Election Law @ Moritz

Commentary...

Recent Opinion Articles

Edward B. Foley
Free & Fair is a collection of writings by Edward B. Foley, one of the nation's preeminent experts on election law.

Weekly Comment

When to Redo an Election?

Print Page

January 25, 2005

Suppose 1,000 illegal aliens vote in an American election, where the margin of victory is less than 1,000. Wouldn't it be necessary to redo the election, to remove the taint of the 1,000 illegal votes?

Not necessarily. Suppose that these aliens had managed to register to vote - improperly, of course - but successfully nonetheless. And suppose that all candidates and their political parties had a fair opportunity before the election to review the list of registered voters and to challenge any names on the list they believed to be improper, but had failed to catch the 1,000 aliens on the list. In this situation, the law might plausibly take the position that the candidates and their parties are barred from contesting the election on the ground that it improperly contains the votes of these aliens.

Once the list of participants has been finalized, as flawed as it may be, the results should not be set aside on the ground that some of the participants should not have participated after all. Settling upon the list of eligible participants is something that should occur ahead of time, before the event is held. Mistakes can be made in this determination, as they can in any human endeavor. But some mistakes cannot, or should not, be rectified if intervening circumstances dependent on the mistake have subsequently occurred. The undertaking of an election, based on an openly verified list of eligible participants, is one such event that should not be undone on the ground that there was an antecedent mistaken in assembling the eligibility list. Or at least it can be so argued.

Of course, the circumstances are different if 1,000 illegal aliens are permitted to vote on Election Day even if they never previously registered to vote. We can imagine poll workers permitting 1,000 individuals to cast ballots without checking the poll books to see if they are registered voters. If it turned out that these 1,000 individuals had never registered and were ineligible to do so (because they were non-citizens, not entitled to participate), then the election would be indelibly tainted by these 1,000 illegal votes. Just because the count of all ballots cast in this election - including these 1,000 illegal ballots - shows more ballots cast for one candidate than any other, it does not follow that the election has produced a winner. If the margin of ballots by which the one candidate leads over the other is less than the 1,000 illegal ballots, then the illegality of these ballots unavoidably infects the legitimacy of the lead.

In this situation, the leading candidate cannot fairly say, "The election is mine regardless of the mistake." That would be likely claiming victory if 1,000 individuals were permitted by poll workers to vote twice (with the margin of "victory" again less than 1,000 votes). Or if the voting machines improperly counted 1,000 votes twice, although there is no way to know now which 1,000 votes were double strength.

When the election results contain these sorts of illegal extra votes, and no candidate leads by more than this number, then the system is incapable of identifying the winner of the election, and a revote is necessary. It is as if the system failed to produce any result at all in the first place - because that is precisely what happened: the system malfunctioned and could not say who won. "I should be declared the winner and permitted to hold office, because the system recorded 1,000 phantom votes that do not correspond to any registered voter and we cannot tell whether these votes were awarded to me or my opponent," is hardly the equivalent of saying "the count shows that I received more votes from registered voters than my opponent did."

From a distance, it appears that the contested gubernatorial election in Washington State may involve both kinds of problems described above. In challenging votes by felons who were registered to vote, the Washington Republican Party is attacking a mistake in assembling the list of eligible participants, and depending on the particularities of the relevant provisions of Washington law, it may well be that this Republican challenge is barred on the ground that, after the election, it comes too late.

Conversely, however, the Republicans also challenge 437 provisional ballots that were fed into tabulating machines on Election Day without first ascertaining whether these ballots were eligible for inclusion. Superficially, and sitting here in Ohio, it seems that this problem is essentially the same as permitting 1,000 illegal aliens to vote without ever checking whether or not they are registered. Perhaps there is a way to pin down the identity of the individuals who cast these 437 provisional ballots and then determine that these ballots would have been ruled eligible for inclusion had they been properly processed - in which case it would be possible to declare "no harm, no foul." (The media reports I've seen so far are unclear on this point: they all say that it is impossible to know the identity of the candidate for whom these 437 provisional ballots were cast, but they do not specifically clarify whether it is now impossible to know the identities - and therefore the eligibility - of these 437 provisional voters.)

But if the provenance of these 437 provisional ballots is entirely unknowable, then they function as essentially phantom ballots - improperly included in the election results, not corresponding to any registered voter. It is as though the system arbitrarily included an extra 437 votes - for which candidate, who knows - making it inherently impossible to say which candidate actually received the most real votes. And the problem is not in counting the ballots: even assuming that the 129-vote margin is perfectly accurate, the total includes 437 phantom ballots that have no business being in the mix and inevitably pollute the result. Moreover, the fact that a candidate has been "certified" the "winner" based on this defective non-result would not seem to obviate the need for a do-over.

There are even more issues affecting the contest of the Washington gubernatorial election. For example, there is the allegation that dead people voted. Or, to be more precise, that votes were cast in the name of registered voters who are now deceased. This problem falls somewhere in the middle of the two categories previously identified: on the one hand, it is a mistake in the list of eligible voters, a mistake which at least theoretically could have been corrected ahead of time by removing their names from the list of registered voters; on the other hand, the individual who cast the ballot on behalf of the dead voter clearly is not the same individual as the deceased registrant, so the fraud occurred at the polls on Election Day.

Washington State will need to resolve these and other issues according to its own election laws. But the difficulties that Washington is having should remind all other states of the value of a well-designed electoral system. One key component of a well-designed system is clarity: it clearly identifies far in advance of Election Day the permissible grounds and procedures for contesting a close election. If the well-designed system has clearly established a mechanism for verifying and challenging the names on the list of eligible participants, then the system can choose to remove this kind of challenge as a basis for contesting an election afterwards. But at the very least the system will have clarified this point ahead of time.