No. 77504. Affirmed.
Opinion filed February 17, 1995.
Appeal from the Appellate Court for the Third District; heard in that court on appeal from the Circuit Court of Bureau County, the Hon. Paul E. Root, Judge, presiding.
Mathias W. Delort and Burton S. Odelson, of Odelson Sterk, Ltd., of Evergreen Park, for appellant.
J. Paul Aplington and Thomas L. McClintock, of Aplington, Kaufman, McClintock, Steele Barry, Ltd., of La Salle, for appellee.
Paul Bazydlo contested the outcome of an election for village president of Ladd, Illinois. The trial court excluded 28 uninitialled ballots from the election count and declared Bazydlo the winner. The appellate court reversed, concluding that those ballots should have been included in the election results, thereby making Bazydlo's opponent, Harry Volant, the winner. ( 264 Ill. App.3d 105.) We allowed Bazydlo's petition for leave to appeal (145 Ill.2d R. 315(a)), and now affirm the appellate court.
On April 20, 1993, the Village of Ladd held a general election for village president. Bazydlo and Volant were the only candidates. The official vote canvass showed that Volant had won by one vote, 344 to 343.
A discovery recount (see 10 ILCS 5/22-9.1 (West 1992)) revealed that a total of 28 ballots were uninitialled. Of these, 25 were cast for Volant and 3 were cast for Bazydlo.
Bazydlo filed a petition in the circuit court of Bureau County contesting the outcome of the election. (See 10 ILCS 5/23-1.1a et seq. (West 1992).) The trial court conducted a full recount of the two Hall Township precincts in which the village election was held, precincts Nos. 4 and 9. The 28 uninitialled ballots were cast in precinct No. 9.
All five election judges from precinct No. 9 testified at an evidentiary hearing. Each testified that all of the in-precinct ballots were initialled, counted, and verified prior to the opening of the envelope containing the absentee ballots. Further, the number of in-precinct ballots matched the number of in-precinct voters that day.
The envelope containing the 52 absentee ballots was then opened. These ballots were separately stacked and counted. The number of absentee ballots was compared with the number of absentee ballot applications; only one absentee ballot had not been returned. Three of the election judges stacked the absentee ballots. They formed more than one stack because it was not possible to place 52 ballots in one stack.
After the absentee ballots were counted, election judge Linda Peterson was to initial them and their attached ballot stubs. Although she had more than one stack in front of her, she clearly remembered initialling only one stack. The election judges did not know how many ballots were in each stack. No election judge actually saw any uninitialled ballots on election night. However, Peterson believed that the only explanation for the 28 uninitialled ballots was that one stack of absentee ballots had not been initialled. Only after the absentee ballots had been counted, the stubs removed, and separately stacked were they then combined with the in-precinct ballots and taken to the central tabulation station.
Volant testified that during the discovery recount, which Bazydlo also attended, 27 uninitialled ballots were found clustered in the total number of ballots cast. Likewise, detached ballot stubs to the 28 uninitialled ballots formed a cluster in the total number of stubs. The twenty-eighth uninitialled ballot, referred to at trial as "spoiled ballot 4," was damaged and, consequently, rejected by the ballot counting machine. A visual inspection revealed that this ballot was cast for Volant. It was not included with the other 27 uninitialled ballots due to the physical anomaly.
The trial court noted that there were no allegations or evidence of voting fraud. However, the court found that the 28 uninitialled ballots could not be readily identified as absentee ballots without "speculation or conjecture." Further, the court found that the cluster of uninitialled ballots and the separate cluster of corresponding ballot stubs constituted merely a "fortuitous circumstance." Accordingly, the court excluded the 28 uninitialled ballots from the election count, making Bazydlo the winner.
The appellate court reversed the trial court's judgment. The appellate court concluded that the 28 uninitialled ballots were absentee ballots and that they should be included in the election results, thereby making Volant the winner. ( 264 Ill. App.3d at 110-11.) Bazydlo appeals.
Election Code section 24A-10(1)(b) provides that ballots that have not been initialled by an election judge must be marked "Defective" and not counted. ( 10 ILCS 5/24A-10(1)(b) (West 1992).) Statutes that require election judges to initial ballots are mandatory; uninitialled ballots may not be counted. Pullen v. Mulligan (1990), 138 Ill.2d 21, 49.
However, in Craig v. Peterson (1968), 39 Ill.2d 191, this court found an exception to this otherwise mandatory requirement. A court will consider the initialling requirement to be merely directory and allow the counting of uninitialled absentee ballots only if: "(1) the absentee ballots can be identified and distinguished from in-precinct ballots; and (2) the initialling requirement does not contribute to the integrity of the election process." ( Pullen, 138 Ill.2d at 52.) The parties may offer trial testimony and other evidence on the issue of whether the Craig exception has been met. See Pullen, 138 Ill.2d at 53; Goble v. Board of Education of Iuka Community Consolidated School District No. 7 (1980), 83 Ill. App.3d 284.
CN Standard of Proof
At the outset, we discuss the appropriate standard of proof in establishing the Craig exception. The Election Code provides that election contests "shall be tried in like manner as other civil cases." ( 10 ILCS 5/23-23 (West 1992).) In the ordinary civil case, the decision determines merely which party must bear an economic loss. Because there are no sound reasons for favoring one party over another, the party with the burden of persuasion must prove his or her case by a preponderance of the evidence. ( In re Stephenson (1977), 67 Ill.2d 544, 553.) A proposition proved by a preponderance of the evidence is one that has been found to be more probably true than not true. In re Estate of Ragen (1979), 79 Ill. App.3d 8, 13.
Occasionally, however, constitutional or policy considerations require a court to impose a higher standard of proof. In such a case, the party with the burden of persuasion must prove his or her case by clear and convincing evidence. 29 Am.Jur.2d Evidence § 157 (1994); see, e.g., Stephenson, 67 Ill.2d at 556-59.
Courts have defined "clear and convincing" evidence most often as the quantum of proof that leaves no reasonable doubt in the mind of the fact finder as to the truth of the proposition in question. Although stated in terms of reasonable doubt, courts consider clear and convincing evidence to be more than a preponderance while not quite approaching the degree of proof necessary to convict a person of a criminal offense. See Ragen, 79 Ill. App.3d at 13-14 (and authorities cited therein).
In the present case, the appellate court held that the standard of proof in establishing the Craig exception was proof by clear and convincing evidence. The court reasoned that "a clear and convincing standard adequately balances the conflicting interests in preserving the integrity of the election and avoiding unnecessary disenfranchisement of qualified absentee voters." 264 Ill. App.3d at 108.
We agree and so hold. Cases that require either a higher (see Snow v. Natzke (1986), 140 Ill. App.3d 367, 370) or lower standard of proof are overruled on this point.
In the present case, the appellate court correctly concluded that the Craig exception had been met. Addressing the first prong of the Craig exception, the evidence clearly and convincingly shows that the uninitialled absentee ballots can be identified and distinguished from the in-precinct ballots. The following facts are undisputed. All of the in-precinct ballots were initialled, counted, and verified before the envelope containing the absentee ballots was opened. The number of in-precinct ballots matched the number of in-precinct voters. Each in-precinct ballot had been initialled.
Further, the following evidence was uncontradicted and unimpeached. The number of absentee ballots corresponded to the number of absentee ballot applications. The election judges formed more than one stack of absentee ballots. Election judge Peterson remembered initialling only one of these stacks. Moreover, during the discovery recount, 27 uninitialled ballots were clustered together in the total number of ballots. Spoiled ballot 4 was physically separated due to its rejection by the ballot counting machine. Likewise, the 28 corresponding uninitialled ballot stubs were clustered together in the total number of stubs.
The trial court, however, discounted much of this evidence. The trial court found that election judge Peterson's testimony was "simply not sufficient to identify the uninitialled ballots as absentee ballots without having to resort to speculation or conjecture." The trial court likewise found Volant's testimony regarding the clusters of the uninitialled ballots and stubs to be merely a "fortuitous circumstance."
A reviewing court should not overturn a trial court's findings merely because it does not agree with the lower court or because it might have reached a different conclusion had it been the fact finder. The trial judge, as the trier of fact, is in a position superior to a reviewing court to observe witnesses while testifying, to judge their credibility, and to determine the weight their testimony should receive. Consequently, where the testimony is conflicting in a bench trial, the trial court's findings will not be disturbed unless they are against the manifest weight of the evidence. ( In re Application of the County Treasurer (1989), 131 Ill.2d 541, 549.) A judgment is against the manifest weight of the evidence only when an opposite conclusion is apparent or when findings appear to be unreasonable, arbitrary, or not based on evidence. See Johnson v. Abbott Laboratories, Inc. (1992), 238 Ill. App.3d 898, 905; Briggs v. Gaddis (1985), 133 Ill. App.3d 704, 707; Anderson v. Beers (1979), 74 Ill. App.3d 619, 623.
Nonetheless, the fact finder may not arbitrarily or capriciously reject unimpeached testimony. Where the testimony of a witness is neither contradicted, either by positive testimony or by circumstances, nor inherently improbable, and the witness has not been impeached, that testimony cannot be disregarded by the trier of fact. People ex rel. Brown v. Baker (1981), 88 Ill.2d 81, 85; Kelly v. Jones (1919), 290 Ill. 375, 378.
In the present case, election judge Peterson testified that the absentee ballots were formed into more than one stack and that she remembered initialling only one stack. Peterson's testimony did not depend on "speculation or conjecture" to identify the uninitialled ballots as absentee ballots. Rather, such identification is a logical inference from the uncontradicted evidence.
Undoubtedly, the uninitialled ballots could only be absentee. It is undisputed that all of the in-precinct ballots were initialled, counted, and out of the way when the absentee ballots were counted. As the appellate court noted, this fact distinguishes this case from McDunn v. Williams (1993), 156 Ill.2d 288. "In McDunn, the absentee and in-precinct ballots were facially indistinguishable, but the initialed and uninitialled ballots had been combined prior to counting, unlike the present case." (Emphasis in original.) ( 264 Ill. App.3d at 109.) Further, in the present case, the total number of absentee ballots corresponded to the number of absentee ballot applications.
Additionally, Volant testified that at the discovery recount, which Bazydlo also attended, the uninitialled ballots were clustered together in the total number of ballots. The corresponding uninitialled ballot stubs were likewise clustered together in the total number of stubs. This uncontradicted fact was not merely a "fortuitous circumstance." It is true that uninitialled absentee ballots cannot be identified and distinguished from inprecinct ballots based on how the ballot is folded, which this court has described as a "fortuitous circumstance" ( Morandi v. Heiman (1961), 23 Ill.2d 365, 373-75), or marked ( Crum v. Green (1966), 68 Ill. App.2d 246, 252). Using these means, individuals could manufacture false ballots and jeopardize the integrity of an election.
In the present case, however, the uncontradicted evidence shows that the uninitialled ballot and ballot stub clusters were not manufactured. Again, the absentee ballots were counted separately only after the in-precinct ballots were out of the way. More than one stack of absentee ballots was made; only one stack was definitely initialled. The absentee ballots and stubs were then combined with the precinct totals, which were then taken to the central tabulation station. When all of the ballots and stubs were subsequently recounted, the uninitialled absentee ballots and stubs from their respective stacks were discovered. We hold that the trial court's findings that discounted this evidence were against the manifest weight of the evidence. The evidence clearly and convincingly shows that the uninitialled absentee ballots can be identified and distinguished from the in-precinct ballots.
We now address the second prong of the Craig exception for uninitialled absentee ballots: whether the initialling requirement contributes to the integrity of this election. This court has held that, because absentee ballots are not cast in the polling place and are not opened until after the polls have closed, application of the initialling requirement to absentee ballots is not necessary to prevent voters from fraudulently stuffing the ballot box. Pullen, 138 Ill.2d at 53.
In the present case, neither party questioned the legitimacy of the uninitialled ballots or alleged any fraud or other irregularity. ( 264 Ill. App.3d at 110.) Consequently, application of the initialling requirement is not necessary to preserve the integrity of this election. We hold that the second prong of the Craig exception has been clearly and convincingly met.
After carefully reviewing the entire record, we hold that the Craig exception to the statutory initialling requirement applies to this election. Accordingly, the 28 uninitialled ballots are properly included in the election count. As the appellate court found, the final vote is 344 votes for Volant and 343 votes for Bazydlo.
For the foregoing reasons, the judgment of the appellate court is affirmed.