An Economic Theory of Democracy

Chapter 6 begins with the insight that uncertainty divides voters based on levels of information and the intensity of their convictions. Leadership is necessary precisely because voters rely on those more knowledgeable and motivated than themselves to enact policy. Uncertainty also enables some citizens to influence others who are less confident in their vote. Downs’s theory assumes that influencers will not lie, but might withhold information that may undermine their purposes. Other better-informed voters may choose not to influence others, and some better-informed citizens may not vote if they lack a strong preference. Others have no idea how to vote, have made only a tentative decision, or fail to see any meaningful difference among alternatives. Others still become used to voting for a party that has habitually served their interests, and so they eventually stop doing the research to check if that is still the case. The least informed are presumed to be those most open to persuasion; those who would bring someone into a higher category of confidence and involvement must have a sense of how much information is necessary to convince them. The only reason for expending the energy necessary for disseminating such information is that leaders expect their own position to improve with more people voting in line with their preferences. Political parties push for policies which they will implement, while interest groups advocate for the government to adopt policies which they believe will benefit themselves. Another group, “favor-buyers,” are transactional, influencing voters on behalf of a party only when a party promises them a specific benefit.

Uncertainty requires intermediaries to operate between government and citizenry. A government uses representatives to advertise the benefits of its policies for the people as well as to analyze what people want. This requires a government to diffuse its energies, making it less efficient in other functions, and so it will only expend such energy if it provides an overall benefit to the party’s fortunes. Some people will represent themselves as the voice of the majority to increase their influence and, while this claim is often exaggerated, such people may prove useful for buying favors, since it is more efficient to buy the support of an influential person than to seek the support of a large number of citizens. This gives added weight to the views of some citizens at the expense of others, although the precise composition of influential voters may vary with different governments. Uncertainty, and the resultant asymmetry of information, precludes the ideal of democratic equality from prevailing in political practice.

Chapter 7 seeks to deal with the topic of ideology within a model that treats all political actors as rational and self-interested. Within the model, an ideology communicates an image of the good life to voters, although this is strictly a means of helping the party that espouses it to hold and retain power. Even under conditions of pure rationality, parties have to distinguish themselves from one another with respect to policy; under conditions of uncertainty, voters may be drawn to philosophical appeals rather than the substance of each policy proposal. Ideologies help sharpen the differences between parties and provide a shortcut for voters to make their decisions. In the long run, voting on the basis of ideology is rational only if it correlates to policy differences that impact the voter, since ideology in and of itself is nothing but a series of statements. Ideology might conflict with the party’s goal of appealing to a majority, as it presupposes a much starker difference than specific stances on policies, but there are profound ideological divisions within modern societies which a party may easily exploit. Ideologies are a much better vehicle for organizing coalitions of minority positions than trying to assemble each of their discrete preferences. Ideologies are fluid and so they may change in order to build new coalitions in response to political developments. In the rational model, parties in power must still deliver concrete benefits to their constituents; ideological orthodoxy will accomplish nothing on its own.

Although ideologies may be mutable, their statements must align with their actions and/or statements in the previous election cycle. An ideology helps to establish a party’s reputation for reliability (predictable action) and responsibility (keeping promises). A party may theoretically be reliable by constantly reversing itself, and an incumbent can dispense with reliability so long as makes its constituents happy, but the two qualities usually go hand in hand. Opposition parties need to be reliable in the conventional sense that they must demonstrate continuity between their previous time in power and current conditions. A party that changes its position too frequently will develop a reputation for untrustworthiness. Ideologies need to be coherent and refrain from changing doctrine except when circumstances change dramatically. A change of leadership is often a necessary precursor to a change in doctrine. Periodic changes will be necessary, but a party will struggle to adjust their doctrines to circumstances. Maintaining ideology consistency may sometimes restrict efforts to attract votes, but this is only part of a broader strategy to retain the largest overall vote share and not due to any sincere commitment to ideological principles. In sum, ideologies can be loose enough to attract a diverse array of voters, but they must be internally consistent and offer a reasonably consistent external message over time in order to attract and hold voters.

Chapter 8 focuses on the question of why ideologies either change or stay the same. It proposes that a two-party democracy requires a degree of consensus that limits ideological polarization, whereas a multiparty system encourages parties to differentiate from one another. The number and ideological diversity of parties depends on the relative stability of public preferences. Finally, a two-party system encourages both parties to keep their ideologies vague enough to justify resemblances in policy. The model adopts an ideological scale from 0 to 100, where each voter has a number signifying their position and favors other positions based on its degree of proximity (in either direction) to their own. The model is exemplified best with one issue at a time: Downs uses the example of how much the government should intervene in the economy. In a two-party system, both parties will converge toward the middle to broaden their base, although this runs the risk of alienating their more extreme voters who might be willing to punish their own party in the short term to deter any further ideological drift. If the voters are themselves extreme, policy will swing toward one side or the other with each election, and the preponderance of one or another social class is likely to give a distinct advantage to a corresponding ideology. In a winner-take-all voting system—where the party with the most votes is the victor—the tendency will be toward a de-facto two-party system, but more parties are possible when representation is based on the percentage of votes. The existence of one structure or another can also help condition the ideological distribution among voters, who learn to accept the convergence of two parties or the fracturing of multiple parties as normal. A multiparty system lacks the incentives for convergence, as the presence of three or more alternatives along the 0-100 scale means that any efforts to court voters on one side will incur losses on the other.

New parties occasionally emerge, some of whom wish to win outright and others of whom want to influence the existing parties. Downs uses the example of the1948 “Dixiecrats” whose main goal was to prevent the Democratic party from embracing civil rights. New parties are most likely when there are fissures within existing parties due to sudden shifts in the ideological composition of the electorate. There may also be situations where partisan gridlock can prompt the emergence of centrist parties, and a major social disruption such as a war or revolution can completely upend a political system. Ideally, a party can resolve the dilemma between broad appeal and ideological coherence by embracing different positions across the 0-100 scale, while packaging them under a common ideological identity. However, voters may choose a party that aligns more closely to their own range of ideological positions, even if the aggregate mean of all those positions is not similar to their own. Ranges for parties in a multiparty system will be significantly narrower. In a two-party system, it is possible for some parties to overlap with one another in particular areas, offsetting this with policies to please their most extreme voters. They want to conceal the aggregate mean of their positions so as to convince both moderates and extremists that they are on their side. This in turn makes it more difficult for their voters to make a fully informed choice, so that the rational attempt of the party to maximize their vote share interferes with the rationality of the electorate. The only check on this tendency is the rationality of the electorate. Overall, the central variable is the ideological distribution of the electorate, as well as the structure of elections.

Chapter 9 turns to the issue of governments composed of different parties, none of whom can earn a majority on their own. Downs hypothesizes that in coalition governments, voters have trouble voting rationally, because parties are constantly converging and diverging. Once again, a key variable is the ideological distribution of the electorate. Downs examines the voting model—found outside the US—where parties receive seats in the legislature based on their percentage of votes, and the legislature selects the prime minister and cabinet by majority. The coalition government has the same powers as a single-party government while in office, but will dissolve upon the next election. Downs leaves aside the intricacies of intraparty negotiations, and assumes that voters care more about the government than the legislature and that their decision-making is oriented around future coalitions instead of present ones. The rational voter must know which parties are willing to enter coalitions with which other parties, and which policy positions they would adjust in order to participate in those coalitions. Such calculations are easier when there are fewer parties, or more extremist parties disinclined to compromise. Lack of knowledge about which coalition will form makes it difficult for a voter to ascertain which party will benefit them the most. At the same time, it magnifies the importance of each vote since there are so many possible outcomes, and there is a far greater risk of a coalition implementing profoundly unfavorable policies than in a system of two converging parties. A key consideration is how voters are able to discern one another’s preferences. This is extremely difficult to do, and voters have a limited timeframe in which to make their assessments. Some voters may be so confused or disincentivized that they decline to vote. Others may vote as they please without bothering to calculate the preferences of others; this is irrational since the mere expression of a preference is not the same as advancing one’s self-interest. The more people who vote simply to express their preferences, the easier it is for other voters to rationally calculate their advantages, since they can predict the behavior of at least some of the electorate. The sheer number of possible outcomes encourages voters to be irrational, not in the sense of being foolish, but because they are now selecting the legislature rather than the government, the real locus of policy.

An electorate divided among many parties presents a legislature with the problem of selecting a government that can be said to reflect the will of the majority. A coalition government will not be able to implement an integrated policy platform because it is by nature an amalgamation of different ideologies. People vote for a party with a narrow range of positions, only to have it join a coalition which then implements policies across a very broad ideological range. Once a party has joined a coalition, they have an interest in cooperating with the others because otherwise nothing will get done. It is possible for them to converge around the ideological mean of the parties. This is easier if voters tend to be moderate. The parties may also stress their differences so as to broaden the appeal of the entire coalition, so that each party may be reelected. Yet parties are still more ultimately concerned with maximizing their own position, and will not settle for a minor role within a coalition if it has a chance to increase its share. Parties in a multiparty system have an incentive to be vague about what compromises they might make as part of a coalition, so as not to alienate potential voters. Here again, a party’s rational strategy for winning an election makes it harder for voters to behave rationally. This tension is inherent to political life, where everyone cannot pursue all of their goals all of the time. The possibility of a rational electorate depends on the degree of ideological consensus among citizens.

Chapter 10 turns to the relationship between government and the private sector. It focuses on how government efforts to maximize votes affect the profit-maximizing efforts of private economic actors and the level of approximation to a “Paretian optimum position”—a state of complete economic efficiency. Downs hypothesizes that some degree of government action is required to achieve a Pareto optimum, but that a democracy is extremely unlikely to achieve such a condition since it conflicts with the government’s efforts to win votes. In the private sector, firms are trying to maximize profits, particularly by maximizing assets with the highest present value, as present enjoyment generally outweighs future enjoyment. People can and will sell anything short of their own selves, whether their personal assets or their time, in order to prioritize their highest-paying investments. This makes the economy distinct from politics, where no one can buy or sell a vote. Additionally, politics entails the provision of public goods, coercing citizens to pay for services through taxes even if they personally have no use for them. By doing so, the government invariably makes someone better off by making someone else worse off, therefore preventing a Pareto optimal condition. Government action does not obey the principles of the market, as it has the authority to coordinate collective action rather than convince actors to do something out of their self-interest. For all of its power, a government cannot trade assets in order to maximize the present value of its investments, since the only metric of success is the vote total in the next election. Governments accordingly dispense money based on the votes it can produce, and those who might lose out from such transactions can do nothing but vote against that government in the future. Governments may seek a more equitable distribution of resources, thereby negatively affecting the status of the wealthy, or fall short of its duty to provide public goods. Even if governments had perfect information regarding voter preferences, there is no way they could appeal to the preferences of some without negatively affecting others. Given actual conditions of uncertainty, parties will not be able to accurately anticipate the effect of their policies on each segment of the electorate. Influential citizens may constrain the power of government to ensure that it does not take actions detrimental to their interests. In any case, governments enact policies that apply equally to everyone, so they cannot coordinate the preferences of individual citizens, even were they to have perfect knowledge. A government also cannot weigh the competing interests of citizens based on their respective value, since the only value the government cares about is the vote.

The only way for a government to bring the economy into a Pareto optimum, even theoretically, would be by having the power to buy and sell votes. In such a case, advantages would fall to wealthier citizens, forcing less wealthy citizens to band together, as larger groups marginalize the value of each individual member. Assuming such groups were able to form, politics would turn into a battle for influence between the rich who buy votes and the poor who sell them. This would cost the wealthy money, and the poor their influence and, so long as the poor care about government policy and not just cash in their pocket, they will want to ensure equitability of voting power. This would be mitigated if the vote-buyers and vote-sellers each formed one large group that dealt exclusively with one another, but asymmetry of preferences makes this essentially impossible. There is almost no way, except by pure luck, that a person’s contributions to the government will precisely match the benefits they receive, and the government can coerce someone who is not paying what they legally owe. While some coercion is present in the private sector, only the government can exercise a monopoly on certain kinds of transactions, and back its power up with legal force. Everyone has an obvious benefit in keeping a functioning government, even if it is inefficient and causes some measure of harm, because of the necessary benefits it uniquely provides.

The government can choose to redistribute wealth in a way that significantly changes the ratio of payouts to benefits for various citizens. Even so, governments will be reluctant to take too much from the wealthy, since they cannot gauge everyone’s potential income and therefore cannot tax their way into a condition of Pareto optimum. In sum, inequalities in the influence of different citizens, the coercive power of the government and its focus on maximizing votes rather than profits, the indispensability of government functions, and the inability to buy or sell votes makes it impossible for a democracy to reach a Pareto optimum. The power of the wealthy offsets a government’s tendency to secure votes by appealing to the less well-off. There is no way to achieve total equilibrium in the economy without an intolerable violation of political rights.