concurring specially.
I agree with the majority that the trial court’s order must be reversed. Howell v. Fears, 275 Ga. 627 (571 SE2d 392) (2002) and the other cases relied on by the dissent have been perpetuating a misunderstanding in the law stemming from unfortunate language used in Taggart v. Phillips, 242 Ga. 454, 455 (249 SE2d 245) (1978). Taggart was an election case in which the contestor challenged the outcome of the election based both on the casting of votes by unqualified, i.e., “illegal,” voters and on an irregularity in the ballot causedby misaligned voting machines. Taggart acknowledged those illegal voter cases that limited review of election cases to those votes recorded in a particular contest, noting that “to cast doubt on an election it is only necessary to show (1) that electors voted in the particular contest being challenged and (2) a sufficient number of them were not qualified to vote so as to cast doubt on the election. [Cit.]” Id. This limitation makes sense in “illegal voter” cases, given that if the unqualified voter did not vote in a particular contest the voter’s unqualified status could have had no practical impact on the outcome of that contest. As to the irregular ballot issue, Taggart noted that “[similarly, doubt may be cast on an election by showing improper maintenance of the voting machines resulting in votes being miscast. [Cit.]” Id. Unfortunately, the Taggart opinion then inartfully combined these two separate problems when it stated that the contestor “must show that a sufficient number of electors voted illegally or were irregularly recorded in the contest being challenged to change or cast doubt on the election.”1 This language has been cited in Howell and other cases for the proposition that the “contest being challenged” limitation applies to irregular ballot cases.2 However, *275unlike the illegal voter cases, application of this limitation makes no sense in a situation where the voters, all qualified, were issued irregular ballots by election officials. That is because we cannot know what effect the irregularity in the issued ballots had on the electors. The irregularity may be the very reason an elector declined to vote in that particular contest or chose not to vote at all. It may also be the reason an elector, confused by the irregularity, chose instead to cast a vote for another candidate in that contest. We do not know and should not surmise what an elector’s intent may have been in irregular ballot cases, for we may well come to the wrong conclusion.3 It is for this reason that I would disapprove Howell v. Fears, supra, Bailey v. Colwell, 263 Ga. 111 (428 SE2d 570) (1993) and any other case that indicates that only the votes cast in a particular contest may be considered in irregular ballot cases. In their stead I would hold that in an irregular ballot case, it is only necessary for a contestor to show that there were enough irregular ballots issued4 to change or cast doubt on the election.
I agree with the majority that not all errors on a ballot rise to a level sufficient to render the issuance of the flawed ballot an “irregularity” on the part of a primary or election official under OCGA § 21-2-522 (1). Assessment of such errors must necessarily be handled on a case-by-case basis.5 Furthermore, only those irregularities sufficient to change or place in doubt the result can serve to invalidate an election. Id. In this case I agree with the majority that the irregularity in the absentee ballots issued by Laurens County election officials was sufficient to place in doubt the result of the Court of Appeals contest. Thus, I agree with the majority that the trial court’s judgment must be reversed and the case remanded with the direction *276that the trial court act expeditiously in accordance with OCGA § 21-2-527 (d).
The more appropriate phrasing for that sentence would have been the contestor “must show that a sufficient number of electors voted illegally in the contest being challenged or were irregularly recorded to change or cast doubt on the election.”
While Taggart speaks in terms of “irregularly recorded” votes, Howell and other cases *275have expanded it to include any ballot irregularity. Hence, I would draw no distinction between irregular ballots and irregularly recorded votes.
Thus I would also reject the alternate basis for the trial court’s dismissal of Mead’s petition, which involved subtracting the number of votes cast by electors for candidates other than Mead in the Court of Appeals contest from the total number of irregular ballots issued and concluding that the remaining votes failed to cast in doubt the outcome of the election. This reasoning likewise involves improper speculation regarding the intent of the electors.
It follows that I must disagree with the majority that our consideration is limited to the 481 absentee ballots that were returned and counted. Instead I would include all absentee ballots that were issued because I consider it improper to speculate regarding the intent of those electors who failed to return the ballots or surmise that their decision was unrelated to the irregularity in the ballots.
Alexander v. Ryan, 202 Ga. 578, 582 (3) (43 SE2d 654) (1947) (finding illegal certain ballots altered by rubber stamp to add name of candidate not qualified to be on ballot), the adverse impact of the ill-considered language in Taggart stands as an example why this Court should take care in following casual language in cases lacking a thorough analysis of the pertinent law.