first the question of the coherence between those bans and other structurally similar practices of democracy that are commonly accepted. They include campaign promises, campaign finance contributions, logrolling, and earmarking. There is also the relationship between the normative basis for bans on markets in votes and the secret ballot. Due to his conception of voting as a trust rather than as a moral right, Mill argued that ‘the duty of voting, like any other public duty, should be performed under the eye and criticism of the public’ (1861: 490). This is because secret voting suggests to the voter that ‘suffrage is given to himself; for his particular use and benefits’ (p. 488). ¹ Mill’s argument against the secret ballot is thus no different than one of the standard rationales against markets in votes.

Third, it is tempting to rely on the widespread acceptance of the illegality of markets in votes as a starting point for thinking about the desirability of extending markets to other domains. It is worth examining whether this is justified. Our initial concerns might be the same, such as the potential distributional effects and impact of introducing monetary incentives in domains mainly regulated by the existence of social norms. Our final judgments might, however, rest on alternative grounds. The literature on the limits of markets over contested commodities is largely concerned with the potential consequences of ‘market’ rather than ‘non-market’ inalienability. But the fundamental concern with markets in votes might arise directly from the latter type of alienability. Discussions on the limits of markets over contested commodities are also mostly concerned with specifying the limits of individuals’ moral claims to exchange what we tend to recognize as theirs. In dealing with markets in votes, however, part of what is in contention is the precise meaning of the ownership relationship individuals’ have over their votes.

Finally, it is common to conceive of the undesirability of political clientelism as largely derived from the wrongness of vote buying. It is worth considering whether the relation is the inverse.

I offer a reconstruction of the two standard rationales against markets in votes in the first section. In the second one, I develop a positive argument for markets in votes that has not yet received attention. According to this argument, markets in votes would align voters’ economic incentives and the fulfillment of their civic obligation to abstain from voting when this is called for. Assessed against this argument, the two standard rationales might not be deemed sufficiently robust. In the third section, however, I argue that the incentive alignment argument and the two standard rationales share the same failing. This is a failing to appreciate the implications of the mere alienability of the right to vote for the prevalence of clientelistic practices. In the conclusion, I highlight the implications of this perspective for the related issues noted earlier.

The two standard rationales

The legitimacy of the democratic process might not require strict equal political influence. But it is generally thought to require a distribution of influence that is insensitive to normatively irrelevant factors. Those factors would include wealth, race, gender, or religious beliefs. It would be acceptable that individuals’ capacity for political influence is unequal if only due to their greater persuasive skills, for example. For many, however, markets in votes would yield an unequal distribution of political influence between groups of voters based on one of those morally irrelevant factors: wealth. ² This is because votes could be regarded as a normal good, that is, a good for which demand increases when income increases. Heterogeneity in income would then imply heterogeneity in individuals’ valuations of votes and room for trade. The source of the expected unacceptable unequal distribution of political influence would thus come both from the wealthy’s greater willingness to buy votes and from the poor’s greater willingness to sell theirs.

As Downs notes in his seminal analysis of democracy, every rational voter would have a low reservation price on his own individual vote due to the indecisiveness of voting characteristic of large-scale elections (1957: 189). Among wealthy voters, however, such reservation prices would tend to be higher. Due to their other most urgent material needs being already satisfied, they would place more value on the psychological satisfaction that voting provides relative to poor voters. Poor voters could thus supply votes at a lower cost to themselves than rich voters could. Wealthy voters could buy votes at a lower cost to themselves than poor voters could. Although the low reservation price of both poor and wealthy voters does not entail that the prices of votes must also be low, it is reasonable to suppose that this would be so. Vote sellers would find it difficult to coordinate among themselves, and those who hold on their votes for too long might get nothing in return once the demand for an electoral majority is met (Levmore, 2000: 123; Taylor, 2016b: 909–910). For a much smaller proportion of the well-off, the potential forgone monetary benefit would thus be high enough to move them to sell their own votes. We could thus expect the wealthy to be overrepresented in the demand side of the market and the poor to be overrepresented in the supply side.

The poor could only sell their votes to those with whom they identify. The ability to sell their votes, we might think, could then only give them an opportunity to benefit that they currently lack. Alternatively, the poor could be willing to receive smaller payments for their votes when selling them to those whom they identify. Although perhaps more willing to sell their votes, we could claim that the poor’s underrepresentation should be thought to be a matter of their own making. A market in votes, however, would create economic incentives against carrying out those previous intentions. Since an individual poor voter could not control what other voters do, it would make more economic sense for him to sell his vote to the highest bidder while perhaps hoping that others who share his political views do not do the same. A market in votes is thus likely to place poor voters into a multilateral Prisoner’s Dilemma game with respect to the sale of their political voice. ³

If the conditions of that multilateral Prisoner’s Dilemma obtain, the usual considerations supporting markets would fail to hold. Those who would decide to sell their votes could not be understood as offering their genuine consent or agreement to the policies proposed by those who would buy them. Neither could it be possible to deduce any sort of value from either the mere voluntariness of the vote exchanges or the increased options that voters would face. If those conditions obtain, we could not infer a preference for having the legal liberty to trade from the mere willingness to trade when having such liberty. Taylor (2016a, 2017b, 2017c) has analyzed these implications in detail. Those same reasons suggest why a market in votes could fail to capture the intensity of voters’ political preferences. ⁴ For voters’ decisions might not represent the strength of their preferences but rather a mere consequence of the indecisiveness of their own individual choices. What many voters might prefer is merely not to express their true political preferences when there is no prospect of gaining any benefit from doing so.

The second standard concern supporting the illegality of markets in votes focuses on the consequences that the commodification of votes would have on citizens’ understanding of their civic responsibility. The idea is that it would encourage voters to view their votes as a means to pursue their own private interests, rather than to express their views concerning the public good. ⁵ The most plausible version of this concern might take an instrumental character. The concern would be about how the alleged effect on citizens’ understanding of their civic responsibility affects the quality of democratic choices. Sunstein suggests the following, for example: ‘If votes were freely tradable, we would have a different conception of what voting is for – about the values that it embodies – and this changed conception would have corrosive effects on politics’. (1994: 849)

The rationale for the ban of markets in votes from the undue influence of wealth on voting decisions tends to be based on fair proceduralist grounds. ⁶ The justification of democratic voting would rest, at least partially, on its property of being a procedure that provides equal treatment of voters. The rationale from the undue influence of self-interest relies rather on a substantive conception of democratic voting. The purpose of voting would be to allow us to express our views about the public good. Yet markets in votes would make it harder for citizens to conceive of voting in those terms. Although substantive, this view is central to a number of democratic theories. ⁷ More importantly, it is not particularly controversial when understood in an austere sense as implying the existence of certain basic duties in the act of voting. ⁸ Our electoral choices have the potential to generate consequences that affect others. We tend to see those types of choices as constrained by moral considerations that include some sort of impartial weighting of the relative importance of all the interests involved.

The plausibility of this second objection against markets in votes does not rest exclusively on the plausibility of its underlying substantive conception of democratic voting. It also rests on the plausibility of a psychological claim. This psychological claim would seem to ground objections to extensions of markets in other domains. The claim is the following: Allowing the option of securing a monetary benefit by the foregoing of duties of charity, care, assistance, or a civic duty in the case of voting might have a negative impact on people’s attitudes toward the proper discharging of such moral obligations.

Gneezy and Rustichini’s (2000) field study in day care centers might provide some evidence in support of that claim. The result of the introduction of a monetary fine for late-coming parents was a steady increase in the number of parents coming late. One plausible explanatory hypothesis is that the fine changes the agents’ understanding of the social norms at work (pp. 13–14). The fine is understood as a mere price that one is free to pay in exchange for the service provided by the day care after closing time. The guilt and shame felt for taking advantage of other people’s time is no longer felt. The sense of obligation that those emotions might ground is thus weakened as a result of the introduction of a monetary price. A monetary price high enough would certainly reduce the number of parents coming late. The issue is, however, that the mere introduction of monetary prices might crowd out the considerations behind the shadow or implicit prices that may govern some of our social interactions in nonmonetary settings. As Sandel (2012: 42–43) points out, the lesson is that the stigma associated with violating a social norm might sometimes be worth preserving.

Markets in votes would add an additional cost to the act of voting: the forgone monetary benefit that voters could gain by selling their votes. Granting individuals the liberty to sell their votes would amount to asking them to pay an additional price for exercising their own right to vote. Those who take this option, that is, those who choose not to sell their votes, might then face a comparable psychological mechanism to the one illustrated by the day care study. Their decisions to vote rather than to sell their votes could generate a sense of entitlement over their right to vote. The psychological pressure to exercising it with a view on the common good could thus be diminished as a result. Admittedly, psychological mechanisms of this kind are rarely universal. Not all voters might perceive the forgone monetary benefit of voting in the same way that they perceive regular prices. As Levmore (2000) notes, it is also conceivable that for some individuals ‘vote trades might focus deliberation rather than supplant it’ (p. 11). These qualifications, however, do not seem to be sufficient to disregard the underlying concern.

The two standard rationales for the ban on markets in votes have been recently challenged. Freiman (2014) and Taylor (2016c) have argued that the rationale regarding the undue influence of wealth is insufficient to justify a prohibition in markets in votes. Their answers are both grounded on the same idea. An unconstrained market in votes might very well clash with a basic procedural principle of democracy by making poor voters worse off in terms of the representation of their political voice. There are, however, multiple other paths through which wealth might have an impact on that representation under current electoral systems that ban vote trades. Campaign contributions is one of those paths. The concern for the undue influence of wealth in democracies is not taken to justify the ban on those contributions because of the availability of relatively effective regulatory mechanisms. Therefore, the argument against markets in votes would be incomplete at best. Its argumentative force would be contingent on establishing the unavailability of those mechanisms for markets in votes. Both Freiman and Taylor suggest that this cannot be shown. ⁹

Freiman has also questioned the rationale from the erosive effects on civic virtue or responsibility. In Freiman’s view, this rationale conceives the protection of civic virtue as one of the functions of a legal order (2014: 769–773). This would be problematic insofar as it seems to imply objectionable abridgments of equal suffrage. We might believe that concerns for the potential intrusiveness or stigmatization involved in evaluating voting positions could explain why that function does not imply such abridgments. Freiman notes, however, that votes that fail to satisfy some standard of the common good could be simply disallowed or disregarded. This procedure ‘would not stigmatize, excessively coerce, or intrude upon citizens’ (2014: 772). Nevertheless, Freiman notes that ‘many democrats would, at a minimum, be uneasy with this restriction, if not outright opposed to it’ (p. 772). Since our concerns for the erosion of civic virtue is an inadequate basis for the justification of legal restrictions to the right to vote, Freiman thus concludes that it is an inadequate basis for the justification of the illegality of market in votes.

Well-established practices of contemporary democracies, such as campaign contributions and unrestricted voting, are in tension with concerns regarding the undue influence of wealth and self-interest in voting decisions. In those cases, however, neither outright bans nor criminal prosecutions are seen as appropriate. The rationales for bans on markets in votes that rely on those concerns should then be judged to be inadequate. This is the essence of the abovementioned challenge to the two standard rationales against markets in votes. Its main weakness rests on its failure to identify a collective interest that would be promoted by the legalization of markets in votes.

The illegality of markets in votes does not preclude the influence either of money or of self-interest considerations in politics. The extent to which such undesirable influences are tolerated, however, might be best understood as the inevitable cost of protecting other fundamental values. ¹⁰ Campaign contributors could be banned only at the expense of free speech. The self-interest effects of unrestricted voting could be mitigated only by assuming the risk of granting government the power to assess the qualifications and merits of voters (or of the value of their votes). In this regard, contrary to Freiman’s suggestion, the illegality of markets in votes would not seem to contradict the demands of democratic legitimacy even if understood in robust terms. This is because a ban of vote markets does not preclude citizens’ judgments to be taken into account as inputs in the collective decision procedure. At most, it constraints the form in which such judgments must be expressed. ¹¹ Therefore, justifying a ban on vote exchanges by appealing to their expected erosive effects on civic virtue would not commit us to restricting voting.

Regulatory mechanisms could be conceived to mitigate the influence of wealth in markets in votes. They would not come without costs, however. Those costs would include the costs involved in the mere provisions of such mechanisms. They would also include the costs associated with the imperfect nature of all regulatory measures along the specific domains over which they apply. For this reason, the availability of relatively effective mechanism to curb the influence of wealth within markets in votes does not entail the constancy of the overall influence of wealth in electoral outcomes. If democratic legitimacy is not at stake, it is not initially clear what would make the previous costs, along the risks of affecting individuals’ sense of civic responsibility, worth incurring. ¹² When it comes to voting, markets are needed neither to fix output levels nor to minimize the undesirable consequences of widespread illegality.

The capacity of markets in votes to reflect the intensity of voters’ preferences is often considered as its main potential benefit. However, this capacity is questionable due to the collective action structure of electoral outcomes implied by the indecisiveness of each individual voter. Such structure would not be altered by regulatory measures oriented to curb the influence of wealth. Furthermore, the value of capturing preference intensities would be ambiguous at best. The intensity of citizens’ preferences rarely correlates with their deeper understanding of the problems at hand or with their dispassionate assessment of the relevant moral matters. Markets in votes would allow voters to benefit monetarily from the sale of their own votes. These monetary benefits would not seem to have a significant moral weight, however. This is because voters do not have moral entitlements to their votes in the same form they might have them to their bodies or to the security of their lives and property. As Mill observed, the right to vote entails a right to power over others (1861: 488). Such right is granted to individuals only due to the moral demands that come with the political organization of society. Individuals would then not have fundamental claims to the alleged private benefits that an alternative voting system might confer upon them independently of the resulting impact on that political organization.

The incentive alignment argument

The challenge by Freiman and Taylor to the two standard rationales for prohibiting markets in votes reviewed earlier was found wanting. But it was found wanting on the basis of its alleged failure to identify some significant collective interest compromised by the legality of markets in votes. It is then an open question whether a case against markets in votes grounded on the two standard concerns is robust enough. This is particularly clear once we appreciate that equal political voice and the exercise of civic responsibility are not the only considerations worth appealing to when evaluating democratic procedures. ¹³ The identification of a currently under-appreciated collective benefit from the implementation of markets in votes could thus unveiled a positive impact along an alternative normative dimension. This section presents a case for markets in votes along those lines. It conceives a market in votes as a mechanism to mitigate the effects of voter incompetence on the quality of democratic decision-making. The alleged unappreciated value of markets in votes would thus rest along the epistemic dimension. The significance of this dimension is widely recognized by several theories of democracy. ¹⁴

The manner in which markets in votes could mitigate the effects of voter incompetence is suggested by Downs’ analysis of the causes and effects of voting abstention (1957: 260–276). Downs notes that we can expect voters to be motivated to some extent by a sense of social responsibility relatively independent of their own short run and losses. This is the value of voting per se. Its magnitude depends in part by ‘how much [a voter] values living in a democracy’ (p. 270). The value of voting per se is one part of the return from voting. The other part of the return from voting comes from the strength of a voter’s desire to see one party win instead of others. Voters tend to discount both values to different degrees to allow for the influence of other voters. In particular, voters consider how close the election would be and how many other citizens would vote. Information plays a crucial role in voters’ expected returns from voting. Information can be taken to reduce voters’ uncertainty in terms of their confidence over either the intrinsic or instrumental value of their policy preferences, their expectation that such policies will be pursued by their preferred party, and their estimation of the likely behavior of others. The more informed a voter is the less he would tend to discount his estimated return from voting correctly. Voters with lower levels of information would tend to face lower returns from voting. Since voting is inherently costly, the less-informed voters would then abstain at a higher rate due to their lower returns from voting.

If voters were to have the option to gain a monetary benefit by selling their vote, the costs of voting would increase. Voters would have to incur the opportunity costs of their time but also of the monetary benefit that is forgone by not selling their votes. Those voters we would not expect to vote in the presence of a vote-buying ban would now have a reason to sell their votes rather than merely to abstain. Their direct electoral impact would remain the same. A market in votes, however, would reduce the electoral impact of those voters who face marginally superior returns from voting. For these voters, the returns of voting are just high enough to offset the current costs of voting. A small increase in those costs, such as the ones entailed by the opportunity to sell their votes, would move them to refrain from voting. ¹⁵ If we assume that prices for votes would tend to be low (or if the market in votes were somehow regulated as to avoid that competition among those in the possession of vote-buying capital funds were to drive the prices of votes well above the reservation price of those marginal voters), we should then expect an increase in the political competence of the median voter relative to the scenario of a vote-buying ban.

The argument has been formulated by an appeal to the role information plays in the electoral participation of eligible voters. It need not be understood exclusively in those terms, however. Some voters might have an inclination to discount future aggregate, non-salient costs and to assign greater value to immediate tangible benefits. ¹⁶ A market in votes might grant voters with such temporal preferences, and who would tend to vote for misguided present-oriented policies, an attractive option to exit from the collective decision. Furthermore, those who assign very little value to voting per se due to their low levels of appreciation of the value of living in a democracy would also be induced not to take part in elections at a higher rate. This would be a positive result under the assumption that those individuals tend to exhibit low levels of civic virtue in their voting choices. Finally, it is questionable whether low levels of information must necessarily yield higher discounting rates in calculating the returns from voting. For some group of voters, access to additional information might affect their sense of identity in a negative way. This is part of the explanation of voters’ tendencies to engage in motivated reasoning and cast their votes on purely expressive grounds. ¹⁷ The availability of a monetary benefit would, however, increase the opportunity costs of voting on those grounds.

Insofar as it is generally agreed that widespread political ignorance, low levels of civic commitment, and biased voting are undesirable aspect of contemporary democracies, this argument does not rest on a controversial normative goal. This would be so regardless of how much weight we place on the epistemic function of democracy. It is also worth noting that the argument does not need to deny the instrumental value of efforts to reduce political ignorance, lack of commitment, and bias through education or other means. A market in votes could be advanced simply as a mechanism to reduce the impact of those aspects of political decision-making that are regarded as being beyond the reach of current or feasible educational efforts. In that sense, this epistemic argument would also be compatible with a general positive appreciation of the values of political participation and universal suffrage.

Electoral systems already grant individuals the option of not taking part in elections either through non-compulsory voting systems or through the issuing of blank votes. It is reasonable to think that individuals who lack some minimal knowledge regarding the issues involved should make the required corrective efforts foremost, but abstain from voting if unable to do so. Markets in votes could thus be seen as providing the appropriate alignment between voters’ private incentives and the fulfillment of that civic obligation. ¹⁸ The incentive alignment mechanism would thus allegedly increase the quality of democratic decisions without granting the government a potentially discretionary power to limit the right to vote. Analyzed in terms of this epistemic case for markets in votes, the two standard rationales for their illegality might not be sufficiently robust.

As Downs noted, low-income groups in society are likely to have less political power than their numbers warrant due to their greater propensity to abstain. The cost of voting is harder for them to bear. The cost of information is also harder for them to bear, and this entails that their returns of voting are also likely to be lower due to the greater uncertainty they face. A ban on vote markets would not eliminate this source of bias against the representation of low-income citizens. It would merely restrict it. The case for markets in votes does not need to hold that the restricted magnitude is morally insignificant. This is because the reduction in bias against low-income groups is gained at the expense of a reduction in the relative impact of political competence in electoral outcomes. Such a reduction might entail potentially high costs on others or a depreciation of the overall epistemic value of democracy. It is then not evident that they are worth paying for the sake of a lower (but not zero) representational bias against low-income groups. This argument could only be strengthened if we recognize the availability of relatively effective regulatory solutions to mitigate wealth effects within markets in votes. ¹⁹

Similarly, the case for markets in votes can hold even if such markets have a significant negative impact in individuals’ understanding of their civic responsibility. The standard concern is grounded on the expected effects of commodification in augmenting the role of self-interested considerations in voting decisions. If this concern is understood in instrumental terms, what ultimately matters is how self-interested considerations affect the quality of democratic outcomes. According to the incentive alignment argument, those self-interest considerations might also have a positive effect along that same dimension. By increasing the costs of voting, ‘low-quality’ voters would self-select themselves out of the electoral process due to the lower returns they expect from voting. The effect of their voting decisions would thus be nullified. Conditional on the magnitude of both types of effects, we might then be justified in increasing the role of self-interested considerations in voting decisions if the benefits due to the increase in the political competence of the median voter are large enough.

Alienability and strict secrecy

The two standard rationales for the illegality of markets in votes might not be sufficiently robust to answer the incentive alignment argument. In this section, however, I argue that both the incentive alignment argument and the two standard rationales fail to appreciate the consequential significance of the mere alienability of the right to vote. This alienability would lower the transaction costs of certain political exchanges that are widely regarded as either intrinsically or instrumentally disvaluable. This concern was previously suggested by Epstein (1985, 1995) and dismissed by Hasen (2000) and Levmore (2000). Epstein did not provide a clear articulation of its underlying logic, however. ²⁰ In particular, he did not offer an account as to how the alienability of votes would relax the strict secrecy of voting and the importance of this effect. ²¹ The underlying assumption behind discussion on markets in votes is that they would be compatible with the secrecy of voting. This is because the commodity to be exchanged would be the right to issue a vote. In order to accomplish this, electoral authorities could issue tradable voting coupons to all registered voters. The act of voting would thus remain unobservable to others. This is, however, only one of the two dimensions of voter secrecy that is secured by the secret ballot. The strict secrecy of voting is a function of the two dimensions.

The secret ballot is supposed to break the connection between the identity of the voter and his vote choice. To do this, it prevents others from observing the content of the vote. But the secret ballot also prevents voters themselves from proving to others how they vote. This second dimension of secrecy is secured by its mandatory status. Under perfect enforcement of strict secrecy, it is thus impossible for an agent to communicate credibly to others how he acted since the content of the vote is unobservable to them, and since there is no means he can rely upon to reveal such content. Under strict secrecy, there is no incentive to engage in intimidatory efforts. Schelling (1980) states it well:

when the voter is powerless to prove that he complied with the threat, both he and those who would threaten him know that any punishment would be unrelated to the way he actually voted. And the threat, being useless, goes idle. (p. 148)

As already noted, a legal market in votes does not need to compromise the first dimension of secrecy. The actual content of votes need not be observable to others. A market in votes, however, would facilitate voters proving others whom they support. This is because voters would have the power to transfer their voting coupons to those who ask for them on behalf of a given party. Equally significant, the refusal to engage in this exchange would be taken as an indication of their unwillingness to support the party in question. ²³ Individuals could be offered the option of entering into anonymous forms of vote trading. The central issue, however, is whether they could be protected from revealing the information implicitly contained in their decisions when faced with offers to trade. Thus, regardless of the presence of such legal option, as a matter of practice, the alienability of votes would entail a significant weakening of the strict secrecy of voting.

In the absence of effective protections of strict secrecy, we should thus expect markets in votes to increase voters’ vulnerability to intimidation by reducing the uncertainty regarding their voting intentions. The criticism to markets in votes from the undue influence of wealth points out to the distorting effects that access to capital funds would have on the electoral process. What would matter in this particular regard would be rather the access to an organized apparatus of intimidation, such as those that could be provided by different types of organizations or the same institutions of the state. Governments are rarely able to sustain themselves through intimidation alone, however. The costs of enforcement and the prospects for instability are both significantly lower in the presence of wide loyal support. Yet the legalization of markets in votes would also increase the ability of political actors to enter into a particular undesirable type of exchange. This is the exchange of targeted government benefits in return for political support. By having the ability to transfer their votes publicly, voters would acquire the ability to commit their support to particular political candidates in a credible manner. The transaction costs of clientelistic bargains would thus be significantly reduced. Clientelistic strategies would thus yield larger electoral returns relative to alternative strategies of electoral persuasion.

Anderson and Tollison (1990) argue that markets in votes allow the rich to invest in rent protection by purchasing large blocks of the votes from lower income citizens. In doing so, the wealthy minority are less vulnerable to the depredations of the non-wealthy majority ‘because the wealthy could buy a (temporary) majority as needed’ (p. 293). This particular line of reasoning, as well as the more general concern for the undue influence of wealth based on procedural grounds, rests on a lack of appreciation of how the introduction of markets in votes changes the options faced by all. The wealthy might be willing to allocate great sums in vote buying as a means to advancing policies that protect or advance their own interests. Parties interested in capturing votes from poor voters, however, are no longer constrained by their current endowments. Populist leaders could improve on the offers of the rich through promises of future payments financed by the expropriation of their assets. This is an offer that the rich would have no financial interest in fully matching. ²⁴ Promises of future payments are always uncertain. However, those who gain power gain access to both a stream of income and a large pool of public offices that could compensate for that uncertainty. As in all sequential exchanges, parties’ promises to reciprocate to voters with targeted benefits could always go unfulfilled. As in all repeated sequential exchanges, however, the cost of unfulfilling one’s promises is the loss of credibility and, in turn, a reduced power of making promises. ²⁵

The plausibility of the incentive alignment argument offered in the previous section also overlooks that alteration of the strategic setting faced by political actors as a result of the alienability of votes. That argument holds that voters who lack adequate political knowledge or dispositions would rather sell their votes even when their monetary gain is low. However, the transferability of votes implies that political actors’ strategies might now include promises of significant targeted future benefits such as public employment or access to subsided housing. Those particular voters might as a result find no reason to self-select away from the electoral process for small present payments. Political actors’ strategies might also include intimidatory threats toward voters when unwilling to sell their votes. In this case, there is no guarantee that the voters who would ‘self-exclude’ from the electoral process belong to the set assumed by the incentive alignment argument. There is then no reason to believe that the alleged benefits advanced by the incentive alignment argument in terms of improving the quality of democratic governance would be realized. On the contrary, the greater enforceability of clientelistic political exchanges would seem to yield a series of undesirable consequences along that same dimension.

Regulatory measures such as spending caps and proportional allocation quotas for vote buying funds as suggested by Taylor (2016c) would not address the problem at hand. This is because the fundamental issue is related to the alteration of the strategic setting that political actors would face due to the weakening of strict secrecy that the mere alienability of votes implies. In other words, contrary to the focus on commodification might make us believe, these problems would still arise even if only donations of votes were allowed. If voters were to have the simple legal ability to transfer their voting coupons to others, those others would have the ability to gain control of those votes by either promises of targeted future benefits or outright intimidation. Due to the low deterrent effects resulting from the immense challenges involved in policing the legitimacy of all voters’ exchanges, the illegitimacy of any such means would be expected to make little difference.

Institutional constraints in general might also fail to provide adequate checks to the detrimental consequences of political clientelism. A strictly defined sphere of public action, guaranteed by institutional mechanisms and protections, diminishes the incentives for investing in political influence. However, a properly confined area of public action is less likely to remain confined in the long term when transactions between political actors are easily enforced. Upon the mutual advantage of electoral transactions, important loyalties could easily be formed, increased, and maintained. In the presence of such loyalties, intimidatory efforts face lower costs due to the greater ease in identifying possible targets and the greater availability of enforcement mechanisms. Public rules of distribution of government benefits can also be subverted with greater ease. In extreme cases, government programs and public employment could be exclusively reserved to party members and those memberships instantiated by the continued transfers of individuals’ votes to the party authorities. In order to secure the long-term stability of the clientelism arrangement and the vested interests of all involved, there could be greater pressure and increased ability to relax inherited institutional constraints. The inalienability of votes might thus constitute an important although overlooked condition for the long-term efficacy of such constraints.

Experiences of clientelism

The lessons from societies with strong clientelistic political cultures support the previous theoretical considerations. No legal markets for votes exist in such societies, yet informal vote buying is widespread nonetheless due to their particular social and economic conditions. The regrettable implications of the alienability of votes drawn above in terms of the distributions of government benefits are in those societies frequently observed. The rule of law and the fundamental values that it protects are also usually weak where political clientelism is strong. Widespread clientelism is commonly thought to be a consequence of weak institutional systems. The weakness of such systems could, however, be partially explained by the widespread reliance on clientelistic practices that certain social and economic conditions allow.

In contemporary clientelistic societies, the uncertainty regarding voters’ choices force party leaders to rely on brokers. ²⁶ These brokers are community leaders who interact directly with voters. Although direct vote-buying sustained on fraudulent voting methods are not uncommon, indirect forms of vote-buying supported by informal mechanisms of enforcement are more significant. Some distribution of resources and favors to citizens in need is often made in advance with the expectation that the acceptance of such gifts will generate in the recipients either an obligation of reciprocity or the fear of being left-out of future rewards. The informational demands of this type of electoral campaigning are considerable since it is difficult to know who will take the benefits offered and fail to express their gratitude in the voting booth. The use of brokers is required by these informational demands. Their sustained and frequent engagement with voters allows them to have valuable local knowledge of both their needs and political inclinations. This personalization of the trade relationship between brokers and voters is also important for instilling voters with the sense of obligation that underlies gratitude-based strategies. ²⁷

Stokes et al. (2013) have thoroughly analyzed the implications of the extensive use of brokers required by political clientelism. Their central finding is that the costs implied by the need to rely on brokers go beyond the payments offered to them. In the same way in which voters are imperfect agents of brokers due to the imperfect monitoring created by the secret ballot, brokers are imperfect agents of party leaders. Brokers cannot be perfectly monitored either, and therefore, party leaders are uncertain regarding the impact provided by their services. The empirical evidence suggests that clientelistic politics does not give priority to swing-voters but rather mostly reward loyal supporters. ²⁸ This is explained by brokers’ status as imperfect agents. Since they are not easily monitored, and since their contribution in terms of the composition of the votes delivered is unobservable, brokers face incentives to increase their observable size of their networks by targeting loyal voters with greater emphasis if the party’s probability of victory is not greatly affected. These voters are expected to be cheaper, and brokers tend to garner larger rents in the process. ²⁹ All these are the agency costs party leaders incur when relying on brokers.

Another implication of the required extensive reliance on brokers is that the costs of clientelism tend to be higher in wealthy countries. In such countries, brokers face many other potential sources of income and this raises the costs of hiring them. The intensity and the frequency of the interactions that allow brokers to monitor voters are compromised as a result of urbanization and population growth. Brokers’ informational-gathering practices tend to be less effective. ³⁰ As poverty decline, brokers’ favors are more easily turn down. Due to the presence of all those increased costs, and due to the presence of reduced costs to mass communication, the electoral returns of clientelistic strategies are thus diminished in wealthy countries relative to the returns of more formal or ‘programmatic’ means of offering government benefits in exchanges for votes.

The legalization of markets in votes would, however, minimize the informational demands of current forms of clientelism due to the weakening of strict secrecy. The need for an army of local-level brokers would then be reduced as well, if not eliminated. A simple registry of vote purchases could be substituted for such a costly army needed to ensure vote buyers get the votes that they pay for. Local knowledge might no longer be needed, and the challenges that population growth and urbanization pose to brokers could be handled by modern forms of communication. Contrary to the constant or diminishing returns provided by an expending network of brokers, a simple registry of vote exchanges could face increasing returns to scale. Even if markets in votes would not make brokers unnecessary, their agency costs would in all likelihood be greatly reduced due to the greater ease in identifying the previous voting choices of their networks. One important implication is that the primary targets of clientelistic offers would no longer be mostly limited to party loyalists.

Under certain social and economic conditions, formal guarantees of strict secrecy are insufficient to reduce the returns of clientelistic strategies to the point at which their importance is negligible. However, the legality of markets in votes would dramatically lower the transaction costs of the exchanges upon which clientelistic strategies are based. Societies in where such political practices are still widespread would likely see an even greater reliance on them. Societies currently facing more favorable conditions for the effective observance of strict secrecy would likely fare no better. The corresponding legal guarantees to strict secrecy would be relaxed as a matter of practice. Due to the centrality of the enforcement mechanisms that such relaxation allows, the more favorable socioeconomic conditions of those societies might turn to be insufficient in the mitigation of the prevalence of clientelistic bargains. The increased downward pressure that modernization brings to the profitability of political clientelism would thus likely be eased relative to the main electoral strategies seen today in the developed world.

There is an element of truth in both standard rationales for the illegality of markets in votes. They would likely compromise basic fair proceduralist norms. They would also likely have a negative impact on citizens understandings of their civic responsibility. ³¹ The reasons why this is so would, however, be mostly related to the logic of political clientelism, rather than the differential wealth of voters or the psychological effects of the commodification of votes. In particular, it is worth considering that voters might understand both the detrimental consequences of widespread clientelistic practices and their corruptive influence on the legitimacy of the democratic process. Their individual electoral rejection of those practices, however, would have no consequential effects in the mitigation of those undesirable effects. Being individually powerless to mitigate those effects, voters would face powerful incentives to maximize their own private gains.

Conclusion

Contrary to the common denunciation of political clientelism by reference to the most basic illegitimacy of the exchanges upon which it rests, the case I have presented derives the illegitimacy of the latter from the undesirability of the former. This undesirability, in turn, seems to be normatively overdetermined. The intimidatory dimensions of political clientelistic cultures are in clear tension with the autonomy of voters. The granting of conditional benefits might erode the conditions required for civic deliberation. Clientelistic forms of redistribution could be seen either as involving straightforward infringements of property rights or as violating procedural norms of distribution. Recurrent confiscations of private resources for political gain result in lack of savings and investment, and thus of economic growth. Voters’ welfare is then expected to be diminished. The robust majorities that clientelistic strategies tend to generate jeopardize the efficacy and stability of institutional constraints on the use of power. Due to this normative overdetermination, contrary to what the two standard rationales suggest, the justification of a ban on markets in votes does not seem to be dependent on any particular conceptions of democracy. There is a plurality of concerns that would justify their current legal status.

It is now worth noting the most straightforward implications that follows from this argument for the related issues that I mentioned in the introduction. If we grant wide discretion to individuals’ exercise of their right to vote, on what grounds can we justify the legal ban on their decisions to sell it? Mill believed that there were no such grounds and concluded that voting should then be understood as a trust to be performed in public. Lever (2007) rejects Mill’s argument on the basis that all significant rights ‘are really a bundle of rights, rather than a single claim to do something’ (p. 362). The mere idea of a right to vote, even when understood in discretionary terms, would then not commit us to endorsing all possible separate components of the full bundle. This point is merely conceptual, however. At least for our purposes, Mill’s challenge is best understood in terms of the specific normative grounds for acknowledging a bundle constituted by a practically unrestricted right of use and a nonexistent right of disposition.

Lever also appeals to the familiar notion of inalienable rights and suggests that such rights are founded on the strength and moral or political importance of the interest they protect. That being so, ‘…the idea that people may not sell their vote is consistent with the idea that voting is a right, albeit a right that people should be able to renounce, though not sell’ (2007: 362). Yet it is unclear that the interest protected by the right to vote has the same strength and importance of the interests protected by some the typical examples of inalienable rights, such as the right not to be tortured or sold into slavery. It is also unclear why such an interest would establish the permissibly of renouncing to the right to vote, though not to selling it. Lever’s (2007, 2015) rejection of Mill’s argument for public voting is based on a plausible understanding of individuals’ interest in not subjecting their opinions to the criticism of others if they wish not to do so. In itself, this argument, however, is also insufficient to justify the mandatory status of the secrecy of voting. The argument I have presented provides an alternative account that avoids these problems.

An example is instructive in this regard. Until recently, the legal system in Pakistan allowed victims of attempted honor killings and their family members in case those victims did not survive, to forgive the accused relatives who committed the crime. A case against this feature of a legal system need not be based on the significance or strength of individuals’ interest in the very act of punishing their offenders. Due to the risks victims face if they choose not to forgive their aggressors, this additional legal liberty diminishes the protection, security, or value of their own other liberties as well as that of others by diminishing the deterrence effects of the legal system. The alienability of the right to vote could place individuals in a structurally similar situation. The protection of individuals’ interests that the right to vote provides would be diminished by granting them the liberty to alienate that right. This is because such alienability implies the weakening of strict secrecy, and in turn, the reduction of the transaction costs of intimidatory threats and clientelistic exchanges.

The reason why individuals should be able to abstain from voting though they should not be permitted to sell their votes is thus straightforward. The former legal permission does not alter the strategic setting they typically face, and the latter does. Due to the centrality of the weakening of strict secrecy in altering that strategic setting, this argument against markets in votes is in its essence an argument against ‘unveiling’ the vote. Mill’s argument for open voting based on the notion that voting should be performed under the ‘eye and criticism of the public’ (1861: 490) thus loses much of its appeal. Regardless of whether we believe individuals are simply entitled not to be forced to share their opinions with others, it is the strict secrecy of voting that would diminish the most the effectiveness of clientelistic exchanges and thus of voting for private gain. ³² As Brennan and Pettit (1990) note, the significance of self-interest voting should not be assumed to be significant within the context of a large-scale electorate due to the indecisiveness of each individual vote. In their view, the advantage of open voting rests rather on its increasing of the discursive pressures on individuals and on its decreasing of the safeguards on ill-considered or symbolically motivated voting. These alleged benefits of unveiling the vote would be realizable, however, only in the absence of the widespread clientelistic practices that, contrary to what Brennan and Pettit suggest, we might have reasons to expect.

The implications of this argument for our understanding of other democratic practices are also worth highlighting. Lippert-Rasmussen (2011) suggests that the central objections to markets in votes per se have no less force when applied to electoral promises per se. He argues that in most realistic conditions, vote buying is democratically problematic because of its effects on political equality and democratic deliberation. Under the same set of realistic conditions, several electoral strategies such as election promises may also both distort democratic deliberation and increase political inequality. This reasoning leads Lippert-Rasmussen to conclude that we should, ‘on pains of inconsistency’, adopt much sterner views on election promises (2011: 144). Similarly, Sandel (2010) infers the wrongness of earmarking, that is, the allocation of public funds by public officials with the intent of winning the electorate’s support, from the wrongness of vote buying (116–117). Freiman (2014) also notes the equivalence between vote markets and legislative vote trading or ‘logrolling’. Contrary to Sandel, Freiman suggests that since there are no clear grounds to make logrolling and earmarking criminal offenses, there are no clear grounds to make exchanges in votes a criminal offense either.

All these previous symmetry claims have been resisted on an alleged series of disanalogies between the practices in question. ³³ Regardless of their merits, the argument I have presented also provides a straightforward rationale for the different legal status of markets in votes and the other practices. While structurally similar, markets in votes are significantly dissimilar to campaign promises and earmarking in their capacity to alter the strategic setting faced by the relevant political actors. This is due to the crucial credibility-inducing mechanism of the former. Bans on markets in votes, on the other hand, do not face the difficulties involved in possible bans of campaign promises and earmarks due to the inherent ambiguity of such practices. Secret voting in legislative assemblies could minimize the opportunities for logrolling by diminishing the credibility to reciprocate through the uncertainty regarding the identity of the legislative voters. In doing so, however, secret voting would also minimize representatives’ accountability and preclude the potential functional benefit of the practice. ³⁴

Normative consistency might indeed call for the adopting of much sterner views at least on some of those structurally similar practices to markets in votes. There is a simple rationale for their dissimilar legal status with markets in votes, however. Consistency, on the other hand, does not seem to commit us in any clear direction when it comes to the more general discussion on the limits of markets in other contested commodities. This is because votes would be a very special type of commodity. The control over this commodity amounts to a capability to exercise power over others. The distribution of such power to all would be of value only when no power to alienate it whatsoever is distributed along. These special features could guide our thinking when it comes to certain specific others areas. The ban on markets in votes should not, however, be taken as a paradigmatic case of the limits of markets when it comes to several others.

I will end with a qualification. I have understood the likely operations of a market in votes based on the analysis of a simple system of tradable voting coupons, where voters are free to choose their trading partners. Within the confines of such a decentralized market, alternative institutional mechanisms could be conceived to prevent the erosion of strict privacy that the simple system of tradable coupons allows. The case against a decentralized markets in votes presented here is thus not entirely conclusive. But it would be a mistake to expect otherwise. The philosophical treatments of markets in votes have tended to ignore questions of institutional design when it comes to the institutional framework that they would require. Freiman notes this possible limitation of his own argument when writing that he addresses the in-principle permissibility of vote markets, without addressing the question, ‘of precisely how – or, critically, whether – to implement vote markets in practice’ (2014: 761). By asking the question of in-principle permissibility, we could be asking whether there are some conceivable conditions under which an institution is justifiable. In that sense, I have not shown that markets in votes are in principle impermissible. However, no plausible argument might yield such conclusion. In general, we tend to be concerned with assessing the permissibility of an institution within the much smaller set of realistic conditions we expect to face. This can only be fully done when the institution in question is conceived as being realized in some form of concrete design. For this reason, our negative moral assessment of that institution would be necessarily inconclusive relative to some other possible designs.