Allan J. Stevo v. John R. Keith, et al., in their capacities as members of the Illinois State Board of…

In the
United States Court of Appeals
For the Seventh Circuit
No. 08-3218
ALLAN J. STEVO,
Plaintiff-Appellant,
v.
JOHN R. KEITH, et al., in their capacities as
members of the Illinois State Board of Elections,
Defendants-Appellees.
Appeal from the United States District Court
for the Central District of Illinois.
No. 08-3162—Richard Mills, Judge.
ARGUED SEPTEMBER 25, 2008—DECIDED OCTOBER 1, 2008
Before CUDAHY, POSNER, and FLAUM, Circuit Judges.
POSNER, Circuit Judge. The plaintiff wants to run as an
independent candidate for the U.S. House of Representa-
tives from Illinois’s Tenth Congressional District, which
encompasses parts of Lake and Cook Counties. His petition
to appear on the November 2008 general election ballot
was challenged by a local resident because the plaintiff
had failed to submit with his petition the requisite mini-

-- 1 of 8 --

2 No. 08-3218
mum number of signatures of persons qualified to vote
in the election. That number is 5 percent of the number of
people who voted in the district in the last congressional
election. 10 ILCS 5/10-3. For the Tenth Congressional
District, the minimum required number of valid signa-
tures is 10,285, and the plaintiff claims that he had more
than 7,200. The Illinois State Board of Elections ruled
that he had only 6,978 valid signatures; in any event he
does not claim that his “more than 7,200” reached 10,285.
The plaintiff claims that the 5 percent requirement denies
equal protection of the laws and infringes First Amend-
ment rights to stand for public office and to vote for the
candidate of one’s choice. Jenness v. Fortson, 403 U.S. 431,
434 (1971); Bullock v. Carter, 405 U.S. 134, 143 (1972); James
A. Gardner, “Deliberation or Tabulation? The Self-Under-
mining Constitutional Architecture of Election Cam-
paigns,” 54 Buff. L. Rev. 1413, 1432 (2007). A candidate
denied a place on the Illinois ballot can, it is true, con-
duct a write-in campaign. “State of Illinois Candi-
date’s Guide 2008” 41 www.elections.il.gov/Downloads/
ElectionInformation/PDF/08CanGuide.pdf (visited Sept. 30,
2008). But that is an inferior alternative to having one’s
name on the ballot. U.S. Term Limits, Inc. v. Thornton, 514
U.S. 779, 830-31 (1995); Lubin v. Panish, 415 U.S. 709, 719
n. 5 (1974).
The district court dismissed the complaint for failure to
state a claim. The appeal relies mainly on equal protection;
the First Amendment claim is not developed.
All congressional districts must be redistricted after
each decennial census. Georgia v. Ashcroft, 539 U.S. 461, 488

-- 2 of 8 --

No. 08-3218 3
n. 2 (2003); Ill. Const. art. IV § 3; 10 ILCS 76/1-76/99; Adam
Cox, “Partisan Fairness and Redistricting Politics,” 79
N.Y.U. L. Rev. 751, 752-53 (2004). And for the first election
after the decennial census Illinois requires independent
candidates to obtain only 5,000 valid signatures from
qualified voters, rather than the 5 percent required in the
other elections. 10 ILCS 5/10-3; Libertarian Party v. Rednour,
108 F.3d 768, 771 (7th Cir. 1997). In one congressional
district, the Fourth, 5,000 is more than 5 percent of the
voters in the last (2006) congressional election (5 percent
in that district is only 4,293), but in the others it is less
and in the Tenth Congressional District, with its require-
ment of 10,285 signatures, 5,000 is less than half of
5 percent of the votes cast in the last election.
The average number of required signatures per district
is 9,442. The Fourth, with only 4,293, is an outlier; the
next lowest is the Fifth, with 7,713. The highest is the
Nineteenth, with 12,205. (The source of these figures
is “Signature Requirements and Forms—U.S. Representa-
tive in Congress,” www.elections.state.il.us/Downloads/
ElectionInformation/PDF/usrep.pdf (visited Sept. 30,
2008).) Although the Supreme Court’s reapportionment
jurisprudence requires that congressional districts be of
equal population, Karcher v. Daggett, 462 U.S. 725, 730-31
(1983); Wesberry v. Sanders, 376 U.S. 1, 17-18 (1964), the
5 percent rule produces different numbers in different
districts without violating the Constitution because it is
5 percent of the number of persons who actually voted in
a district, not 5 percent of the district’s population.
The last decennial census was in 2000, and no Illinois
congressional district was redistricted after the most recent

-- 3 of 8 --

4 No. 08-3218
congressional election, held in 2006. But the plaintiff
argues that there is no difference between a newly redis-
tricted district and a district that is unchanged since the
last election, and therefore the state’s judgment that
5,000 signatures is enough in a newly redistricted district
proves that 5 percent is too stringent a requirement in
any district in which the 5 percent formula yields a re-
quirement of more than 5,000 signatures, such as the Tenth.
The Supreme Court has held that 5 percent is a permissi-
ble minimum signature requirement for placing third-party
or independent candidates on the ballot, Jenness v. Fortson,
supra, 403 U.S. at 439-41; Nader v. Keith, 385 F.3d 729, 733
(7th Cir. 2004), provided that there is not only a write-in
alternative but also other means of getting one’s candidacy
before the electorate, such as finding sponsorship by a
political organization, Jenness v. Fortson, supra, 403 U.S. at
438; Hall v. Simcox, 766 F.2d 1171, 1174 (7th Cir. 1985), and
provided also that the state does not impose “suffocating
restrictions” on ballot access. Jenness v. Fortson, supra, 403
U.S. at 438. Illinois does not. See Nader v. Keith, supra, 385
F.3d at 734-35; 10 ILCS 5/10-6, 5/10-8 to 5/10-10. But the
plaintiff argues that Illinois’s disparate treatment of the
two types of district shows that a 5 percent minimum is
arbitrary, at least in Illinois. He is using the 5,000-signa-
tures provision of the law just to show that if it is good
enough in newly redistricted districts, it is good enough
in all districts.
The state defends the disparity in treatment between the
two types of district on the ground that it is impossible to
calculate a percentage of the votes in the previous election

-- 4 of 8 --

No. 08-3218 5
in a redistricted district because by definition there was
no previous election in that district—the district didn’t
exist. But vote totals are reported for each precinct, and, if
a district’s boundaries are changed, the votes cast in the
previous election in the precincts within the new bound-
aries can be added up and the sum will be the total number
of votes that were cast in that area in the last election.
Indiana does that in determining, when a district is
redistricted, the number of votes that was cast in the last
election in the area embraced by the new district; that
number is then multiplied by the required percentage of
valid signatures to determine whether the candidate has
the signatures of enough qualified voters to get on the
ballot. See Indiana Code §§ 3-8-6-3(a), 12(b)(7), (c), (d).
But this suit fails even if Indiana’s procedure for deter-
mining the previous turnout in the area enclosed by new
district boundaries is entirely feasible (as we’ll assume,
though district boundaries sometimes slice through
precincts and the turnout in each slice could not be deter-
mined with precision or perhaps at all). Redistricting is
a disorienting event for voters and candidates alike,
since it changes the electorate, usually with an eye to
improving the electoral prospects of the majority party in
the legislature doing the redistricting. Gaffney v. Cummings,
412 U.S. 735, 753-54 (1973); Shaw v. Reno, 509 U.S. 630, 661
(1993) (White, J., dissenting); La Porte County Republican
Central Committee v. Board of Commissioners, 43 F.3d 1126,
1130 (7th Cir. 1994). Candidates and voters alike must
adjust to the new political landscape. It is plausible that
it would be more difficult for candidates to obtain signa-

-- 5 of 8 --

6 No. 08-3218
tures in such circumstances, and so the required number
is reduced.
The method that Illinois uses to reduce the required
number is crude, however, because districts vary in the
number of people who voted in the last election. Hence
the anomaly of the district in which the requirement of
5,000 signatures is more demanding than the 5 percent
requirement, making it more difficult for candidates to
obtain the required number of signatures in election
years in which the district has been redistricted even
though the objective of requiring “only” 5,000 signatures
in new districts is to make the process of qualifying
for a place on the ballot less difficult (albeit it is more
difficult in only that one district). But the plaintiff’s
proposal—a requirement in every election of just 5,000
signatures—is as or more arbitrary, since, other things
being equal, it is easier to obtain a specified number of
signatures the larger the turnout in the district in the
preceding election and hence the smaller the percentage
of qualified voters whom the candidate is required to sign
up to reach the minimum number and so get onto the
ballot. There are more fish in the pond, so it is easier to
catch the required number.
Granted, other things may not be equal. The more
signatures that are required, the greater the cost, and the
difference may be important for a minor-party or inde-
pendent candidate. Hall v. Simcox, supra, 766 F.2d at 1174.
But at best this point makes the choice between num-
ber and percentage a standoff; it does not justify invalidat-
ing the percentage approach.

-- 6 of 8 --

No. 08-3218 7
Suppose the Indiana approach, whereby disaggregation
of votes to the precinct level in the preceding election
enables the same percentage to be required in districts
that have recently been redistricted and districts that
have not been, is indeed a compellingly superior approach;
nevertheless the plaintiff would not have standing to urge
its adoption because he could not show harm. The Indiana
approach is a solution to the problem of determining
previous voter turnout after a redistricting, and the
Tenth Congressional District is not a newly redistricted
district. So the plaintiff is forced to argue that a uniform
rule requiring 5,000 signatures (or a slightly higher num-
ber, provided it is below the number of valid signatures
he was able to obtain) is so far superior to the present
system that the Constitution requires that it be sub-
stituted for it. That is wrong, given the disparity in voter
turnout in the different districts. We have just seen that
requiring a number rather than a percentage may well
be a bad feature of Illinois’s treatment of elections in
recently redistricted districts; the plaintiff urges us to
impose that quite possibly inferior method in all elec-
tions, rather than, as at present, in just the elections
in newly redistricted districts.
We warned in Crawford v. Marion County Election Board,
472 F.3d 949, 954 (7th Cir. 2007), affirmed, 128 S. Ct. 1610
(2008), against federal judicial micromanagement of state
regulation of elections. See also Clingman v. Beaver, 544
U.S. 581, 593 (2005); Burdick v. Takushi, 504 U.S. 428, 432-34
(1992); Storer v. Brown, 415 U.S. 724, 729-30 (1974). But
the appeal in this case fails on a more basic level: the
change the plaintiff asks us to make in the Illinois voting

-- 7 of 8 --

8 No. 08-3218
system might well make that system more arbitrary than
it already is.
AFFIRMED.
10-9-08

-- 8 of 8 --