Alleviating the Problem of Rational Voter Ignorance: A Proposal for a "Ballot Portal"

Political scientists have long observed that what is good for the individual may be bad for the group. Under a ruthless dictatorship, for example, the people might want to overthrow their leaders and establish a democratic government. But it is very risky for any particular individual to participate in such an uprising. The individual bears all the costs, but the public at large receives the lion's share of the benefits. Economists call such goods "public goods." National Defense is a classic example. It is impractical to exclude any individual from the benefits of national defense, so national defense must be provided for everyone.

It turns out that voter acquisition of political information is much like this. Consider this scenario. You are an average citizen that lives in my home town of Severna Park, Maryland, and you are confronted with this choice: 1) research candidates for 18 elected offices on a ballot, or 2) research a one week family vacation to the Caribbean. The year is 2002 and every elected office at the county and state level is on the ballot. The state has a budget of a bit over $20 billion per year and the county (with approximately one tenth the population of the state) about $1 billion per year. In an upper middle-class community such as Severna Park, that comes to about $50,000 in taxes per household per election cycle. The election winners will have the responsibility for spending these funds on the citizen's behalf. In contrast, the vacation will cost $2,500.

On which decision will you invest more energy? The odds are it will be the vacation, despite the fact that the candidate decision involves twenty times as much money.

The explanation is a simple cost-benefit analysis. The benefits of price and quality shopping for the vacation will fully accrue to you whereas the benefits of choosing competent elected officials will be shared by all citizens. And even where your preferences significantly differ from others, the probability of your research making a difference in an election outcome is negligible whereas the probability of your research making a difference in your vacation outcome is great. Meanwhile, the cost of comparison shopping for Caribbean vacations is likely to be much less than the cost of candidate shopping. Expedia, Travelocity, Fodor's, and Let's Go, for example, offer a degree of affordable, high quality comparative vacation information that is simply unavailable for political contests except for the most visible offices such as U.S. President and state governor, which constitute only a tiny fraction of the more than 500,000 elected offices in the United States.

Public policy can do little to change the fact that individual citizens cannot reap the full benefits of becoming an informed voter and that the probability of such information making a difference is small. But it can do quite a bit to reduce the voter's costs of becoming informed. The government already does this in a host of direct and indirect ways.

• Public meeting and record laws are designed to reduce voter costs of becoming informed about the actions of their elected officials.
• Ballots are printed and publicized at public expense so that voters will know their choices.
• Periodicals receive subsidized 2nd class postage.
• Newspapers receive free space on public streets to place coin operated newspaper boxes.
• Broadcasters receive free use of the public airwaves.
• National Public Radio and local public TV receive public funds.
• Cable companies are given rights of way on public streets in return for promises to provide public, educational, and government access channels.
• Telephone companies are given privileged rights of way on public streets in return for commitments to universal service.
• Public officials are given large budgets for press secretaries so the public may know what they're doing.
• Courts give more libel protection to political speech than commercial speech because they recognize not only that such information is vital to a healthy democracy but also that the incentives to produce and acquire it are weak.

One way to reduce the public's burden in accessing political information is through new public policies that exploit the potential of new information technologies. Reformed telecommunications, copyright, and parliamentary policies are all important parts of this solution. But the idea I want to focus on here is what I call the ballot portal.

The core idea behind a ballot portal is that an online ballot could provide a highly efficient, content neutral interface for voters to gather information about candidates. At its simplest, the ballot portal allows voters to click on a candidate's name and link to the candidate's official website. But links on a ballot needn't be limited to candidate websites. They could also link, in a content neutral way, to information sources about candidates.

Today, governments print static ballots about all candidates. In many states, such as my home state of Maryland, candidates for important offices must jump through multiple hoops to get on the ballot. They must prove their identity and residence, they must register by a particular point in time, they must form a committee with a separate auditor for campaign receipts and expenditures, they must provide detailed and regular disclosures of their assets and sources of revenues, and they must get a certain number of signatures or win a preliminary election (e.g., a Democratic or Republican primary) to get on the general election ballot. Parties other than the Democrats and Republicans need 10,000 signatures to get on the ballot.

The underlying rationale for all these laws tied to getting on the ballot is to reduce the voter's information burden. My proposal is that these principles be extended to information sources about candidates.

To get on a ballot to provide information about a particular election, an information source would have to register by the same date candidates have to register. They would need to get as many signatures, if any, as an independent candidate. They would need to prove their identity and contact information. And they would need to disclose their assets and all sources of revenue. Information sources purporting to represent an organization could also be required to disclose their bylaws.

Next to each race on the ballot could be a simple link to a list of information sources providing information about the candidates in that race. Each of the links to these information sources, in turn, would go to the official website of that information source. These information sources could include political parties, interest groups, the press, and individuals. No restrictions would be made on the content of candidate evaluations. The goal would be to provide diverse and antagonist viewpoints, not a particular viewpoint. To get on and stay on the ballot, all an information source would have to do is satisfy the registration and regular disclosure requirements.

An additional type of information source could provide information about the candidate information sources. By adding this type of information source to the mix, the number of diverse and antagonistic sources of information would be further increased.

All information sources would have the rights and obligations of the press. They could publish on their websites anonymous information, but they would be legally liable for its content. Like the press, they would have reputations to be gained or lost based on what they published and the transparency with which they operated.

An online ballot is all about creating new options for voters, so a voter who wanted an old-fashioned print ballot would have that option. Voters who wanted to rely on conventional, non-ballot information sources would, of course, have that option, too. Another role for ballot portals might be to help facilitate a new system of campaign public finance. Instead of giving public money to candidates, money could be given to voters in the form of vouchers. Voters could then apply those vouchers, via the ballot portal, to either candidates or information sources. Professors Bruce Ackerman and Ian Ayres, in Voting with Dollars, have already developed a proposal for a similar voucher-based system in great depth. But whereas in their proposal voters can only spend their vouchers on candidates (who will then spend the money on information campaigns), this proposal allows voters to spend the money directly on information sources.

Surely, a ballot portal would have many biased information sources, since the goal of a ballot portal is not to eliminate information bias but to facilitate voter access to diverse and antagonistic sources of information. In this way a ballot portal would be very different from government provided explanations of referendum items, which strive for objectivity.

Because elected officials face an obvious conflict of interest in administering a ballot portal (like letting the "foxes guard the chickens"), it is essential to set up an independent entity to administer the system and resolve conflicts. Many election boards, ranging from the Federal Elections Commission (FEC) down to local town boards, achieve this independence by being non-partisan. The FEC, for example, has an equal number of Democrats and Republicans. But non-partisan boards may still have a pro-incumbent bias. To eliminate both partisan and incumbent bias, I suggest a system based on a large-scale, randomly selected citizen jury. A randomly selected public body has democratic legitimacy because it is representative of the general populace. The purpose of such a body is to eliminate both partisan and incumbent bias. Professors Robert Dahl and James Fishkin have proposed the use of large-scale, randomly selected citizen bodies to evaluate issues and candidates. The difference in this proposal is the specific emphasis on using such bodies for decisions where elected officials have a conflict of interest.

Today, it is almost always illegal to provide candidate information near a ballot box. Many jurisdictions draw lines outside places of election and won't allow any campaign literature to be distributed beyond that line. Voters, however, find information most useful when directly looking at a ballot and needing to make a decision. Ballot portals offer this information convenience without the intimidation associated with physical distribution.

In conclusion, political scientists have often observed that it is rational for voters to be ignorant and that political information has the characteristics of a public good. A ballot portal would not transform voters into democratic Einsteins. But as new options for voters, some or all of the proposals above could be significant improvement over the status quo.