(dissenting)-
I dissent. The burdens placed upon voters by the two-year purge statute to obtain reinstatement are coneededly not onerous, nevertheless, I conclude that the burden, light as it is, is not constitutionally permissible.
The Supreme Court has held that statutes placing restrictions on the right to vote can be sustained only upon a showing that they are necessary to promote a compelling state interest. City of Phoenix v. Kolodziejski, 399 U.S. 204, 90 S.Ct. 1990, 26 L.Ed.2d 523 (1970); Cipriano v. City of Houma, 395 U.S. 701, 89 S.Ct. 1897, 23 L.Ed.2d 647 (1969); Kramer v. Union Free School District, 395 U.S. 621, 89 S.Ct. 1886, 23 L.Ed.2d 583 (1969). Certainly prevention of voter fraud is a compelling state interest. Where I part company with the majority is on the matter of whether there has been here demonstrated a relationship between voter fraud and failure to vote.
There is a surface validity to defendants’ contention that the two-year purge statute provides a safeguard against “phantom” voters. The presence on a voter registration list of the names of persons no longer eligible to vote undoubtedly creates a potential for fraud, i.e. the possibility that evil persons with knowledge of the true facts may attempt to cast a “phantom” vote in the name and place of the non-existent voter. It also appears reasonable to assume (and we so noted in ruling on the preliminary injunction), that one who has failed to vote for two years is either no longer a resident, or if a resident, is no longer *654interested in voting. The evidence presented, however, does not bear out those surface appearances. As noted in the majority opinion, the survey indicated that 58.5 % of the persons in the sample stricken for non-voting were found to be actually still living at the place from which they had last voted. In 1971, approximately 11,000 persons responded to the first notice and an additional 2,000 responded to the second notice. If the numbers of those responding to the notices to strike were added to the projections from the survey it would appear that somewhat in excess of 64% of the persons to whom notices were sent for failure to vote still maintained the same residence. It is quite clear therefore that non-voting for two years is not a valid indicator of non-residence.
As for the inference of lack of interest, the evidence presented casts considerable doubt on that as well. The two-year non-vote purge produces approximately three times as many strike offs in the third year following a presidential election as there are in other years.1
The indication is quite clear, therefore, that many persons, a substantial number, exercise the right to vote only in presidential elections. I cannot say that such voters are “in fact substantially less interested.” Kramer v. Union Free School District, supra, at 632, 89 S.Ct. 1886.
In my view the defendants have presented no evidence whatsoever of a relationship between vote fraud and non-voting. The majority has noted that “the two-year period allows removal of the names before the political operatives can take advantage of the situation.” I am not aware of the factual basis for that statement and I disagree with the conclusion. It appears to me that if, indeed, a no longer eligible voter’s name is used by a “phantom” to east a fraudulent vote, that very fact will prevent disclosure of the fact of ineligibility through the non-vote purge provision. As to the majority’s statement that “such considerations led the state legislature to change the purge period from four to two years in 1941 after considerable public concern over voting fraud,” the only “evidence” I find in support of that is a statement by counsel for the defendants. There is no legislative finding, to my knowledge, in the statute itself or in any legislative history. The fact, of which the majority took judicial notice, that in 1941 a Special Grand Jury found that there were 50,000 ineligible voters on the Philadelphia registration lists does not cast any light on the relationship, if any, between non-voting and voter fraud, which is the issue before us.
In any event, even if evidence had been produced to establish that removal from the voting lists of the names of those who fail to vote for two years does tend to prevent voter fraud, the procedure does not accomplish the purpose with sufficient precision to justify denial of the right to vote to those who are still bona fide residents of the voting district. The survey of the sample of 491 persons reveals that the two-year non-vote purge identified only 144 persons (29.4%) who had moved, died or were otherwise not eligible to vote. From a projection of that information, it appears that 82,352 persons were stricken from the voter registration list by the two-year purge statute in order to effect the removal of only 24,210 who *655were no longer eligible to vote, a ratio of about 3y2 to 1. Such a procedure is too sweeping and does not conform to the exacting standard of precision required of statutes which selectively distribute the franchise. Kramer v. Union Free School District, supra; see Shapiro v. Thompson, 394 U.S. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969); United States v. Robel, 389 U.S. 258, 88 S.Ct. 419, 19 L.Ed.2d 508 (1967). At best, it should serve only as a basis for further investigation through the other statutory means ¿vailable (e.g. 25 P.S. § 623-31 (Official Report); § 623-32 (Mail Canvass); § 623-33 (Inspectors’ Canvass); § 623-35 (Strike-off Petition)) to ascertain whether, in fact, the voter is no longer eligible by non-residence or otherwise.
I would hold that the two-year non-vote purge provision (25 P.S. § 623-40) of “The First Class City Permanent Registration Act” of Pennsylvania is unconstitutional as imposing an unnecessary burden on the right to vote.
. 1967 1968 1969 1970 1971 First notices 94,000 35,000 31.000 33,000 92,000 Second notices 80,000 31,000 29.000 31,000 81,000 Williams v. Osser, supra, 326 F.Supp. at p. 1143.