Gaffney v. Cummings

LOCATION:American Trust & Security Company

DOCKET NO.: 71-1476
DECIDED BY: Burger Court (1972-1975)

CITATION: 412 US 772 (1973)
DECIDED: Jun 18, 1973
ARGUED: Feb 26, 1973 / Feb 27, 1973

Robert G. Dixon, Jr. – for appellant
Robert A. Satter
Robert Satter – for appellees

Facts of the case


  • Oral Argument – February 27, 1973
  • Audio Transcription for Oral Argument – February 27, 1973 in Gaffney v. Cummings

    Audio Transcription for Oral Argument – February 26, 1973 in Gaffney v. Cummings

    Warren E. Burger:

    — arguments next in 71-1476, Gaffney against Cummings.

    Mr. Dixon you may proceed.

    Robert G. Dixon, Jr.:

    Mr. Chief Justice, may it please the Court.

    This case is here on direct appeal from a three-judge District Court in Connecticut which invalidated the official state legislative enforcement plan.

    We appear here in defense of that official state plan.

    The plan was devised by a bipartisan board of three members including a tie-breaker under the state constitution which constitution thereby tries to incorporate an anti-gerrymandering spirit in the very reprocess of the apportionment activity.

    The plan of the board was subjected to litigation promptly in both state and federal court.

    The federal court denied a plea of extension and on April 4, 1972, he did invalidate that plan.

    An appeal was promptly taken and also a submission of a motion for a stay order, the motion was referred by Mr. Justice Marshall to the Court and on June 12, 1972, this Court stayed the adverse District Court judgment.

    That had two effects. One was to bar implementation of a Master’s plan which had been under preparation, under now stayed judgment and more importantly perhaps and also on the way for conducting the 1972 election of state legislators in Connecticut under the official state plan, that was done.

    Along the way, there was a supplementary mandamus action in state court to clarify one or two clerical errors.

    The claim was reviewed by the state court and the State Supreme Court found that no constitutional impediment under the state constitution to use of the state plan.

    The large question therefore here is whether Connecticut can continue the use of this official state reinforcement plan, gerrymander of the 1970 decade.

    In a more precise fashion, the federal issues are three, population equality, justification of deviations as such be needed and the plaintiff-appellees’ allegations of gerrymandering.

    Turning first to the population issue, we assert, and most strongly assert that trivial deviations of the sort in Connecticut and I’ll mention those in a second, trivial deviations present no prima facie case.

    Therefore, the plan would remain presumptively constitutional unless plaintiff-appellees could introduce proof of a non-population nature of discrimination.

    We think that a concept of a presumptive constitutionality concept keyed to a lack of a prima facie case, if there are trivial deviations is implicit in Swann against Adams, and was not settled in Swann against Adams and is implicit in the dissented opinions in Kirkpatrick against Preisler.

    What are these Connecticut deviations which we find so trivial?

    In percentage terms, the most deviant district from ideal in the Connecticut Senate deviates by only 0.9%.

    For the Lower House, most deviant district from ideal deviates by only 3.9%.

    Over these percentage terms, must be put in a context of the size of the district being created to become meaningful and likewise the available census data we look to the census data.

    We find that the most deviant district in either house in Connecticut deviates from ideal by fewer than 800 census-persons, for the Lower House, 789, for the Senate 787.

    You might well ask if the population deviation in real census terms is identical virtually why a percentage — or the percentages is so different, the answer is in the size of the districts.

    Connecticut has 36 State Senate seats, average population, 84,000 roughly, 800 population deviants, census body is deviant on that base because there’s only 0.9% deviation.

    Connecticut has 151 assembly districts, average population about 20,000 on that base the same trivial 800 person deviation, gives you only a — since they raise the percent deviation to a 3.9%.

    In terms of the census data which was the building unit for the apportionment board of the 2,700 census units available to the board, 88% in exceed 400 in census population and yet the average deviation in Connecticut in either House is under 400.

    By contrast, Howell against Mahan — Mahan against Howell last week, excuse me, the most deviant district from ideal was I believe was about 4,400.

    The figures of the Connecticut type, we suggest are really on target in the Reynolds against Sims sense of substantial equality or of not — or in showing that the vote of each voter in a state, be approximate or equal to that of any other on target and therefore constitute no dilution of voting power.

    Hence we suggest that this case could fit the general principle of treating trivial deviations of this sort as constituting no prima facie case of dilution of voting power as being presumptively constitutional.

    More precisely and we’ve suggest this for Connecticut and not for the nation necessarily about a rule if we divide as follows for Connecticut making more precise this presumptive constitutionality concept.

    Robert G. Dixon, Jr.:

    It could run this way that where the maximum deviation in census population terms, actual census bodies in a district, in most deviant district, in a state plan is less than the average population of the census units being constructed, that plan should be presumptively constitutional.

    And the average population of the census units in Connecticut is 1,100, the median unit, census unit is 1,000.

    And as I said the most deviant actual district from ideal is only 800.

    Without such a presumptive constitutionality rule keyed to lack of showing a prima facie case with trivial deviations of this sort, we feel the Courts remain not only in the business of stirring up the political thicket, that plaintiff instigation that may take up residence in a permanent basis in the political thicket.

    Why do we say this? Because plaintiffs can always show a slightly less deviant plan then laboriously constructed official state plan at any given level of the percentage deviation.

    With a hand calculator, you could device hundreds of plans, each one though having a quite different political effect because no district line is neutral.

    Lower courts have tended to follow this approach in the past.

    In fact we’re having a past and we’re still today to an extent a one body better rule.

    That is plaintiff’s comment with the plan often to the last minute at trial perhaps showing their own self serving plan, just no statewide input in it, showing it is one body better so to speak than the state plan and if it — that the election is immanent and the state cannot justify or hasn’t built under the — its process, the state plan goes down the drain.

    Now, Mahan against Howell last week, that does not solve this problem of avoiding casual invalidations on a one body better rule because the state which has tried to be directly on target and does not crank into its process any corollary rational state purposes than an attempt to be directly on target in population could not qualify to utilize the Mahan against Howell justifications process.

    However with — what we do — I refer to here is a presumptive constitutionality rule, keyed a very low deviations of this sort, we will achieve some legislative certainty, the bodies can apportion without the sense of futility as often been the case in the last several years, litigation have been massive as we all know even after 1970 census.

    As far as courts are concerned of this approach a finding that there is no prima facie case with low deviations of this sort, trivial deviations would flush out many petty cases.

    Reserved court time for those cases for justification does seem needed because the size of the deviation or for the more complex issue of gerrymandering if the court proceeds in that direction.

    Potter Stewart:

    Mr. Dixon, do I understand your argument to be just — you told us a couple of minutes ago that you cannot rely on the Virginia cases decided last week —

    Robert G. Dixon, Jr.:


    Excuse me, I said —

    Potter Stewart:

    For the very reason —

    Robert G. Dixon, Jr.:

    (Voice Overlap) —

    Potter Stewart:

    For just a moment, in fact, so I can ask my question, so I’ll be sure understand it.

    Robert G. Dixon, Jr.:

    I apologize.

    Potter Stewart:

    You cannot rely on it for the very reason that you have in order to approach to a maximum numerical proportionality, you have disregarded state boundaries and town boundaries and so on, is that it?

    And therefore you can’t rationalize the tiny deviations that remain in terms of any legitimate state interest as Virginia could in respecting local political boundaries, is that your argument or just I misunderstand it?

    Robert G. Dixon, Jr.:

    Mr. Justice Stewart, my point was for — was made in general terms that a state might have only a straight population purpose, that is not Connecticut.

    Connecticut has three purposes to be served in its reapportionment process.

    And we’re suggesting that we would not reach the need to justify, shouldn’t have reached the need to justify until it is first determined that the deviations of a given plan and I’m not talking about Connecticut with its trivial deviations, are high enough to warrant constitutional concern.

    Potter Stewart:


    Robert G. Dixon, Jr.:

    We fell that if the deviations are smaller, not when deviations are smaller than the census average population — the average population census unit is being used that it is really on target and should be declared constitutional.

    Byron R. White:

    Without getting to any figures, without getting any reasons?

    Robert G. Dixon, Jr.:

    Without getting to it, the process justification is correct Mr. Justice White.

    Potter Stewart:

    Well, the point is though, the more a state in order to approach numerical perfection disregards its other state interest unless it can justify any tiny deviations in terms of any state interest, isn’t that correct, because it has disregarded them?

    Robert G. Dixon, Jr.:

    That would depend upon Mr. Justice Stewart, depend upon the rational state policy at issue and I think that is an next logical consideration for the Connecticut case.

    We have sought not merely to achieve population equality directly on target as a pre-eminent goal in order to serve the overall purpose of fair and defect in representation.

    I mentioned several times Reynolds against Sims.

    We have also added two corollary policies which we think would support this plan under the decision of Mahan against Howell last week.

    In short, if we do not prevail on a theory of presumptive constitutionality because deviations are very trivial, then there are the following justifications to be made.

    The Apportionment Board seeking close population equality also had two corollary policies, a Connecticut town policy and a policy of political balance or fairness which I’ll define in just a moment.

    Now turning first to the town policy which is the analogue of the no-county line policy in Mahan against Howell last week, we find that where Virginia could honor that fully and still stay inside a top to bottom variation of 16%.

    The Connecticut sought to honor its town — its no-town line cut fully for the Lower House, it does not apply to the Senate but for the Lower House.

    The population deviations would reach a 111% if we bring forth the 1965 plan, and cut no-town lines and test it out under the 1970 census.

    110%, 11% is way beyond Howell against Mahan even if we think a no-cut policy is good.

    We do think that — an important to note, the purposes lying behind is no local division line cuts in order to see if they can be served by some other premises.

    Mahan against Howell last week, speaking of the Virginia county policy, analogous to the no-cut policy in Connecticut, said that it served the purpose of furthering a political voice for the counties.

    We see in this a judicial recognition of community of interest factor in representation that also spoke of the no-cut policy asserting an important anti-gerrymandering function.

    We see in this a judicial recognition of the danger of blind or invidious line drawing.

    The apportionment board in Connecticut honored the town — the no town line cut policy substantially.

    Approximate 3/4 for the line of the towns are not cut.

    The board cut 47 towns out of 169.

    It then — seeking to serve — preserve the substance if not the latter on all cut policy and the purposes of political voice and of anti-gerrymandering used a complementary policy of political fairness or political balance which requires some definition.

    This political balance or political fairness policy but that is found.

    The board with knowledge of public voting patterns in past elections tried to divide equal population districts which would and knowing out of hope of avoiding the minority election, it actually occurred in 1970 in Connecticut but would actually — of our hope of providing each party with a seat gain, the legislature approximately proportionate to its percent or total of — percent total of popular vote.

    Now a functional trend — you’ll see this is being a vital arrogant, anti-gerrymandering principle and an attempt to safeguard a political opportunity for effective political action by all of the voters.

    The essence of this political fairness or political balance principle uses a corollary of principle where they attempt to cut no town lines and they only cut 3/4, I know it would become — what they felt with the on target in equality terms.

    The assessment says, “A direct turning away from a gerrymandering purpose, one other result from the plan that was achieved.”

    We think that in regarding 1966 to 1970 election data, the plan is conceitedly fair, demonstrably so.

    Regarding 1972, we would suggest that it is also demonstrably fair, all class before the Court.District Court, would’ve achieved somewhat the same outcome in 1972 under all plans, it would’ve been an excess seat gained for Republican Party.


    Because that was a landslide election here in Connecticut and past election data in Connecticut and indeed a civil law of political science indicates that when you have a landslide election, the seat gained runs substantially ahead of the percent of the — of vote cast.

    And also, the charts on this point in our reply brief indicate that the factor of (Inaudible) seats is endemic and all the plans before the District Court and the — is no different under the official state plan than other plans.

    Byron R. White:

    Do you find any difference here if they had followed a plan of political unfairness?

    Assume all the districts were right on target as you put it number wise, population wise, the only thing is that they did follow a policy of maximizing either Republican or Democratic representation but I — drawing it, put in the districts in places where one party or the other would benefit the most, would you reach the same result or?

    Robert G. Dixon, Jr.:

    Mr. Justice White, we find the — we would find in that instance that we have — presented to us that issue of gerrymandering whether or not with the equality satisfied, there were nevertheless because of district pattern as mentioned in Fortson against Dorsey, a substantial submergence of a political arrangement within the population which had bring — which would then —

    Byron R. White:

    Do — you would say it turns from, into a fair to unfair when each — either party’s representation is not maximized if you could draw equally numbered districts in a way that one — that one party would have more or less or what?

    Both parties together have to have the maximum number?

    Robert G. Dixon, Jr.:


    Mr. Justice White, our proposition would be this that what is — if a state does not violate any constitutional requirement of the federal constitution by seeking a policy of fairness along with a policy of population equality.

    Byron R. White:

    Now but fairness means —

    Robert G. Dixon, Jr.:

    And by political fairness in this sense we do mean something called political balance does require consulting past election data and testing out various plans possible to see if they would become — would appear very unfair under no bordering pattern (Voice Overlap) —

    Potter Stewart:

    Oh, you’re talking about benevolent gerrymandering, aren’t you?

    Robert G. Dixon, Jr.:

    We could use that term not required by this Court —

    Potter Stewart:

    And the courts never held that the ordinary kind of gerrymandering as a justiciable question or any business of ours in any of these reapportionment cases so far but to whatever the ultimate answer to that question might be, your point is that benevolent gerrymandering at least is neutral that it doesn’t make it otherwise a valid plan invalid, is that it?

    Robert G. Dixon, Jr.:

    Yes Mr. Justice Stewart, that is precisely our position here (Voice Overlap) —

    Warren E. Burger:

    I should think you’d take some — I should think you’d take some comfort from the arguments in the preceding cases where at least I understood them to say that if it isn’t politically fair, it’s unconstitutional, that would — that argument is the one — you would think supports your position?

    Robert G. Dixon, Jr.:

    Our position on that is that that is a mater for plaintiff proof, for plaintiff attack on a state plan and that because gerrymandering of an invidious sort is not clearly subject to close policing by this Court yet an attempt to be fair should not be questionable either.

    My last point as to mention the plaintiff-appellee’s gerrymandering argument, they do allege that the plan is a gerrymander of an invidious sense that we see in this record not one iota of proof regarding the standards suggested in Whitcomb against Chavis following on that from Fortson against Dorsey for attacks on plans on an impact basis rather than a population basis.

    This Court said in Whitcomb that when that kind of claim is made there must a showing that the districts were conceded or operated as purposeful devices to further discrimination, racial economic, you can say a fortiori political fits that that principle and the proof on Whitcomb must be real life proof of impact on — in view of voting power.

    So in short we suggest that at one level we have a plan with deviation, somewhat trivial that should be deemed to be presumptively constitutional, absent a gerrymandering type proof.

    That is not thought to be the case, not figured enough and we’ll suggest too rational state policy to justify the deviations preserving as many town lines as possible or about 3/4 and also making a good faith attempt to avoid unfairness.

    And last we find no evidence at all here of any overt invidious gerrymandering of the sort which this Court has spoken of in Whitcomb against Chavis in similar cases.

    May it please the Court, I like to reserve the remainder of my time for rebuttal.

    Warren E. Burger:

    Very well.

    Mr. Satter.

    Robert A. Satter:

    Mr. Chief Justice, may it please the Court.

    This Court scored a stunning six sets in the one man one vote doctrine in the cases which enunciated that doctrine.

    In the space of a few years, this Court has come close to achieving or leading this country to achieve the fundamental purpose of that doctrine and has thereby enhanced a fundamental civil right, namely the right of each individual citizen to an equally weighted vote.

    And as this Court said in Reynolds to the extent that a citizen — the citizens’ right to vote is debased, he is that much less a citizen.

    And so these one man one vote cases are important and they put us on a high road toward improving the quality of our democracy and the passion and the high hopes with which this Court started that journey with Baker against Karr should not play and particularly it should not play in light of how close we are to an ultimate success.

    Warren E. Burger:

    Do you think these principles have the same validity whether people do or do not vote, whether 30% of the people vote, 50% or 75%?

    Robert A. Satter:

    I think they have that same principle, yes.

    Whether people do or do not vote doesn’t change the quality of the — change the necessity of having equally weighted votes or giving people the opportunity to be in districts where they have an equal voice with every other person.

    Yes sir.

    Warren E. Burger:

    If only 30% of the people in a whole in an entire district actually vote, then of course you have no assurance that to that result represents the majority view, do you?

    Robert A. Satter:

    Of course you don’t. I might say in Connecticut that the history is that we vote close to 80% in many of our elections and in fact some of our presidential elections, we have voted close to 90% and that’s a remarkable fact but it’s true.

    Now why do I say that we have come so close to success in these one man one vote doc — in these one man one vote cases.

    Because the facts are, that in 53% of the states, the extent of a deviation is 5% or less.

    The range of deviation is 5% or less in their — in the reapportionment of their Senate and 47 — and in 47% of the states, the range of deviation is 5% or less in the House, in their Lower House and by range of deviation I mean plus or minus to and a half percent.

    Now the states did not come to that voluntarily.

    The states were not anxious to accord this extent of equality to their citizens’ tree on their own free will.

    Reapportionment is a very painful process.

    I know it because I have been in the state legislature and I have been in the state legislature in which I was required to reapportion myself and it is a very painful process.

    It is a painful process whether the legislators are doing it or whether others are doing it.

    And the concern for the equality of the vote, of the individual citizen on the part of the reapportioners if left to their own devices is very slight, except when this Court has required them to do so.

    Potter Stewart:

    Mr. Satter, what is your — what are your figures as to the deviation here, in this case?

    Robert A. Satter:

    We submitted a plan of deviation of 2.16%.

    Potter Stewart:

    No, as to the deviation that was struck down by the District Court?

    Robert A. Satter:

    7.8%, the range of deviation is 7.8% Your Honor.

    Potter Stewart:

    The range, meaning what, the maximum?

    Robert A. Satter:

    No, the range from the highest to the lowest.

    Potter Stewart:

    Now what does that mean?

    Robert A. Satter:

    That means that it was plus 3.9 and approximately minus 3.9 Your Honor.

    Potter Stewart:

    So you don’t disagree with the figures?

    Robert A. Satter:

    Oh, no.

    We don’t disagree with the figures at all.

    Potter Stewart:

    Well, many in the cases, the parties can’t seem to get together on the figures.

    Robert A. Satter:

    No Your Honor.

    We recognize that the range of deviation is 7.8%.

    Potter Stewart:

    From a maximum of 3.93 to a minimum of 3.9 in percentage points?

    Robert A. Satter:


    Warren E. Burger:

    That was for the House?

    Robert A. Satter:

    That was to the House, yes.

    We make no issue whatever in this respect.

    Warren E. Burger:

    The issue is not here?

    Robert A. Satter:

    It’s not here.

    We obviously recognize —

    Potter Stewart:

    And the average, do you agree that for all the districts and the House, that’s 1.9 and the median 1.8?

    Robert A. Satter:

    I think the figures that are in the — in a stipulation are correct, yes and that’s that I think the Court found.

    Potter Stewart:

    At least we don’t have that kind of an argument in this case?

    Robert A. Satter:

    Not have that argument.

    And so, may it please the Court, the point I am making is that the process is a painful one that the rights of the individual citizen are only going to be protected by a Court having a very firm and severe stand on the requirement to achieve equality.

    And that thus the states have to be ordered first, to reapportion and Baker against Karr made that issue sufficiently litigable for them to realize they had to and secondly that they had to achieve equality.

    The Reynolds case established the standard of equality as equal, as practicable.

    There is no question that it had some caveats in it.

    It’s expressly stated that absolute maximum exactness or precision was not required.

    And taking the very tolerant approach that appeared to be implicit in the Reynolds decision, the Courts replied and in the period after 1964, the pattern of approved deviation in state legislative cases ranged substantially above 15%.

    And in fact, our record shows that prior to 1969, 88% of our states had deviations of over 15% with respect to their Senate and 96% had deviations with respect to their House of over 15%.

    And then came Kirkpatrick, fortuitously in 1969, prior to the 1970 census and prior to the time when the vast majority of the states had to reapportion and the Kirkpatrick decision had two parts to it.

    The first part was that the standard of equality was going to be strict as equal as practicable manner that the states had to seek to achieve precise mathematical equality and they were to be permitted only those deviations in which the variations were unavoidable.

    And the second aspect of it was that virtually all excuses or justifications at least as applied to congressional reapportionment, were going to be ruled out.

    And the states listened to that and they knew that this Court meant what it was saying and they felt that they had to apply those rules in the Kirkpatrick case to their own state legislative reapportionment.

    They were very attuned as all reapportioning authorities are to the nuisances of the language of this Court’s opinion where they can find a escape hole, where they can interpret an out, they will do so because achieving equality is not a high priority on the part of reapportioners.

    They have other purposes.

    They do not have purpose unless a Court tells them that they have to have it, but they read Kirkpatrick to say it applies to us and we’re going to listen to it and we’re going to apply it and that’s why we have 53% of the states keeping within a deviation of 5% or less.

    Thurgood Marshall:

    How many states is 53%?

    Potter Stewart:


    Robert A. Satter:

    How many states?

    Your Honor its 53% is all but it — what?

    Potter Stewart:

    Some are between 26 and 27 states.

    Robert A. Satter:


    Thank you Your Honor.

    And what this has demonstrated is and I know this line has been said to you so many times it’s a frightful cliché that the life of the law is not logic but experience and experience has proven the wisdom of Justice Brennan’s statement in Kirkpatrick that a state legislature which tries can achieve complete numerical equality among all of the states districts.

    And this is the answer to the — this Court’s carping critics who have run their hands over an over insistence upon equality among the districts.

    Robert A. Satter:

    That is an over insistence upon achieving an equally weighted vote for our citizens as if we can ever have an over insistence upon perpetuating valid and important constitutional rights and that was the purpose —

    Warren E. Burger:

    We’ll resume at this point in the morning counsel.

    Robert A. Satter:

    Yes sir.