« Welcome | Main | PETA vs. The World »

January 20, 2005

The Question of Election Fraud

There is an old saying about sausage that goes something like this: if you like sausage, don't look too closely at how it is made. It is becoming increasingly clear that American electoral politics suffers a similar failing. Consider: in 2000, the Presidential election lingered for six weeks due to multiple problems with Florida's vote, with President Bush finally winning the state by 537 votes. In 2002, Jon Thune lost his bid for the Senate by 527 votes despite numerous allegations of voting abuses. In 2004, Dino Rossi lost by 130 votes to Christine Gregoire only after two recounts produced an additional 391 votes for Gregoire. Some Democrats still maintain that President Bush stole Ohio and the election with over 118,000 fraudulent votes. And now Stranded on Blue Islands asks if President Bush might not have really won Wisconsin (HT: PoliPundit), as it appears that some 9,000 ballots in Milwaukee alone are of questionable provenance, a not-insignificant issue in a state won by 11,384 votes.

To clear my own biases first, I don't buy the arguments that the Republicans stole Florida in 2000, nor that they stole Ohio in 2004. Florida appeared to be a comedy of errors in 2000: a butterfly ballot intended to help the elderly instead confusing them; the networks calling Florida before the polls closed; Republican intimidation of Democrat election workers; God knows how many 2-1 votes after attempts to discern voter intent. Ultimately, we'll never know who would have won Florida without all of the errors. What we do know is that a media recount of the ballots showed that Bush won a majority of the ballots cast under the rules in place on election day. As for Ohio in 2004, if the Republicans really can generate 120,000 fraudulent ballots, I have no problem with them winning as long as they put those guys to work solving our real problems in between elections.

The pattern seems pretty clear, regardless of where you want to lay the blame. Our voting system guarantees that, in a close vote, nobody will ever really know who won. This is not healthy for a country that depends on the assumption that the government is ultimately accountable to the people.

Many liberals hated Reagan. Many conservatives hated Clinton. But I don't recall either of them claiming that their respective white whales were illegitimate. There may have been anger on both sides at the voters, but ultimately each side accepted the fact that their nemesis was, in fact, President. This is a good thing.

I have argued in the past that the notion that government depends on the consent of the governed is a fiction. Nobody from the government will ask you if you are willing to submit to the United States government. If you're in the country you're bound to live under its laws, and the government will not hesitate to use the power of the state to enforce that. As Randy Barnett argues in his excellent Restoring the Lost Constitution, it's simply not possible to have a government that is truly run at the consent of the governed; the best we can hope for is a government that is just. A government that depends on popular support requires a certain level of faith among the electorate: if the system is perceived as just, the people will accept the decisions imposed upon them even if they disagree with those decisions. But if the perception grows that government is not just and does not respond to the will of the people, the support for government is eroded and will eventually reach the point of collapse. Ultimately, our government works because enough people believe it works to support it. Should that balance ever shift, the result would likely be a constitutional crisis or convention, or (less likely), another civil war.

Belief that elections are no longer producing correct results is one of the most likely routes to that end. As long as a large enough fraction of the population believes that President Bush really did win the 2004 election, the government remains secure. If belief in election fraud were prevalent enough, the government could begin to see larger and larger numbers of people refusing to obey the laws because they believe the government in power is illegitimate. It would likely take a major shift in our culture before such a threat could emerge; even in the wake of the 2000 election, which was functionally a tie, large-scale civil disobedience never became a problem.

Each election cycle that passes with claims of fraud moves our culture closer to one that might someday refuse to accept the results of an election. The number of people who believe that the Bush campaign was somehow able to steal 119,000 votes is small, but vocal, and one hard fact of life is that a statement repeated often enough will begin to be believed. Right now, the margin of the vote in Ohio will probably keep all but diehards from believing that Ohio's vote count was fraudulent, but the other examples I cited above were so close as to make questioning the results quite natural. And people who begin to question the accuracy of the vote counts will then look to see what is being done to resolve the problem. As of right now, the answer seems to be not very much. That answer can only exacerbate the concerns of people worried about vote fraud.

The country as a whole seems to be moving towards electronic voting machines, an understandable move given the technological marvels of the day. Electronic voting is far more likely to provide accurate results than the antiquated punch card machines used in many places. It eliminates the problems of dimpled chads/multiple votes/etc. that set thousands of people to staring at punch cards for weeks in November 2000. A properly programmed electronic voting machine can remind voters when they haven't cast ballots on all issues and can sum up their choices before they cast their ballots, reducing the risks of people making errors when they vote. But electronic voting also risks nearly untraceable fraud, as there is not necessarily a paper trail associated with electronic voting. No matter how carefully you check to make sure you've cast your votes for the proper candidate, once you hit the submit button, you have no way to go back and make sure that your ballot was counted properly. The data enters the machine and can be altered as the programmer sees fit. If the count is close, there can be no recounts, because the data is all the same.

There are alternatives that will overcome this. The machine can print a receipt (or more than one) that displays the voter's selections without displaying the voter's name, maintaining confidentiality while creating a hard record of the voter's choices. I cannot recall where I saw this suggestion, but someone suggested putting a unique ID on the ballot and allowing the voter to bring a copy of the receipt home. The data would be placed on the Internet, allowing voters to enter their ballot number to confirm that it was counted for the right candidates. The one downside to this is that unscrupulous types could then demand to see people's ballots to confirm that they voted the 'right' way. Laws would bar this, of course, but there are many laws in this country that people obey only when they feel they must. If we create a receipt that proves how people voted, it is certain as night follows day that some people will demand to see that receipt. Nonetheless, that might be a cost people see as worth bearing to ensure that their vote is properly counted. On the other hand, many people would likely be content simply to know that a hard copy of their vote was kept at their voting place, so it could be checked against the machine count if necessary.

We could also stick with a tried and true method of balloting: paper ballots. When I lived in Massachusetts we used paper ballots that were then optically scanned, a method that seemed to combine the best of all worlds: an easy to use and understand ballot with rapid initial counting of those ballots and the ability to count the votes manually later if necessary. Paper ballots and optical scanners are also likely to be cheaper than more complex electronic voting machines, no small consideration for districts where money for voting is tight.

Which brings us to another consideration: do we continue to permit local communities to control how they vote for President, or do we federalize the system to ensure that everyone votes in the same way, minimizing the risk of votes being counted differently in different locations? The Constitution is quite clear on this: each state legislature has the power to determine how it selects its electors. A federal ballot also would require people to vote two separate ballots: one just for President, and one for all other state and local races. A minor inconvenience, perhaps, but not so necessarily so minor for the thousands of officials who would have to deal with it. Still, it might be worth it to ensure that people's faith in the Presidential election was maintained. Such a change would require more than just general agreement; it would require a Constitutional amendment. That process takes time, although if popular opinion were in favor of such an amendment, it could probably be adopted quickly. Still, many people are leery of altering the Constitution without a clear need to do so, and that feeling could delay or even prevent the adoption of such an amendment.

A quicker solution would be for Congress to make funds available to local communities to shift to better voting machines. Each community could choose not to use the funds if they so chose, and Congress could likely select several options for the funds to be spent on, so long as the options were selected for their resistance to fraud. I suspect few communities would not choose to make use of funds for improved voting machines if they were available, so we would likely end up with a system far more resistant to election fraud without having to amend the Constitution and place those decisions in the hands of Congress.

This leaves only the question of whether or not it is appropriate to spend federal dollars on voting machines. Unfortunately, a review of Article I, Section 8, indicates that Congress is not granted any authority over local elections, and therefore does not have the power to appropriate money for such activities. So we're back to a Constitutional amendment. Given that we would have to amend the Constitution in either case, I would prefer to leave the control of elections in the hands of local officials and simply give Congress the ability to pay the costs of national elections. Leaving power at the lowest levels is in keeping with the traditions of federalism, and would allow communities to reduce election fraud while still minimizing the troubles with merging local and federal issues on a single ballot.

Given how little attention Congress has paid to the concern of election fraud to date, despite numerous problems over the past four years, it is unlikely this solution will even be considered in the short run. This is unfortunate, as each successive election that has real problems with voting will further degrade the faith the public holds in the accuracy of their elections. At some point, we may reach a level of distrust such that no action of Congress would suffice to regain the trust of the people. It would be far better to eliminate the cause of such mistrust now.