People ex rel. Williams v. Cicott

16 Mich. 283 | Mich. | 1868

Lead Opinion

Campbell J.

The jury sworn to try this cause have found a verdict that defendant received the greatest number of votes for the office of sheriff of Wayne County. A motion is now made for a new trial, based on legal rulings complained of, and on the ground that the verdict was against evidence.

The chief controversy upon the trial, as well as on this motion, has arisen out of the alleged reception on either side of votes from unqualified persons. These voters are alleged to have been disqualified respectively by non-age, non-residence, or mental incapacity. Other controversies also arise which will be referred to in their place.

The first inquiry, therefore, is, whether an election can be defeated as to any candidate by showing him to have received illegal votes.

The authorities upon election questions are in this country neither numerous nor satisfactory. In England, where votes are given viva voce, it is always easy to determine how any voter has given his voice. And in some states of the ITnion, a system seems to prevail of numbering each ballot as given, and also numbering the'voter’s name on' the poll *295list, so as to furnish means of verification when necessary. It has always been held and is not disputed, that illegal votes do not avoid an election, unless it can be shown that their reception affects the result. And where the illegality consists in the casting of votes by persons unqualified, unless it is shown for whom they voted, it cannot be allowed to change the result.

The question of the power of courts to inquire into the action of the authorities in receiving or rejecting votes is, therefore, very closely connected with the power of inquiring what persons were voted for by those whose qualifications are denied. It is argued for the relator that neither of these inquiries can be made.

No use can fairly be made in such a controversy as the present, of decisions or practice arising out of any system of open voting. The ballot system was designed to prevent such publicity, and not to encourage it. And the course adopted by legislative bodies can not be regarded as a safe guide for courts of justice. There is little uniformity in it, and much of it is based on English precedents, belonging to a different practice. The view taken of contested elections by these popular bodies, is not always accurate or consistent with any settled principles.

There is no case, .so far as I have been able to discover, under any system of voting by closed ballot, which has held that any account could be taken of rejected votes, in a suit to try title for office. The statutes here, and probably elsewhere, require the election to be made out by the votes given. But it is plain enough that in most cases it would be quite as easy to determine for whom a rejected voter would have voted, as for whom any other actually did vote. In many cases it would be easier, because the vote is always ready and tendered, with better opportunities of observation than are given where it is received and deposited. But the element of uncertainty has been regarded as sufficient to cause the rejection of any *296such, inquiry, and in most cases, probably, it would not be admissible under the statutes. But the policy which leads to this result must have some bearing upon the construction of the whole system.

So far as I have been able to discover, by means of the somewhat imperfect indexes on this head, there is but one case in which the decision has turned upon the propriety of allowing inquiry into the qualifications of voters, and the identification of their tickets when claimed to be disqualified. That case was the case of People v. Pease, 27 N. Y. 45. In the Supreme Court the judges, although arriving at a general result, were equally divided on this point. In the Court of Appeals the judges elected to that tribunal were also equally divided, and a majority of the Supreme Court judges belonging to it by rotation turned the scale, and decided that the inquiry was proper. The decision was based chiefly upon English authorities, the previous New York decisions having turned principally on other errors, which rest upon somewhat different grounds.

New York, so far as may be inferred from the absence of decisions elsewhere, seems until recently to have been the only state, preserving the ballot system, in which the right to office by election is open to examination on the merits to any considerable extent. The courts of that state have gone further than any others in opening the door to parol proof. Some of the western states have, upon the authority of the New York ca'ses, permitted some of these matters to be litigated, but they are not in any majority. And it is quite manifest that the decisions have not in general acted upon any careful consideration of the important questions of public policy underlying the ballot system, which are so forcibly explained by Denio, O. J. in his opinion in People v. Pease. And it is a little remarkable that in New York, while so many doors have been opened by the decisions, the law requires all the ballots, except a single specimen of each kind, to be destroyed; thus leaving *297the number of votes of each kind, in all cases, to be determined by the inspectors, and rendering any correction impossible. I think the weight of reasoning is in favor of the view of Judge Denio, that no inquiry can be made into the legality of votes actually deposited by a' voter, upon any ground of personal, right as an elector.

The reasons why such an inquiry should be prevented do not necessarily rest on any assumption that the inspectors act throughout judicially, although under our registration system that objection has a force which would not otherwise be so obvious. Neither do they rest in any degree upon the assumption that one rule or another is most likely to induce perjury — as very hastily intimated in People v. Ferguson, 8 Cow. 102. But a very strong ground for them is found in the fact that our whole ballot system is based upon the idea that unless inviolable secrecy is preserved concerning every voter’s action, there can be no safety against those personal or political influences which destroy individual freedom of choice.

It is altogether idle to expect that there can be any such protection where the voter is only allowed to withhold his own oath concerning the ticket he has voted, while any other prying meddler can be permitted in a court of justice to guess under oath at its contents. If the law could permit an inquiry at all, there is no reason whatever for preventing, an inquiry from the voter himself, who alone can actually know how he voted, and who can suffer no more by being compelled to answer, than by having the fact established otherwise. The reason why the ballot is made obligatory by our constitution is to secure every one the right of preventing any one else from knowing how he voted; and there is no propriety in any rule which renders such a safeguard valueless.

It has always been the case that the rules of evidence have, on grounds of public policy, excluded proof tending to explain how individuals have acted, in positions where *298secresy was designed for their protection or that of the public. No grand juror could be permitted to disclose as a' witness the ballots given by himself or others upon investigations of crime; informers can not be compelled to disclose to whom they have given their information; and many official facts are denied publicity. In all of these cases, the rule is not confined to one person any more than to another, for public policy is against publication from any source. And if, as is clear, a man is entitled to keep his vote secret, it is difficult to see how any testimony whatever can be allowed, from any source, to identify and explain it.

The statutes contain some provisions bearing upon these topics with considerable force. By section 47 of the Oomp. L. every voter is compelled to deliver his ballot folded, and by section 52 the inspector is prohibited from either opening or permitting it to be opened. .

The devices adopted for creating different appearances in the ballots of different parties are. such palpable evasions of the spirit of the law, as to go very far towards destroying the immunity of the voter, and in some states it has been found desirable to attempt, by statute, the prevention of such tricks; but the difficulty of doing this effectually is exemplified in People v. Kilduff, 15 Ill. 492, where the evidence seems to have shown that a uniform variation may be entirely accidental. Unless some such difference exists, it would be idle to attempt any proof how a person voted; and it would be better to do away at once with the whole ballot, than to have legal tribunals give any aid or countenance to indirect violations of its security. And the evidence received in the present case exemplifies the impropriety of such investigations. In some instances, at least, the only proof that a voter, complained of as illegal, cast his ballot for one or the other of these candidates, was that he voted a ticket externally appearing to belong to one of two political parties, and containing *299names of both state and county officers. To allow such proof to be received in favor of or against any particular candidate on the ticket, is to allow very remote circumstances indeed to assume the name of evidence. And the necessity for resorting to such out of the way proofs, only puts in a clearer light the impropriety and illegality of entering upon any such inquiry, when the law sedulously destroys the only real proofs, and will not tolerate a resort to them. And the whole state is much more interested than any single citizen can be, in emancipating elections from- all those sinister influences which have so great a tendency to coerce or deceive electors into becoming the mere instruments of others.

But there are further provisions bearing more directly on the propriety and necessity of allowing no inquisition into individual votes.

County officers are among those included under section 31 of the compiled laws, which declares that “the persons having the greatest number of votes shall be deemed to have been duly elected.” The law does not confine this to votes cast by authorized voters, and can only be applied to votes cast and recorded in the manner provided by law. And although this section, standing alone, might be open to construction, yet, when the whole law is taken together, there are provisions not to be reconciled with any rule allowing single voters and their votes to be made the subjects of inquiry. It will not be denied that any inquiry into the legality of a particular voter’s qualifications, after his vote has been cast, is of a strictly judicial nature. And it can not be proper or legal to allow such an inquiry in one case and not in another. But it will be found not only that the rejection of votes from the count is required to be in such a way as to preclude any consideration of the person giving .or putting them in, but that there are cases where even a legal inqrdry into the ballots themselves is prevented.

*300In the first place, when two or more ballots are so folded together as- to present the appearance of one, and if counted will make the ballots exceed the names on the poll list, they are to be destroyed. — See. 61. And whenever, for any other reason, the number of ballots found in the box exceeds the number of names on the corrected poll lists, the inspectors are required to draw out and destroy unopened a number equal to the excess. — Sec. 62-This is, of course, upon the .assumption that the excess has probably been caused by fraud, -and assumes that no man’s vote ought to be counted unless the testimony of the poll lists shows that he actually handed in his ballot.

It is, therefore, altogether likely, upon any theory of probabilities, that in drawing out these extra ballots they will really be ballots lawfully put in, and this probability is in the ratio furnished by a comparison of numbers between lawful and unlawful votes. In other words, it is more than likely to punish the innocent instead of the guilty. The true method of arriving at the truth would be to inquire what vote each voter on the list actually cast, and destroy the remainder. The absurdity of this process upon such a large scale is such as to need no pointing out. But unless something-very like it is done in such a case as the present, the result obtained by any partial inquiry will be no better than guesswork. Where votes are thrown out, no one can tell whether the illegal voter whose vote is sought to be assailed has not already had his vote cancelled. The adoption of the principle of allotment is the most sensible and practical measure which could be devised, and I can not conceive how it can be improved upon by any subsequent search.

But when the inspectors have made their returns to the the county canvassers, and by those returns, a tie vote appears between two or more candidates who are highest on the list, their right to the office is to be determined by lot, and the person drawing the successful slip is to be “deemed *301legally elected to the office in question,”— §§ 76, 132, 133. In case tlie state canvassers (who can only count the votes certified to them) find a tie vote, the Legislature have power to choose between the candidates.— Const. Art. 8, § 5. In these cases there can be no further scrutiny, and in the case of state officers, if such a scrutiny were had, no end could be reached within any reasonable time, and there would be a practical impossibility in attempting to conduct it in any time within the official term, or to approach accuracy in a count of some thousand or more of ballot boxes before a jury. Yet state officers are not less important to the private elector, and, of course, are not to the community at large, than local. And the nearer a vote approaches a tie, the more likely it is that a rigid scrutiny might change its character. There is no more reason for preventing investigation behind the ballots in the one case than in the other.

The statute also takes very efficient measures to prevent any needless litigation, by shutting out any preliminary resort to the means of information. If the officers do their duty, no one else can ever know whether their count is correct or not, until a suit is brought and issue joined upon it. The ballots are required to be sealed up, and not opened except for the inspection of the proper authorities in case of a contest. — § 65. The only ballots open to public inspection are those which are rejected upon the canvass for defects apparent on their face.— § 65. These ballots are not sealed up with the rest, but are filed. While, therefore, it can be determined by inspection whether votes which have been thrown out should have been counted, the law does not seem to favor any unnecessary disturbance of the official returns, and any one who assumes to dispute an election is compelled to begin his suit before he can have access to the means of proof. This is not the usual course of litigation, and the rule has a strong bearing upon the policy to be deduced from the law.

*302Under our statutes there is no general 'provision which makes the canvass for local officers conclusive in all cases, and, therefore, the rule is recognized that the election usually depends upon the ballots, and not upon the returns-These being written and certain, the result of a recount involves no element of difficulty or ambiguity beyond the risk of mistakes in counting or footing up numbers, which may in some respects be more likely in examining the ballots of a whole county, than in telling off those of a town or ward; but which involves no great time or serious disadvantage. But the introduction of parol evidence concerning single voters in a considerable district can rarely reach all cases of illegality effectually, and must so multiply the issues as to seriously complicate the inquiry. And when we consider that for very many years legislation has been often modified for the very purpose of suppressing illegal voting, and when we know that hundreds of elections must have been turned by the ballots of unqualified voters, the absence of any body of decisions upon the subject is very strong proof that inquiry into private ballots is felt to be a violation of the constitutional safeguard on which we pride ourselves, as distinguishing our elections from those which we are wont to regard as conducted on unsafe principles.

No system can be devised which will prevent all illegal voting. But it can not be said our legislation is not as likely to shut it out as any means open to judicial control would be. The registration law forbids the board from recording any name of which they have well-founded doubts, and it is practically impossible for any stranger to succeed in defrauding the law, with the publicity given to all the proceedings. "Where a person - applies for registration on election day, the inspectors act upon discretion, and are not compelled to admit a vote unless satisfied of its legality. The challengers on both sides, as we all know, canvass every district beforehand, and expect to challenge every one who is not known. While the inspectors can not reject a *303registered voter who takes the proper oath, yet the means of previous inquiry, and the imminent risk of detection and punishment, have reduced the dangers of illegal voting within very narrow limits. I do not see bow the law could go further, and allow a scrutiny of single ballots, without violating the constitution, and I think no such intention is fairly deducible from any of our statutes.

I am, therefore, of opinion that the election must be determined solely by the ballots received according to law; and that where the election proceedings are not irregular, and the law has been complied with in correcting the lists and preserving the ballots, the means of determining the result must be in the main arithmetical. The present case, however, presents some points arising out of an excess of ballots in some precincts, an alleged failure in one town to preserve the box in security, and questions concerning the propriety of counting some slips and imperfect tickets.

The most important inquiry among these is what effect the excess of votes over the names of voters appearing on the poll lists has upon the election in the township, or ward, or in the county. v

This can only be determined by ascertaining what the statute has provided concerning the conclusiveness of these lists.

Under the registration law, the inspectors of election are required to keep before them the registered list of voters, and not to receive the vote of any person not registered.— Laws, 1859, p. 488. This renders necessary some system of checking, to prevent double voting; and while the law does not prescribe it, the evidence shows it was resorted to in this case, and seems to be a practical necessity. Two clerks of election are also to be appointed, whose duty it is to keep separate and independent lists, in order that in case of doubt one may furnish the means of correcting the other. At each adjournment, as well as at the final closing of the polls, these lists are to be compared, and any *304mistakes in either corrected by the decision of the board, so that they shall be made to agree, before any further business is done, and before the ballots are counted or examined.— Comp. L. §§ 44, 53, 54, 59, 60.

Before the election commences, the ballot-box is to be thoroughly examined, to be sure that nothing remains in it, and precautions are provided against any subsequent unauthorized opening' — ■ § § 51, 55, 56.

When the poll lists are made to correspond, the ballots are to be counted without opening, “ except so far as may be necessary to ascertain whether each ballot is single: and if two or more ballots shall be found so folded together as to present the appearance of a single ballot, they shall be destroyed, when the number of ballots shall be found not to agree with the poll lists, as provided in the next section.” This directs that, “if the ballots in the box shall be found to exceed in number the whole number of names of electors on the poll lists, they shall be replaced in the box, and one of the inspectors shall publicly draw out and destroy so many ballots therefrom, unopened, as shall be equal to such excess.” — § § 61, 62.

“The ballots and poll lists agreeing, or being made to agree, the board shall then proceed to canvass and estimate the votes, and they shall draw up a statement of the result, and cause a duplicate thereof to be made, which statement and duplicate shall be certified by the' inspectors to be correct, and shall be subscribed with their names.” —§ 63.

The statute contemplates that every ballot found in the box in excess of the number of names on the poll-lists, got there by fraud or accident, and does not represent a voter. The duty of the inspectors is peremptory, and unless they perform it, the consequence must be that the number of votes which belong to each candidate out of the only ballots which the statute regards as actually delivered in, must remain in uncertainty. And, inasmuch as no one can be considered elected who cannot be shown to have received a *305plurality of the statutory ballots, and as votes in excess of the poll-lists cannot be counted in favor of any one, and no process exists whereby it can be known for whom the fictitious tickets were intended, this uncertainty may lead to the most serious results. Various theories have been suggested as to the influence such excess should have upon the election. One is that it invalidates every election in which the vote of the precinct is to be counted. Another is that it annuls the vote of the precinct in which it occurs. And a third is that it only affects such elections as would be turned one way or the other by counting the excessive votes in favor of one or another candidate. There is no good reason for holding elections void for such excess, where it cannot affect the result. No voters who have honestly voted ought to lose their ballots, unless it is impossible to give them effect. And where there is such a plurality in favor of any candidate, that he could afford to allow these doubtful votes to his adversary, and still be in advance of him* there is no difficulty in perceiving that he must have been voted for by a plurality of all who east their ballots, and his election should be established. But where the plurality is so small that the excess would turn the scale if allowed to the opposing party, it cannot be shown that either has a majority, because no one can tell what ballots were improperly introduced, and therefore it cannot be determined who would have been benefitted by their exclusion. An election cannot be allowed by law to depend on an uncertainty. The majority must be susceptible of-proof.

There is no room in such a case for the application of probabilities, for there is nothing whatever on which to base them. In State of Ohio v. Ritt, Am. L. R. Dec. 1867, p. 88, and in the Penn Dist. Elect. 2 Pars. (Pa.) R. 533, an unauthorized closing of the polls was held to avoid an election; and in the Locust Ward case — 4 Pa. L. J. 341 — it was held that if polls were kept open too long, and the votes cast during the excessive period were enough to have possibly *306changed the result, the election must fail, because no inquiry could be allowed into the individual votes.

If in the present case any such' uncertainty exists, the election must be deemed to-have miscarried, and must have the same result as if there had been no election. As the case was presented on the trial, this consequence would follow. In the Sixth Ward, it appears by the count that nine votes were found in excess of the poll lists. In the Seventh Ward there were seven. Thus, as now appearing, there are sixteen votes which cannot be counted; and this number would, as the case now stands, turn the election. But, as upon a new trial, it is impossible for us to determine how the facts may appear, it will become necessary to consider some further questions which will "then be presented.

In the town of Dearborn, the comparison of ballots upon the trial showed a different result from the original return of the inspectors. Instead of sealing up the ballots, as required by law’ (§ 65 ), they were thrown back, open, into the box, which was also left unguarded except by an ordinary lock. The Court charged the jury that inasmuch as the ballots had not been preserved in the manner required by law, there was not the same certainty of their correctness which the law would have otherwise presumed, and it was, therefore, left to them, as a matter of fact, whether they would rely on the count or on the inspectors’ return as most reliable. This course was the proper one under the circumstances. The original count by the inspectors is open and public, and made in the performance of a sworn duty. It furnishes presumptive evidence, until rebutted by a new count upon a legal investigation. The sealing and securing of the ballots is deemed essential by the statute to prevent fraud; and while we held, in People v. Sackett, 14 Mich. 320, that such a neglect would not avoid an election, yet it throws enough suspicion upon it to render it subject to a close scrutiny, and the jury must, therefore, determine the matter. In regard to the slips concerning which a controversy *307has. arisen, I think that unless a person has so pasted on his slip as to show beyond question for whom his vote is cast for a particular office, it cannot be counted. Where it is so placed as to show upon the face of the ballot two distinct names for the same office, it comes, I think, within the express provision of the statute concerning double votes, which can seldom be given except by this mode ( § 48). But if the name is placed above the name of another candidate for the same office, so as to partially obliterate it, I think that would be sufficient to show an intent to vote against the one and for the other. And the omission of the word “for” before the office cannot be material, as the description of the office, even in an abbreviated form, if unequivocal, must be held sufficient. — People v. Matteson, 17 Ill. 167. Upon this principle, the slip in Monguagon should not be counted, but if found attached to the edge of a ballot would make it bad for duplicity. So slips containing only certain parts of a surname should in like manner not be counted as if they had been perfect. The voter’s intention must be expressed in favor of some -one, and half or two-thirds of a name cannot, unless idem sonans, represent the whole.

But a more important inquiry arises concerning the allowance of testimony to show that votes cast for E. V. Cicott, E. B. Cicott, and G. O. Williams, were designed for Edward Y. Cicott, and Gurdon O. Williams. It was held by this Court in People v. Tisdale, 1 Doug. R. 59, that votes cast for persons by their initials could not be counted as if giving the full name. In People v. Higgins, 3 Mich. 233, a similar ruling was made. Between those decisions the election laws were twice revised, and no change was made on this subject. The law has been acquiesced in. In New York the acquiescence of the Legislature in the opposite rule was considerably relied on in People v. Cook, 8 N. Y. 67. This would seem to render it improper to disturb the rule unless upon the most urgent grounds, and unless *308it is one leading to very bad consequences. It is not claimed that the inspectors or canvassers can make any inquiry into the identity of initials with full names. If this is done at all it must be done in courts. The effect of such a doctrine would be that the statements and canvass which the law designs shall be the evidence of all elections, except where there has been some blunder or fraud on the part of the inspectors, may be changed by means not in the power of those officers, who would be compelled to put a man in office, and who would be guilty of a crime in deliberately refusing to do it, and yet he would be liable to ouster by ballots which they could not count. I cannot conceive it possible that the law can contemplate a case where a person cannot lawfully obtain the evidence of title to office except by litigation. The statements of every board of inspectors and canvassers are required to give the names of all candidates “ written out in words at length.”— (§ 64). In tie votes, and in cases where votes are canvassed by state canvassers, these are conclusive as already shown. In registering voters their Christian names are required to be given in full.— (L. 1859, p. 483 ). And in the way in which our elections are conducted, a person cannot, without the grossest carelessness, fail to be informed of the full name of each candidate.

In popular business transactions, the use of initials, or even of a surname alone, may suffice, because evidence is easily found to prove identity by various means. And had the election law not designed to make the ballot the final test of the elector’s will, such looseness would not be so material. But in solemn legal proceedings in courts of justice, it has always been customary to use the full name, and the reasons for requiring it in elections are quite as forcible. The elector’s own oath, which is the best means of explanation, can not be required, nor, in my judgment, permitted. The doctrine allowing names to be made out from initials has, in New York, been supplemented by allowing the *309names actually given to be contradicted. — People v. Cook, supra. This doctrine was connected originally with its necessary supplement, that the voter could swear to his intention. But this is now repudiated.— People v. Saxton, 22 N. Y. 309; People v. Pease, 27 N. Y. p. 45, 85. The New York decisions are the sole foundation for those in "Wisconsin, which accept the rule but deny the right of proof by the intention of the voter himself. On the other hand, it was held in Ohio that when the Legislature held an election, and the journal of one house described the candidate as “Lemuel,” and that of the other as “Samuel,” no evidence could be received to identify them.- — State v. Moffitt, 5 Ham. 358. So when Abel C. Dinsloio was in Maine declared elected, and no such person could be found, it was held that Abel E. Dinsloio could not receive the office.— Opinion of Judges, 38 Me. 597.

The election laws contain so many evidences of a design to require great accuracy in all proceedings, that it would not, I think, carry out public policy to recede from the rules formerly held by this court, until the Legislature see fit to open the door. Thus far there has been great solicitude to prevent uncertainties and frauds. "Until the inspectors or canvassers are permitted to inquire into identity, I can see no propriety in permitting such inquiries by courts. An investigation into the systems of the various states, shows a general disinclination to allow offices to be -subject at all to judicial inquiry. It certainly is not desirable to furnish any more room for parol evidence than is absolutely unavoidable.

I think the ballots with initials can not be allowed.






Concurrence Opinion

Christianoy J.

I concur entirely with my brother Campbell in the views he has expressed as to the slips which were in controversy in this cause, and the mode in which the intention of the voter must be thereby expressed. I concur also in holding *310that the discrepancy between the inspectors’ certificate and the re-count of the votes in Dearborn, was-(owing to the careless manner in which the ballots were kept,) properly left to the jury.

With reference to those ballots in which the Christian names of the respective candidates were expressed only by initials, I concur in adhering to the rule established in People v. Tisdale, 1 Doug. 59, yet I am compelled to say that but for that decision, I should have been disposed to hold that, upon principle, extrinsic evidence might be given tending to show for whom the vote was intended; as that the candidates were in the habit of thus writing their names; that they were as well known, respectively, by the name of E. V. Cicott and Gr. 0. Williams,, as by Edward V. and Gurdon 0.; that they were opposing candidates at the election for the same office, and that no other person of the same surname and the same initials was known to be running for the office. This is in strict accordance with the rule which prevails in the construction of all other written instruments, which are to be read in the light of the surrounding circumstances. But the rule in People v. Tisdale, was recognized in People v. Higgins, 3 Mich. 233, and has now been the settled law of this State for a quarter of a century. This rule has the merit of simplicity and certainty, of being easily understood and applied, leaving no room for discretion in the inspectors ; and, as a general rule, is equally fair and just in its application to all parties. I do not therefore think it wise to disturb it by establishing another rule which to me may seem more sound in principle, but which in its practical application, might not be likely to produce any fairer results. The Legislature have full control over this question, and may change the rule when the public sentiment shall seem to require it.

I concur also in holding that it is utterly useless to inquire into the qualification of voters, when the nature of their votes, or the candidates for whom they have voted *311cannot be ascertained; and I fully agree with my brother Campbell that no qualified elector can be compelled to disclose how he has voted. But I cannot go to the extent of holding that no inquiry is admissible in any case into the qualification of voters, or the nature of the votes given. Such a rule, I admit, would be easy of application, and as a general rule might not be productive of a great amount of injustice, while the multitude of distinct questions of fact in reference to the great number of voters whose qualification may be contested, is liable to lead to some embarrassment, and sometimes to protracted trials, without a more satisfactory result than would have been attained under a rule which should exclude all such inquiries. Still I cannot avoid the conclusion, that in theory and spirit our constitution and our statutes recognize as valid those votes only which are given by electors who possess the constitutional qualifications; that they recognize as valid such elections only as are effected by the votes of a majority of such qualified electors. And though the election boards of inspectors and canvassers, acting only ministerially, are bound in their decisions by the number of votes deposited in accordance with the forms of law regulating their action, it is quite evident that illegal votes may have been admitted by the perjury or other fault of the voters; and that the majority to which the inspectors have been constrained to certify and the canvassers to allow, has been thus wrongfully and illegally secured. And I have not been able to satisfy myself that in such a case, these boards acting thus ministerially, and often compelled to admit votes which they know to be illegal, were intended to constitute tribunals of last resort for the determination of the rights of parties claiming an election. If this were so, and there were no legal redress, I think there would be much reason to apprehend that elections would degenerate into mere contests of fraud.

The person having the greatest number of the votes of legally qualified electors, it seems to me, has a constitutional *312right to the office; and if no inquiry can be had into the qualification of any voter, here is a constitutional right depending upon a mode of trial unknown to the constitution, and as I am strongly inclined to think, opposed to its provisions. I doubt the competency of the Legislature, should they attempt it, which I think they have not, to make the decision of inspectors or canvassers final under our constitution.

The extent of the inquiry into the qualification of voters, and how they have voted, may be limited or qualified by other provisions of the constitution; and I think it has been thus practically limited or modified by that provision of the constitution which requires all votes to be given by ballot. The object of this requirement, when considered with reference to the history of our country and the whole theory of popular governments, where suffrage is practically universal, is too plain to be misunderstood. It was to secure the entire independence of the electors, to enable them to vote according to their OAvn individual convictions of, right and duty, without the fear of giving offense or exciting the hostility of others. And with this view the right is secured to eveiy voter of concealing from all others, or from such of them as he may choose, the nature of his vote, or for what person or party he may have voted. This important object, vital as I think it is in our system of government, would be substantially defeated if the voter could be compelled to disclose, even in a court of justice, how he has voted. The constitution, and our statutes which have followed out its spirit, have thrown over the voter an impenetrable shield, under which he may keep the secret of his vote until he shall see fit to disclose it. The statute has taken pains to guard this secret from the prying curiosity of others, at the polls, when from the necessity of the case the outside of the ballot must be seen. The ticket is required to be handed to the inspector folded, and the inspector is required to deposit it in the box, without opening or *313permitting it to be opened. And it -would effectually defeat the object of this statute if any evidence were allowed to be given of the appearance of the outside of the ballot when thus folded, for the purpose of raising an inference how the elector voted. But there is still another objection to this species of evidence. The fact that the ticket had the appearance of a Democratic or Republican ticket, does not, when so many candidates are voted for on the same ticket, even tend to show that the elector voted for any particular candidate, especially as it appears that great numbers of the tickets on both sides were “split tickets,” having slips with the names of candidates of the opposite party from that indicated by the outside of the ticket.

How an elector may have voted is, under the constitution and the law, a fact which no man has a right to learn, in this or any other manner, till the elector himself may choose to make it public. No man can acquire the right to disclose it without his consent; and least of all, can any knowledge surreptitiously obtained, be given in evidence, as by looking over the voter’s shoulder when he is preparing his ticket, or showing it to a friend; or by an inspector in violation of the statute receiving a ticket unfolded and reading its contents, as in the case-of Oabot.

It must also be borne in mind that by the very spirit of the system of voting by secret ballot, the legal right of deceiving others as to the vote, is implied; and, therefore, as well as because it would be mere hearsay, his mere confession or admission to another person must be incompetent evidence.

But careful as the law has been to guard the secret of the elector’s vote till he sees fit to disclose it, that secresy is, it seems to me, in its nature and purpose, a personal privilege only; and may be waived by the elector at his option. He may, if he see fit, testify in court to the vote which he has given. And though the statute requires the voter to hand in his ballot folded, I am not entirely satisfied *314that it is not competent for him at the time of voting to waive the privilege of secresy; as, for instance, by coming up to the polls and in the presence of the inspectors and bystanders, declaring his intention to make his vote public; openly exhibiting his ticket, publicly folding it, and handing it in. In such a case there may be good ground for holding that he has thereby waived the privilege of secresy.

It is true that it can seldom be a question of any importance, in a case like the present, how a person possessing the constitutional qualifications of an elector has voted; since generally the vote of such person is legal, and must be counted at all events. Still there may be cases in which a person possessing the constitutional qualifications can not legally vote; as when he has neglected to register according to the statute.

The question of privilege becomes much more important when applied to the case of a voter whose qualifications are denied. It is clear that this privilege of secresy is given only to those who are recognized by the constitution as electors. And whenever the person who has voted admits that he was not constitutionally qualified, or the fact clearly appears, so that it no longer remains as a question for the jury, he can claim no protection from this privilege. But when his qualification is a disputed question of fact, which is to be determined by the jury in rendering their verdict, he can not be compelled to disclose his vote.

As to the excess of nine ballots in the Sixth, and seven in the Seventh Ward, beyond the number of votes shown by the poll lists, we are all agreed that, for this excess, the whole vote of these wards is not to be rejected. The electors should not be disfranchised for this error of the inspectors or clerks.

But I can not concur with my brother Campbell that because this excess is greater than the majority for either candidate for sheriff, the whole excess must, in effect, be *315doubled, by deducting tbe full number from tbe vote of each.

It was tbe duty of the inspectors under tbe statute— Qomp. L, §§ 61 and 62 — to have discovered this excess before canvassing the votes or opening tbe ballots; and they were required to draw out from tbe box a number of ballots unopened equal to tbe excess. This provision of tbe statute evidently proceeds upon tbe doctrine of tbe calculation of chances — tbe probability that tbe ballots drawn out will be apportioned among tbe respective parties and candidates substantially according to their respective numbers of votes.

This duty having been neglected by the inspectors, it devolves, I think, upon the court in a trial of the right to an office under this section, to apply to this excess of votes substantially tbe same rule as that prescribed for the inspectors; in other words, a rule which will distribute this excess among the opposing candidates with tbe same degree of certainty, and in tbe like proportion which the rule prescribed by the statute was calculated to produce.

Thus apportioned, taking the two wards, seven and a small fraction would fall to the share of Williams, six and a larger fraction to Cicott, and one and a large fraction to Allison. But, as no fraction of a vote can be counted, eight of this excess must be deducted from the vote of Williams, and seven from that of Cicott.

It remains only to apply these principles to the facts of this case.

Taking tbe inspectors’ returns, as corrected by tbe recount, without reference to any of the specially disputed ballots, and without reference to the qualifications of any voter, and tbe total vote for relator is 6,530; respondent’s, 6,515'; relator’s majority, 15. In, this calculation one vote is taken from respondent for a double ballot in Grosse Point, erroneously counted as one, when both should have been rejected; and one is added to his vote in Sumpter, *316for a ballot wrongfully rejected, because folded in a ballot for the constitutional amendment. Deducting now, from relator’s vote eight, and from respondent’s seven, for their respective portions of the excess in the Sixth and Seventh Wards, and relator’s majority will be reduced to fourteen.

Supposing the jury to have found (as it was competent for them to find) in favor of the inspectors’ certificate, instead of the recount in Dearborn, this would take from the relator eight votes, leaving his majority but six. From this deduct the following votes, shown by the testimony of the voters themselves, to have been given for the relator, and who may have been found by the jury to have been disqualified, viz: Chase, who only says he voted “ Republican ticket,” but who, from the circumstances and the course of examination, would probably have made the exception as to sheriff, if he had voted for the candidate of the opposite party; Burtis, who says directly that he voted for Williams, and Edwards, who only says he voted “Republican ticket,” but admits that if he had voted against Williams he would have been likely to remember it. These three votes deducted from relator’s majority of six, reduce it to three. • These are all the voters claimed to have been unqualified and to have voted for Williams, who are shown by any legal evidence to have voted for him. On the other hand, it appears clearly from the evidence, and is not controverted, that Brennan, who swears he voted for Cicott, was not a voter, not having been a resident of the state for the three months next preceding the election. There was no evidence before the jury from which they could have found him a voter. This was so clear that the respondent’s counsel did not contest the point. This, therefore, reduces respondent’s vote by one, making relator’s majority four.

Allowing, therefore, that the jury upon every contested point in the case, resolved every doubt arising upon all the competent evidence in the case in favor of the respondent and against the relator, the latter was elected by four *317majority. There was no competent evidence in the case to warrant the verdict given, or upon which it can be sustained.

The verdict should, therefore, be set aside and a re-trial ordered. But as the case has already been once tried in the Wayne Circuit, and its facts may have become matters of public notoriety there; and since the case was sent there for trial one of the counsel in the case has become the judge of that circuit, and no other judge is required by law to hold a court there for the trial of this case, and it is desirable that an early trial should be had, we have concluded to send the issues to the Circuit Court for the county of Washtenaw'for trial.

Graves • J. concurred. Cooley Ch. J.

I agree fully with all that is said by my brothers Campbell and Christiancy as to the inadmissibility of evidence to show the nature of an elector’s ballot, in any case where it is not first shown that he waived his constitutional privilege of secrecy. And I concur generally in the conclusions of my brother Christiancy upon other points, except as I shall otherwise indicate.

I regret that my brethren are disposed to still follow the case of People v. Tisdale, 1 Doug. Mich. 59, notwithstanding the majority are of opinion that it is unsound in principle. The case has no support, as I think, either in the authorities or in the analogies of the law; and no court outside the state has ever followed it. It is true, as my brethren have remarked, that it lays down a rule easy of application, and one that is no more unfair to one candidate than to another; but it does not seem to me to be sufficient reason for retaining an unsound rule that it is impartial in its infliction of injustice. In every case where it becomes important to apply the rule at all, it has the effect to defeat the clearly expressed will of the electors; *318and although, if we were to consider voters only as aggregated in two political parties, there might be no reason to complain of it, because it would be likely to defeat one party as often as the other, yet the considerations are very different when we reflect that election contests are between individuals, and that that person is irremediably wronged who, having received a plurality of the suffrages of his fellow citizens, is deprived of his office by the application of some arbitrary and technical rule. It can be little compensation to him that, in the next election, a similar application of the rule deprives a political opponent of an office to which he is fairly chosen, and thereby balances the accounts between the parties by doing equal injustice to an individual member of each. All rules of law which are applied to the expression, in constitutional form, of the popular will, should aim to give effect to the intention of the electors ; and any arbitrary rule which is to have any other effect, without corresponding benefit, is a wrong, both to the parties who chance to be affected by it, and to the public at large. The first are deprived of their offices, and the second of their choice of public servants.

The chief argument in favor of the rule of People v. Tisdale is, that ballots cast for parties by their initials only are so uncertain that they cannot be applied without resort to extrinsic and doubtful evidence, to ascertain the voter’s intention, and therefore should be rejected. But nothing can be more fallacious. It frequently happens that a man is better known by the initials of his baptismal name, than by the name fully expressed; simply because he is not in the habit of writing his name in full, or of being thus addressed in business transactions. I think it highly probable that that is the case with each of the parties before us. In political conventions, or legislative bodies, no one deems it important to write the full name of a candidate for whom he is voting, and no one ever thinks of challenging the vote for uncertainty. ITnder the application of this rule to the *319present case, the curious spectacle will be exhibited, of votes cast for E. V. Cicott and G. 0. Williams being rejected because the courts cannot determine for whom they were intended, while not a single person in the county of Wayne has the slightest doubt that they were cast for Edward V. Cicott and Gurdon 0. Williams, the opposing candidates at this election. Thus the courts are required to close their eyes to what everybody else can see distinctly.

The fallacy of the rule consists in its assuming that a certain form of ballot clearly expresses the voter’s intention, while another form is so uncertain that it is dangerous to attempt to arrive at the meaning by evidence. But in fact no ballot can identify with positive certainty the persons for whom it is cast; and notice must be taken of extrinsic circumstances in order to apply it. It is always possible that other persons may reside in the election district having the same names with some of the candidates; but neither the canvassers nor the courts ever assume that there is any difficulty in these cases, but they count the votes for the persons who have been put forward for the respective offices. And in some cases where an element of uncertainty is introduced into the ballot unnecessarily, as by the addition of an erroneous designation, the courts resolve the difficulty by rejecting the erroneous addition, and counting the ballot for the person for whom it was evidently designed.

If the rule were one which the canvassers could apply in every case, and which left nothing open for controversy in the courts afterwards, it would be less open to objection; but it is not of that character. No one doubts that if votes had been cast in this case for Edward Cicott, or Edward B. Cicott, or Edward Cicott, junior, or Edward Oicott with any other mistaken addition, they must have been counted for the respondent on the facts appearing in this case. — People v. Cook, 14 Barb. 259, and 8 N. Y. 67; Milk v. Christie, 1 Hill, 102; Bratton v. Seymour, 4 Watts, 329. Such ballots would be allowed because the error of *320the voter is not so great, when tbe facts surrounding tbe election are considered, as to leave his intent in real doubt; yet no one can fail to see that in every one of these cases the room for doubt is greater than in the case of ballots for E. Y. Cicott, and that whatever doubt exists is referred to courts and juries for solution in the one case as it would be in the other. The rule therefore rejects a certain class of ballots on reasons which apply with at least equal force to others which are admitted. And its indefensible character is still more apparent when we consider that abbreviated ballots are received, as well as those which are mis-spelled; so that a vote for Ed. Sekut would be counted, though the ÍM. is an abbreviation for several other names besides Edward, and no one would suppose that by this name the person intended was as distinctly pointed out as if the name had been written as it commonly is in business transactions.

The true rule upon this subject I conceive to be this: Evidence of such facts as may be called the circumstances surrounding the election — such as, who were the candidates brought forward by the nominating conventions; whether other persons of the same name resided in the district from which the office was to be filled, and if so, whether they were eligible to the office and were publicly named for it, and the like — is always admissible for the purpose of aiding in the application of any ballot which has been cast: and where the intent of the voter as expressed by his ballot, when considered in the light of such surrounding circumstances, is not doubtful, the ballot should be counted and allowed for the person intended. This rule is just and easy of application; and it has the merit of harmonizing with the rules applied to other written instruments, which I think is no slight recommendation. It is always objectionable and mischievous to lay down different rules for classes of cases which all come within the same reasons.

On the question whether it is competent in this proceeding to inquire into the qualifications of those who voted, I *321concur with, my brothers Christiancy and Graves, but solely on the ground that I think it more important that a rule should be laid down, than that that rule should be either one way or the other.

The only remaining question which I deem it important to refer to, is as to the effect of the failure of the canvassers in two of the wards of Detroit to draw from the ballot box the surplus of votes over the number which should have been found there as shown by the poll-lists. The county canvassers rejected the returns from those wards because of this irregularity. This they had clearly no authority to do. Their duties in canvassing are ministerial, and they are to receive the returns transmitted to them, if in due form and from the proper source, as correct, and to ascertain and declare the result as shown by such returns, without attempting to inquire into and settle any question which may lie back of those returns and affect the result. Ex parte Heath, 3 Hill, 42; Brower v. O’Brien, 2 Ind. 423; People v. Hilliard, 29 Ill. 413; People v. Jones, 19 Ind. 357; Bacon v. York County Commissioners, 13 Shep. 491; Mayo v. Freeland, 10 Mo. 629 ; Thompson v. Circuit Judge, 9 Ala. 338; People v. Kilduff, 15 Ill. 492; O’Farrell v. Colby, 2 Minn. 180; People v. Van Cleve, 1 Mich. 362; People v. Van Slyck, 4 Cow. 297; Dishon v. Smith, 10 Iowa, 212; Attorney General v. Ely, 4 Wis. 420; State v. Governor, 1 Dutcher, 331; People v. Cook, 8 N. Y. 67. The cases on this point are too numerous and uniform to be further cited.

Such canvass and declaration, however, would not be conclusive upon any party who might be injuriously affected by any antecedent irregularity. The question here is whether any one is injuriously affected; and this involves a consideration of the purpose of this drawing as well as of the probable result had the statute been complied with.

I do not suppose that the statute, in providing for this drawing, proceeds upon the idea that because a surplus of ballots appears there is necessarily fraud. The keeping of poll *322lists is required as a guard, cheek and protection, and the drawing is provided for because it is supposed, when the count of the boxes and lists disagrees, the latter are more likely to be correct. But any one who has ever watched the conduct of an election in a populous precinct, where the clerks are very busily employed through the day in writing down the names of voters as they are called off to them, has not failed to observe how liable the call is to be in advance of the writing, so that it is no matter of surprise when a name fails to be taken down. And where such a failure occurs, it is quite likely to be made by both clerks, since those officers sit side by side and rely more or less upon each other’s lists as they proceed. Nevertheless when the discrepancy occurs, the poll lists must govern, as least liable to err, and least open to fraud; and the statute makes the best disposition of the case which the wisdom of the Legislature has been able to devise.

What is that disposition? Here it appears there are a number of ballots in the box which the statute presumes ought not to be counted. Tet there is no mark upon them, and no possible way of distinguishing them from the ballots duly cast. It is quite impossible, therefore, to separate the good from the bad. Shall the election be declared void for that reason? This would be simply to disfranchise the electors because in some unexplainable manner a fraud, accident, or mistake has occurred for which they are in no way responsible, and which may not at all affect the result. The statute says they shall not be thus unjustly disfranchised, but that the votes shall be disposed of by'a rule which will operate as equally and justly between the parties as is possible under the circumstances. That rule is based upon the doctrine of probabilities. It directs the drawing from the box of a number equal to the surplus found there, and the equalization of the count of the box and the lists in that mode.

*323Now it is apparent that in this drawing the chances are very much greater that ballots duly cast will be drawn and destroyed than that the ballots wrongly in the box will be drawn, for the plain reason that the legal votes will always greatly exceed the surplus in number, unless there has been gross fraud on the part of the inspectors. If the surplus is one in a vote of a hundred, the probabilities are ninety-nine to one that the vote drawn will be one of those which was regularly and legally cast, and that thereby a legal voter will be deprived of his suffrage. Nevertheless, as each ballot is usually one of a number designed to be allowed to particular candidates and counted against others given to other candidates, the drawing may still work no injustice, since each candidate will probably lose by it a number proportioned to the relative number of ballots appearing for him in the box, and thus the relative proportions will be preserved. The actual drawing might differ from the antecedent probability; but this is the theory which the statute proceeds upon, and the variance can seldom be great.

Now the relative proportion of votes cast for the two candidates in the two wards in question was such that, had the drawing taken place, Mr. Williams would probably have lost by it one more vote than Mr. Oicott, and no more. The probability is, therefore, that the election, as between these two candidates, has not been affected by the irregular action of the inspectors, beyond a single vote. Can it be held void for this irregularity if the actual difference in favor of Williams is more than a single vote?

Upon this point I refer to the following decisions where, in a great variety of cases, it has been held that statutory provisions prescribing the conduct of elections are to be regarded as directory only, except where they are of such a character that a failure to comply with them would have the effect to prevent or obstruct the complete expression of the popular will, or the production of satisfactory evidence thereof.— People v. Cook, 14 Barb. 259 and 8 N. Y. 67; People *324v. Vail, 20 Wend. 14; Clifton v. Cook, 7 Ala. 114; Dishon v. Smith, 10 Iowa, 212; Attorney General v. Ely, 4 Wis. 420; State v. Jones, 19 Ind. 356; People v. Van Cleve, 1 Mich. 65; People v. Higgins, 3 Mich. 233; People v. Bates, 11 Mich. 362; People v. Sackett, 14 Mich. 320; Taylor v. Taylor, 10 Minn. 112; People v. McManus, 34 Barb. 620; Whipley v McCune, 12 Cal. 352; Piatt v. People, 29 Ill. 54; Ex-parte Heath, 3 Hill, 42; Lanier v. Gallatas, 13 La. An. 175; 2 Strong, Petitioner, 20 Pick. 492.

And even where the statutory provisions are mandatory, they do not necessarily defeat an election actually held, if the means exist of determining the result. A particular act or proceeding may be rendered void, and the election be upheld notwithstanding.

The principle in all these cases is, that an election is not to be set aside because of air irregularity, unless it appears that that irregularity affected the result. In well reasoned cases where the election was disputed on the allegation that the polls were kept open for the reception of votes after the hour when the statute directed them, to be closed, it has been declared that, on general principles, no election could be made void on such grounds, unless it appeared that the votes received after the legal hour of closing changed the result. — Pratt v. People, 29 Ill. 72; People v. Cooke, 14 Barb. 296; Same case, 8 N. Y. 91. And in Lanier v. Gallatas, 13 La. An. 176, where it was charged that fourteen illegal votes were received, and that “if the said illegal votes had been rejected, it would in all probability have changed the result in favor of the petitioner,” the court hold the averment insufficient, and say that “to contest an election there should be an averment that the illegalities charged did alter the result; not that it was probable that the result might have been changed.”

If these authorities are sound — as they unquestionably are — we can not be warranted in declaring an election void *325for an irregularity wbiob probably did not change the result. The courts have always repelled “the idea that the will of the electors, plainly expressed in the forms prescribed by law, can be defeated by the negligence, mistake, or fraud of the officers appointed to register the result of an election.” People v. Cook, 14 Barb. 327; People v. Vail, 20 Wend. 14.

Suppose it be conceded that these sixteen votes are illegally in the boxes, and that they can not be distinguished and separated, the case then is no different from what it would be had these votes been cast by sixteen unqualified voters who refused to disclose for whom they voted. In such a case are the votes to be counted, first against one candidate, and then against the other, in order, if possible, to defeat both? If so, then an important election will seldom occur in close districts where, as to some officer there will not be a failure to elect. Sqch a rule could be based upon no reasons of justice or public policy which suggest themselves to my mind, and I do not discover anything in in the authorities to warrant it. I refer, in addition to the other cases cited, to First Parish in Sudbury v. Stearns, 21 Pick. 148, and Blandford School District v. Gibbs, 2 Cush. 39, as bearing upon this question. Ex-parte Murphy, 7 Cow. 153, covers the whole ground of the present case on this point. The majority in that case was only two, and it affirmatively appeared that two votes had been put into the box by unknown persons in the names of electors who were dead. It was not known for whom they were cast. On motion for leave to file an information, the court say: “ The motion must be .denied. For aught that appears, the spurious ballots were for the ticket which was in the minority. To- warrant setting aside the election, it must appear affirmatively that the successful ticket received a number of improper votes, which, if rejected, would have brought it down to a minority. The mere circumstance that improper votes were received, will not vitiate an election. If this rule were otherwise, hardly an election in the state could be sustained.” *326The general rule is thus applied to a case where the fraudulent vote equals the whole majority; in which particular alone could the case now before us be distinguished from many others I have cited.