WISCONSIN v. NEW YORK
Under the Constitution's Census Clause, Congress is vested with the responsibility of conducting an "actual enumeration" of the American public every ten years, primarily for the purpose of aportioning congressional representation among the states. Congress delegated this responsibility to the Secretary of Commerce who, in the 1990 census, decided not to use a statistical correction, known as the post-enumeration survey (PES), to adjust an undercount in the initial population count. Acting on behalf of several citizens' groups, states, and cities, Wisconsin challenged the Secretary's decision not to use the PES; claiming that it resulted in an undercounting of certain identifiable minority groups.
Did the decision of the Secretary of Commerce not to use the PES in the 1990 census violate the constitutional right of certain minorities to be counted?
Legal provision: Article 1, Section 2, Paragraph 3: Apportionment of Representatives
No. The Court held that the decision not to use the PES-based statistical correction would not be reviewed under a strict scrutiny standard of "one person - one vote" because there was no showing of intentional discrimination by the Secretary of Commerce. The Secretary's decision was consistent with past census practices in which the promotion of distributive accuracy, rather than numerical accuracy, was the primary objective. This is because errors in numerical accuracy are less likely than errors in distributive accuracy to impact on the apportionment of representatives among the states. Moreover, the Court added that significant deference - regarding how best to conduct the census - must be accorded to the Secretary since he directly received such latitude from Congress which, in turn, received such latitude from the Constitution.
Argument of Drew S. Days, III
Chief Justice Rehnquist: We'll hear argument first in No. 94-1614, Wisconsin v. the City of New York, Oklahoma v. the City of New York, Department of Commerce v. the City of New York.
Mr. Days, III: Thank you, Mr. Chief Justice, and may it please the Court--
The true total population of the United States is unknown and perhaps unknowable.
At every 10 years for the last 200, the United States Government has been making the best effort possible to count every person it could in compliance with Article I, section 2, clause 3 of the Constitution.
The 1990 census was no exception.
By all accounts, it was the most well organized and most thorough census in history, accounting for 98.4 percent of the population.
Since at least 1790, people informed about the census taking process--
Unknown Speaker: If the true population is unknown, how can one be sure that the most recent census accounted for 98.4 percent?
Mr. Days, III: --Well, it's an estimation, Mr. Chief Justice.
It's the best we can.
I think everybody has recognized that there's no such thing as a 100 percent accurate census, and that's very much bound up in this case.
Since at least 1790, people familiar with the census taking process have realized that the process does not locate everyone in the country.
In fact, in 1790 George Washington and Thomas Jefferson were convinced that the population of the new nation would be 4 million, and they were not persuaded otherwise when the census came in at 3,890,000.
But since at least 1940, the Census Bureau has recognized that there's something called a differential undercount, that is, that certain minority groups, certain racial groups are undercounted, not captured by the census to the same degree as non minority groups.
In this case the responses... the respondents contend that a statistically adjusted alternative to the actual count should have been adopted by the Secretary of the Commerce because of its alleged superior ability to correct for this differential undercount.
The Secretary, however, declined to do that, relying upon the actual count because he believed that it provided a more reliable basis for apportioning Representatives among the States, which after all is the primary purpose of the census, and he did not believe it promised to cure the differential undercount in any event.
The district court upheld the Secretary's decision.
On appeal, however, the court of appeals determined that the Secretary, by failing to adopt the statistically adjusted alternative, did not make... and I quote... a good faith effort to achieve the objective of equal representation for equal population and remanded the case to the district court where the Secretary was going to be required to show that his decision was essential to the achievement of a legitimate governmental objective in order to avoid a declaration of unconstitutionality.
We're here seeking reversal of the judgment of the court of appeals on remand with directions of entry of judgment for the petitioners because we submit that the Secretary's decision, following a 200-year tradition, was within the range of options afforded by the Constitution to the Secretary and that what the court of appeals did was failed to show the proper level of deference to the Secretary's decision that he is entitled to in carrying out his responsibilities with respect to the census.
One of the first things that the court of appeals failed to do was recognize the broad delegation that the Constitution gives to Congress and Congress in turn gives to the Secretary of Commerce in carrying out the census process.
Article I, section 2, clause 3 says that the Congress shall conduct the census in such a manner as they shall, by law, direct.
And Congress by statute says that the Secretary may conduct the census in such form and content as he may determine.
These are very broad delegations to both the Congress and to the Secretary.
In conformity with his powers under statute and under the Constitution, the Secretary of Commerce made three decisions.
The first was to put distributive accuracy, that is, the ability to allocate population among the States so that apportionment of Representatives among the States can be carried out in a responsible fashion.
He was going to use that as the principal focus of his analysis in comparing the actual count and the statistically adjusted count.
His second determination was that the statistical adjustment had to be shown to be superior with respect to distributive accuracy if he were going to use that instead of the actual count.
And the third was, after great deliberation and analysis, the Secretary decided that the adjusted count was not superior to the actual count and, therefore, he went with the tradition of 200 years and made the actual count the official census of the United States.
Unknown Speaker: General Days, is it clear or does the Government concede that the use of statistical estimates is permissible, let alone that it's mandatory?
I mean, the--
Mr. Days, III: Yes.
Unknown Speaker: --text of the Constitution, as I read it, doesn't... does not say, you know, that there will be an estimate of the number of citizens.
It talks about actual enumeration.
It doesn't even use the word "census".
It says actual enumeration.
Do you think that that could be complied with by Congress saying, well, we think the population of the country overall has increased 5 percent and, therefore, we think every State... chances are every State has increased more or less 5 percent?
Close enough for Government work.
Could they do that?
Mr. Days, III: Well, Justice Scalia, I think the approach that you were just describing was in fact what the Framers engaged in when they came up with Article I, section 2, clause 3.
It was an estimate.
And the interpretation of actual enumeration--
Unknown Speaker: Because they didn't have an actual enumeration, but what they called for was an actual enumeration.
Mr. Days, III: --That's true, but I think that what the term actual enumeration really suggests is that the Framers wanted to leave open to the Congress and later to the Secretary of Commerce the ability to rely upon developments with respect to nose counting or head counting.
But I think one can say about the term "actual enumeration" that there has to be a good faith effort by the Congress and by now the Secretary of Commerce to count bodies, to find people.
And, therefore, a totally synthetic census I think would be contrary to the constitutional objective.
Unknown Speaker: But going further, General Days, and using this PES adjustment, at least the district judge, McLaughlin, was confident that that was an appropriate way.
He said, if I were free to decide on my own, that's what I would use.
And he had a footnote, and I wanted to ask you particularly about that.
He said, in light of recent improvements in statistical tools, that the use of the adjustment in the next census is probably inevitable.
So, my question is, has the Government... it's now 1996.
There must be some planning for the next census.
Mr. Days, III: Yes, there is, Justice Ginsburg.
Unknown Speaker: So, what is the Government's current view on that, the propriety of such an adjustment?
Mr. Days, III: Justice Ginsburg, I don't think any definitive decision has been made, but I don't think that there's any plan to conduct a synthetic census, that is, not going out, as has been the case for 200 years.
Unknown Speaker: That wasn't my question.
My question was after you do that.
Mr. Days, III: After you do that.
Well, I think that the plan is to use some sampling, but I don't think that there is any plan to use a PES approach at this point.
I don't think there have been any final determinations.
Unknown Speaker: But there was no sampling in 1990.
There was no sampling that was used.
It was the enumeration that was used and then any adjustment was considered not sufficiently reliable.
Mr. Days, III: That's correct.
There is ongoing experimentation and study of various statistical forms of adjustment.
In fact, Secretary Mosbacher encouraged the Census Bureau, after he decided in July of 1991, to continue to experiment and do research on the whole idea of statistically adjusted censuses.
But as I was saying, I don't think there's any determination at this point to do a PES in the way that was suggested in this lawsuit and is the subject of the debate between the respondents and us.
Unknown Speaker: Do you know what was the source of the trial judge's confidence that the use of an adjustment after the enumeration would be inevitable?
Mr. Days, III: Well, I'm not certain, Justice Ginsburg, but I think that after he heard 13 days of testimony from experts of the highest caliber--
Unknown Speaker: He was punchy.
Mr. Days, III: --He may have been punchy, but I think he also got the sense that the statistical science was moving forward and there would probably be the ability of the Census Bureau to rely more heavily than in the past on this.
After all, we have seen a progression in 1980 that was an effort by the Census Bureau to look at what was called a post enumeration program, not a PES but a PEP, and it was decided that that was not sufficiently robust or reliable to use.
In 1990 we've moved another step, and I think he was simply suggesting that science and time move on without any definitive understanding of exactly what might happen.
Unknown Speaker: You know that's another lawsuit, though.
I mean, this one will decide whether you must use statistical estimates and the next one will decide whether you may use it.
I assume somebody will have an interest in saying that you can't do it next time around, if you do it.
Mr. Days, III: Well, that may well be but, Justice Scalia, if we are successful to the extent that we would like to be in this suit, we think there would be a cutting back on challenges to determinations made by the Secretary of Commerce in this regard.
If appropriate deference is shown to the Secretary in making these decisions, which are myriad, in coming to a conclusion as to what should be done with respect to the census, what types of adjustments should be made, then I don't think that we have to experience what was the case after the 1980 census in which there were 50 lawsuits that... over 50 lawsuits that weren't concluded until 1987.
Our principal point is that there are some types of challenges that deserve to be in court when there's an indication of intentional racial discrimination, when the Secretary of Commerce wants to hold a decennial census every 15 years rather than every 10 years, or where it's clear that the Secretary is not trying to count anybody.
Those may well be appropriate matters for courts to consider.
But what the respondents have really invited the courts to do in this case... and certainly the court of appeals seems to have accepted the invitation... was for the courts to take sides in a statistical dispute among statisticians and demographers about the propriety of a statistical adjustment.
We do not think that that's the type of dispute that properly belongs in Federal court.
Unknown Speaker: What--
--Would you go--
--What do you say is the applicable standard that we should employ in determining whether the Secretary's decision is consistent with the constitutional goal of equal representation?
Mr. Days, III: Justice O'Connor, certainly that's the way that we would describe the standard.
We think that applying it in this particular case, looking at the three decisions that I just mentioned, insofar as the Secretary is relying upon distributive accuracy as the principal objective, it seems to us that that's consistent with the constitutional language and the goal of equal representation, that that's what the Constitution has in mind how does the Congress and the Secretary of Commerce go about figuring out how Representatives should be allocated among the States, apportioned among the States.
That's perfectly consistent with the constitutional language.
And when we talk about the need for the statistical adjustment to be superior to the actual count, we think that that is consonant with but not dictated by the text and history of the Constitution.
After all, what the Secretary of Commerce is doing or was doing in 1990 was relying upon experience that the Census Bureau had gained ever since marshals went around in 1790 from door to door asking people who lived in their households.
That's what the Secretary was doing.
It may not be that the Secretary has to do precisely what was done in 1790 or 1980, but there has to be some approximation and that's exactly what was going on here.
Unknown Speaker: You're saying in response to Justice O'Connor's question that the Secretary gave an accurate interpretation of the Constitution.
I took it the thrust of her question was what standard should the Court use in assessing the adequacy of his determination.
Mr. Days, III: Well, I--
Unknown Speaker: I assume your answer would be that there would be rational basis--
Mr. Days, III: --That's right.
Reasonableness, and although the district court here used arbitrary and capricious, we think that that's about as far as the Court should go, but we would argue for a more deferential standard, one that apparently eluded the court of appeals in this case.
Unknown Speaker: --Would you take the position that it would satisfy the standard, at least in the absence of the kind of extraordinary circumstances like evidence of intentional undercounting, discrimination, and so on... would you take the position that it would satisfy the standard for the Secretary to take the position as a matter of principle that because all of these adjustment techniques ultimately involve kinds of value choices and are therefore politically manipulable, that it would simply be safer to go with the kind of garden variety actual enumeration, to the exclusion of adjustments as a matter of principle, to avoid the risk of political manipulation?
Would you go so far as to say--
Mr. Days, III: Yes, yes.
I think the--
Unknown Speaker: --if he bases his decision on that principled reasoning, that satisfy rational basis regardless of what arguments might be made that this technique would be helpful and that technique might not be?
Mr. Days, III: --Yes, I think the Secretary could do that.
That's consistent with the constitutional tradition, the history.
Of course, the Secretary could balance off those considerations, as he did here and went through a list of eight considerations, and decided, after looking at all of those, that adjusting was not the right thing to do.
Unknown Speaker: I think he may come regret being so open to the various options for estimation--
Mr. Days, III: Well, Justice--
Unknown Speaker: --which he hasn't been in the past.
Mr. Days, III: --Well, Justice Scalia, I think that a new Secretary of Commerce came in and found himself faced with a lawsuit and felt that he would make a good faith effort to try to sort through this and hear what the experts had to say and evaluate it.
And after going through an administrative process that produced 18,000 pages of administrative record and had 650 comments from outsiders and spent a great deal of time looking at this, he came to the conclusion that he did.
One of the most troubling things for us about the court of appeals decision is that it seems to want to assimilate wholesale the rigid standards of the Karcher decision into an evaluation of disputes over the census.
Unknown Speaker: Of the what?
Mr. Days, III: The Karcher.
Unknown Speaker: Karcher?
Mr. Days, III: Karcher v. Daggett where, in effect, there has to be mathematical equality.
And as this Court mentioned in Montana, although it may be common sensical in some ways to seek a precise mathematical test within each State, when you try to apply that same standard to the Nation as a whole, the common sense force of that becomes quite illusory.
The court of appeals certainly recognized that there were differences between the problems that the Federal Government faces with respect to the census and those that States confront, but did not give, in our estimation, a proper weight to those considerations.
After all, unlike a State, when we're talking about the reapportionment of Congress, there are certain constraints, three major ones.
There has to be at least one Representative per State.
There has to be not exceeding one Representative for every 30,000 people, and the district boundaries may not cross State lines.
These in and of themselves make it very hard to imagine that there is any ability to achieve mathematical equality.
But the court of appeals, nevertheless, said that the Secretary had to make a good faith effort and found that the Secretary had not made a good faith effort under these circumstances.
Unknown Speaker: Is there a way if in fact you did an actual enumeration, i.e., you went out and counted people, and you also thought that the inner cities were being undercounted because people couldn't find everyone... it's hard for the--
Mr. Days, III: Yes.
Unknown Speaker: --Is there a way in a system of actual enumeration to do better at counting the people in the inner cities that's a practical way?
Mr. Days, III: Yes.
Well, I think that this 1990 census reflects probably the most forceful and dedicated effort to do that in history.
There was an effort before the census was started to identify households.
Municipalities and States were given an opportunity to challenge the list.
They were able to do that afterwards.
There were special get out the census count campaigns specifically targeted to hard to count populations around the country.
That's why I think the ultimate census, although, Mr. Chief Justice, we don't know whether it's exactly 98.4, by all estimates it's 98.4.
We're talking about missing 4 million people approximately in a population of 248 to 253 million, which is a pretty good record.
But let me say in closing something about this differential undercount because it has really driven this lawsuit, and I think the Secretary reflected his concern about this, as well he should have been concerned.
And I think all of us are concerned.
But what he established was that the statistical adjustment could not show that there would be reliable distributive accuracy.
What that means is, if there's not distributive accuracy at the State level, one doesn't know where to place the minorities who are presumably undercounted.
It's not clear whether one puts the so called undercount groups in New York or South L.A. or other parts of the country, and that can produce great distortions.
If one looks at the adjusted count's impact on New York City, for example, New York State gained population under the adjustment.
May I finish this comment, Mr. Chief Justice?
Unknown Speaker: Yes.
Mr. Days, III: Gained population but it lost share.
That is, it lost its ranking relative to the other States in the Union.
We think for these reasons the court of appeals judgment should be reversed and the case remanded for the entry of judgment in favor of petitioners.
Thank you very much.
Unknown Speaker: Thank you, General Days.
General Doyle, we'll hear from you.
Argument of James E. Doyle
Mr. Doyle: Thank you, Mr. Chief Justice, and may it please the Court--
The people of the State of Wisconsin stand to lose a seat in Congress to the State of California, which is no longer even pursuing its claim because the City of New York and the other respondents seek to engraft on the Constitution certain statistical procedures about how to take the census over which even statisticians disagree.
Unknown Speaker: Do you agree that the individual voters who claim their districts have been undercounted have standing to bring this suit?
Mr. Doyle: We do not agree with standing in the... in apportionment.
We believe that that standing belongs to the State, not the individual voters, as it does in the Karcher case in which you can actually assess what your proportional share of the voting statistic is.
But in the case of apportionment to the State, we believe that belongs to the State.
As in Montana and in Franklin v. Massachusetts, both of those claims were pursued by the States that stood to lose a seat in Congress if it had been... if the procedure had been adopted.
In this case, California pursued its claim in the district court, was... did not appeal the district court's ruling against it, and we believe the State of California is bound by that judgment.
Unknown Speaker: Do individual voters then not have the standing because the States necessarily supersede their interests?
Mr. Doyle: In our view, it is in this case the States' interest and the individual voter does not have the standing.
But let me emphasize in this case California's case was brought State ex rel. for the people of the State of California by the Attorney General of the State.
That's how the case was brought.
So the Attorney General was bringing it not on behalf of the corporate entity of the State of California, but was bringing it on behalf of the people of the State of California.
And he... and California did not appeal the judgment of the district court.
Unknown Speaker: Why doesn't an individual voter have standing on the theory that if there are more Representatives for that voter's States, the ratio of Representative to voters will be more favorable and, therefore, the voter, as well as the State, has a stake?
Mr. Doyle: The apportion... in my view the apportionment right, the right to a certain number of seats in Congress does not convert into the same kind of statistical how many Representatives do I have, what share of a Representative do I have in Congress.
And there is always going to be wide--
Unknown Speaker: Well, they are conceptually different but they are related because the ratio will depend on the apportionment.
So, it may be that the voter is not bringing in a technical sense an apportionment claim.
The voter is bringing a variety of person to vote claim.
But why doesn't the voter have standing to do that?
Mr. Doyle: --Well, because as I said, I believe it's the State's right.
But let me emphasize this case.
The voters have... the Attorney General was representing the people of the State of California, and certainly the voters of New York do not have any standing to claim that Wisconsin should transfer a seat in Congress from Wisconsin to California.
The only practical result of this statistical adjustment, as it was later corrected... the only practical result of this statistical adjustment, as it was later corrected is that one seat in Congress will be transferred from Wisconsin to California.
Unknown Speaker: What about Federal funding that's based on population and so forth?
Isn't that an injury?
Mr. Doyle: --Federal funding is not a constitutional issue.
There is no... the question before this Court is whether there is a constitutional violation.
You have no... no individual has--
Unknown Speaker: Well, doesn't that show injury?
Mr. Doyle: --Not if there is not any showing of purposeful, intentional discrimination.
There is no constitutional right to per capita receipt of Government largesse.
There's no right that a citizen in Wisconsin has to get equal number of dollars from the Federal Government that a citizen in Illinois gets.
It is... it may be a statutory claim, perhaps reviewable under the APA, but it is not a constitutional claim.
The only constitutional--
Unknown Speaker: Well, it's not a claim, but does it suffice to have the nexus of injury for him to bring the claim?
Mr. Doyle: --It may have the nexus for injury to bring a claim, not the constitutional claim having to do with apportionment, which is the claim before this Court, but a statutory claim.
It may have the nexus of injury for standing.
I don't think it would be a successful lawsuit, but it may be the nexus for standing of a non constitutional claim.
Unknown Speaker: But would you have a different basis for challenging the census as it is used under some of these statutes than you would have for challenging it in this lawsuit which does not relate to a statute but to the constitutional prescription?
Mr. Doyle: I think you may well have a different basis.
You may well have... even under the Franklin case, you may have a Administrative Procedures Act lawsuit.
I don't know--
Unknown Speaker: So, you could come out with different results?
You'd have a census, and you'd have two different censuses, one of which would be valid for apportionment purposes and the other one of which would be valid for all of the statutory apportionments?
Mr. Doyle: --That is true currently, Your Honor.
There are various census numbers that are used for the distribution of Federal funds.
There's a mid decennial census that is taken that is not a constitutional census that is used for purposes of various Government programs.
Unknown Speaker: No, but assume the statute just says according to the census figures.
Is it possible that that would mean one thing for statutory distribution purposes and something else for purposes of apportionment?
What is the census?
Mr. Doyle: If the statute said the official census--
Unknown Speaker: That's what it says.
Mr. Doyle: --then we believe that you could conceivably have two different results, although I think it's unlikely.
But the issue before the Court is one of whether or not in this case there has been--
Unknown Speaker: Oh, okay.
So, we could have another lawsuit about it.
We can do all this over again when somebody comes up with a statutory apportionment that requires it to be apportioned according to the census.
And you say they can distinguish whatever we say in today's case because this only related to the official census.
Mr. Doyle: --My understanding of the statutes having to do with disbursal of Federal funds is you would not have that lawsuit, but perhaps under a hypothetical statute, you might well.
You will have many, many lawsuits coming before you constantly if... before the courts if the Second Circuit's decision is adopted.
If in the year 2000 the census is harmful to the State of... the statistical estimation is done and the statute is harmful... the census is harmful to the State of Wisconsin, we may well be before the Court if the Second Circuit decision is there, arguing that if you use different post strata, if you had grouped Wisconsin with different States, if you had had a different way of imputing the people who you can't match, that we would be... have a seat in Congress.
And under the Second Circuit's decision, not only would we have a claim, but by simply coming forward with a claim that we can do the census better than the Census Bureau did, count more people, we would have heightened scrutiny applied to that claim as well.
And if I might address the question about whether the adjustment of the census itself would be constitutional, the statistical estimation, whether it could constitutionally be done.
There is a significant constitutional problem with that.
The Fourteenth Amendment, section 2, requires that the census be done by a counting of the people in the States.
The way the Wisconsin number would be arrived at, if you accepted this statistical estimation, is that our count would not only be because of counts in our State, but through this post strata system by which people are counted in Ohio and Indiana and Illinois and Michigan, our numbers are adjusted basis... based on counts that are made in the post strata in other States.
In fact, in the post... in the PES only I believe 169 blocks of Wisconsin, about 5,000 people, were included in the post enumeration survey.
And our numbers were being estimated based on that.
Unknown Speaker: General Doyle, then are you disagreeing with General Days who said, in answer to the question might the sampling technique or some statistical adjustment be used, yes, it might?
That's the Secretary is considering.
And there was also the prediction, the forecast, that we wouldn't get a lot of lawsuits because if we gave proper deference to the Secretary, then that would be that, as long as it met a reasonableness test.
But I think you are now suggesting that the Secretary doesn't have discretion.
Mr. Doyle: I am suggesting that the Secretary does not have discretion to do the kind of statistical sampling that was done in this case.
Unknown Speaker: Any kind of statistical sampling?
Mr. Doyle: --No.
There may well be other kinds of statistical sampling that... I'm not a statistician... that may serve the purpose, but I do not believe that you can arrive at the census of Wisconsin based on counts of people that you have made in Illinois, Indiana, and Michigan, which is what--
Unknown Speaker: Because that is inconsistent with the term "actual enumeration".
Counting the whole number.
Mr. Doyle: --It's inconsistent with actual enumeration, but particularly it is inconsistent with the Fourteenth Amendment, section 2, which says counting the whole number of people--
Unknown Speaker: I see.
Mr. Doyle: --in the States.
You count people in Illinois.
You say, oh, there has been an undercount in Illinois.
You group Wisconsin with Illinois, and our undercount's percentage is based on counts of people that are made in other States.
Unknown Speaker: And I take it, you have a... sort of a footnote to that argument too because I gather Wisconsin claims that it made unusual efforts to get out the census so that the very fact that you tried to do a good job would be a further reason, in fact as well as in principle, to say you shouldn't compare us with another State that might not have made that effort.
Mr. Doyle: Well, that's correct, Your Honor.
Wisconsin worked very hard.
And let me say that there's a suggestion that the majority States... that there's a majoritarianism issue here.
Every State has an interest in seeing that everybody in that State is counted, minority and non minority citizens alike.
When a minority citizen in Wisconsin is not counted, my vote... my representation is as depreciated as that minority student... that minority citizen who is not counted.
So, every State has an interest in seeing a full count made.
Unknown Speaker: Thank you, General Doyle.
Mr. Doyle: Thank you, Your Honor.
Unknown Speaker: Mr. Rifkind, we'll hear from you.
Argument of Robert S. Rifkind
Mr. Rifkind: Mr. Chief Justice, and may it please the Court--
Let me say at the beginning that the court of appeals did not hold that Secretary Mosbacher's decision was unconstitutional.
It only held, as I understand it, that the district court had employed an inappropriate standard when it found that the Secretary's decision, though mistaken, must be sustained because not so far beyond the pale of reason as to be arbitrary and capricious.
There should be no question about it.
The district court found that the Secretary's decision was wrong, but not beyond the pale of reason.
Unknown Speaker: Well, on what basis did the district court go about finding that the Secretary's decision was wrong, as you put it?
Mr. Rifkind: The basis was, first of all, 13 days of trial and the consideration of the evidence submitted there, but I think very largely on the basis made of concessions made by the Secretary.
The Secretary never found... never found... that the adjustment was less accurate than the original count.
Unknown Speaker: Perhaps we should refine what you mean by saying that the district court found the Secretary's decision was, quote, wrong, close quote.
Mr. Rifkind: What the district court said... and this is in the Wisconsin appendix at page 77... the Secretary has conceded that the objective criteria used to measure the adjusted counts show a greater numeric accuracy at the national level and that the Census Bureau estimates of distributive accuracy marginally favored the adjusted counts.
Unknown Speaker: How does that add up to being wrong?
Mr. Rifkind: It found... it adds up to being wrong if the standard is, is the corrected count more accurate than the original count, and--
Unknown Speaker: So, you say it's simply a factual inquiry that's up to the district court to decide whether one count was more accurate than the other, regardless of what the Secretary chose to do.
Mr. Rifkind: --I believe, as I believe the Second Circuit did, that where fundamental constitutional rights are impacted, the court must conduct a more searching inquiry than merely beyond the pale of reason.
Unknown Speaker: How could that be the test?
I mean, if in fact the true cause is that people who live in inner cities and are hard to get to are undercounted, people who live in inner cities and are hard to get to also have more diseases and also eat less well.
So, in fact, a census system that adjusted upwards by 1 percent according to diet would be more accurate.
Yet, a system like that couldn't possibly be a more lawful system because you'd produce bizarre results in trying to apportion Representatives among the States.
So, I don't see how that could be the test as opposed to the test being are we going to do our apportionment business of Representatives in Congress better.
Mr. Rifkind: Absolutely.
Unknown Speaker: Well, then as soon as you say that, you run into their argument about, you know, who gets the extra Representative, and there are a thousand different ways to do it, and it's undercounted here and overcounted there, and we don't know the true causes of the undercount, et cetera.
Mr. Rifkind: Well, the... I think the key factors here are that the Census Bureau, the expert agency, to which normally I would suppose considerable deference was due, had spent a great deal of time trying to figure out the solution to the problem, after an enormous amount of effort and a very professional effort had decided that it had figured out the solution, or at least a significant amelioration of the problem, and I must say, Justice Breyer, repeatedly advised the Department of Commerce that there was no other way.
Unknown Speaker: You said to whom a great deference is due, and as soon as we get into a great deference is due, they win.
Mr. Rifkind: Well, what I'm... I guess what I'm saying is a decision that overthrows, as the Secretary did twice... overthrows the recommendation of the expert agency, when the Secretary himself obviously claims--
Unknown Speaker: Well--
Mr. Rifkind: --expressly claims no expertise, it raises a question about who--
Unknown Speaker: --Are you saying, Mr. Rifkind, that a political appointee Secretary, who is authorized by Congress to take a particular action, if he rejects the recommendations of his permanent staff, is therefore very likely of being held arbitrary and capricious?
Mr. Rifkind: --I do not believe that arbitrary and capricious is the appropriate standard.
Unknown Speaker: Well, what--
Mr. Rifkind: What I... I guess what I'm trying to say is where you have... where the Court has decided... and I thought this Court had decided... that it would look searchingly at Government action that undermines fundamental constitutional rights, that under those circumstances, when one considers the whole matrix of factors, the fact that an expert agency had made a recommendation and that the Secretary of Commerce had overturned it would lead a court to at least inquire did he have very good reasons to overturn it.
Unknown Speaker: --And what is the Government action here which you say undermines fundamental rights?
Mr. Rifkind: The promulgation by the Secretary of counts, in which persons are in very large numbers omitted not uniformly across the country but, as we have known since 1940, in those places in which minorities reside... those minorities and the people who live next door to them, as the General said, are undercounted when it comes to drawing districts.
That was established by Professor Kohl at Princeton in 1955 and is conceded.
Unknown Speaker: But that assumes that they're entitled to some different form of count than they got.
Mr. Rifkind: They are entitled to a different form of count if there is an available count that doesn't have that consequence.
Unknown Speaker: I don't think you're right on that.
I think the Secretaries and... the Congress is given great discretion by the Constitution.
Congress in turn confers great discretion on the Secretary, and the idea that you simply decide all over again in the district court whether the Secretary was right or wrong I think is quite unsupported by any case.
Mr. Rifkind: The... there is a tradition that goes back to the beginning of the century of Bureau autonomy.
The remarkable thing about the 1990 census is it is the first occasion on which the Secretary of Commerce has undelegated to the Director of the Bureau the authority that the Director of the Bureau has always had in this respect.
Unknown Speaker: Do you claim that the Secretary did not act in good faith in putting out the actual count census and relying on that?
Mr. Rifkind: I believe that the Secretary did not pursue in good faith the zealous pursuit of equality and accuracy which should be the overriding command.
Unknown Speaker: Is it clear that some kind of statistical substitute for an actual count is permissible under the Constitution?
Mr. Rifkind: I think... let me be clear.
No one has suggested, not the Census Bureau and not we... no one has suggested a substitute... that is, disregarding the enumeration entirely.
The only question is whether it can be improved by statistical means.
I don't hear the Government, the Solicitor General, or anyone else suggest that, except for Wisconsin perhaps, that one could never use a correction of statistical nature to improve the quality.
And, of course--
Unknown Speaker: Well, the Attorney General of Wisconsin is here saying the Constitution doesn't permit any statistical adjustment.
Why do we think it does?
Mr. Rifkind: --I think Wisconsin in that respect is wrong.
Unknown Speaker: Why?
Mr. Rifkind: --the Solicitor General agrees with me on that, and I will say why.
I think that when the Constitution uses the word "enumeration", it means to render in numbers.
Unknown Speaker: It doesn't just say enumerate.
It says actual enumeration.
Mr. Rifkind: I take the word--
Unknown Speaker: Actual enumeration.
And you want to say that since you cannot do a perfect actual enumeration, you will take the actual enumeration and adjust it upward or downward by statistical means.
I don't think that that... it isn't clear to me at least that that constitutes an actual enumeration.
Mr. Rifkind: --I take it the word "actual enumeration" in context there differentiates it from the attribution--
Unknown Speaker: The imaginary enumeration.
Mr. Rifkind: --contained in the... well, in the proceedings clause of the Constitution which says for the first 3 years we'll allocate Congress the following way.
But the fact of the matter is--
Unknown Speaker: That was not... Mr. Rifkind, that was not an issue that was aired below since it was accepted by the Government--
Mr. Rifkind: --It was.
Unknown Speaker: --that the Secretary spent all the time considering this.
So, that isn't an issue.
But I was interested in your answer about this fundamental right, and we know the result.
It was also known that an absolute veterans' preference in Massachusetts would mean that of... the upper echelon jobs, and the vast majority would be reserved to a group overwhelmingly male.
And we have Washington v. Davis.
How can you argue that even though you know that the undercount will disproportionately affect minority members, that for that reason you get this highest level of scrutiny when this Court has rejected that over and over again?
Mr. Rifkind: I'm not sure that one gets it because it disproportionately affects minorities.
I would have thought one got it under the line of cases following Baker v. Carr.
One gets it because when we require of States that they proportion quite precisely equal numbers for equal... equal representation for equal numbers, we don't inquire about their intent, and we don't listen to 200 years of history as an explanation for why they got there because--
Unknown Speaker: But then General Days has told us that you can't compare the national census with apportionment one person/one vote within a State, and General Doyle tells us that what we have to concentrate on, the only thing at the constitutional level before us, is taking away from Wisconsin one seat, adding to California one seat.
That's what he said is before us, nothing else.
Mr. Rifkind: --I don't think that's all that's before us, although that is before us.
Unknown Speaker: Why... explain then.
The Secretary said I'm concentration... concentrating on distributional accuracy within the United States.
The result of doing this PES will mean that California gets another seat, Wisconsin gets one seat less.
Now, why is anything more than that involved?
Mr. Rifkind: Well, we all know... and the Congress has commanded... that the census also be available for the States to use in drawing congressional districts.
Unknown Speaker: If they so will.
If they so will.
They're not required to.
Mr. Rifkind: I think it would be a temeritous State that declined to use the census counts to draw its congressional districts.
After Karcher at least, I think there's a grave doubt whether they're free to do that, but assuming they are, they will certainly be exposed to an array of litigation much larger than has been suggested.
And as a practical matter, this is fairly traceable to what the Secretary does.
As a practical matter, we know the Secretary is commanded by section 141 to give to each of the States the most accurate data they can so that they can draw their congressional districts, State districts, State legislative districts.
And since 1962, this Court has required the States to do just that, to draw equal populace districts.
Unknown Speaker: Mr. Rifkind, why couldn't the State say, we take the enumeration and then for our own purposes of apportioning, we will take this adjustment?
Mr. Rifkind: No State in fact is equipped to do that.
Unknown Speaker: But... because they don't have the numbers to do it?
Mr. Rifkind: Only... in this case we have the numbers because we forced the Government to give them the numbers.
Unknown Speaker: Yes.
Mr. Rifkind: But even so, most States have constitutional requirements of their own and statutes of their own that say you use what the Secretary of Commerce delegates... sends to you pursuant to section 141 of the code.
Unknown Speaker: Well, then maybe each State should redecide how that statute should be and say that we want to have the enumeration plus the statistical adjustment.
And then it's a question for the State's political process.
Mr. Rifkind: --Well, that may be.
As a... in this decade that doesn't solve the problem.
We have a set of data and we have an alternate set of data.
The Secretary chose the data that the experts in the agency, that the Bureau told them were the less reliable, that missed 4 million or 5 million people nationally, that undercounted blacks, Hispanics, Asians, and so on.
Unknown Speaker: And you told me the practical result on the State level, but I'm looking for the constitutional infirmity.
Mr. Rifkind: Well, I think there's at least a suggestion in Karcher that the States are required to use the census counts unless they can prove... prove... that they have something better, at the very least.
Unknown Speaker: Mr. Rifkind, your whole argument seems to rest on the assumption that we're dealing with two sets of figures which are alike in every respect except one is more accurate than the other, and that in fact is not so, is it?
We are, in fact, dealing with a set of figures which are derived, one of them by actual count, the other one by estimate, which depends upon a series of value judgments in order to determine how to make the estimate.
So, we are not comparing a big apple against a little apple.
We are comparing two different kinds of vegetables, aren't we?
Mr. Rifkind: Justice Souter--
Unknown Speaker: I mean, isn't that so?
Mr. Rifkind: --With respect, it is fundamentally wrong.
Unknown Speaker: All right.
Mr. Rifkind: And I'd like to address that.
The census is full of estimation.
The original enumeration, the enumeration, is full of estimation.
Every Director of the Bureau that I have known of... and we quote several of them at page 9 of our brief... has said the notion that there's no estimation in the enumeration doesn't understand what the actual way we count people is.
Unknown Speaker: Would it be fair then to say that the difference between these two figures is a difference on the... a difference depending on the amount of estimation versus actual enumeration which in fact is involved in reaching the number?
Mr. Rifkind: There is--
Unknown Speaker: By definition, there's... I presume there's more estimation in the method that you want used or you wouldn't want it used.
Mr. Rifkind: --There is more estimation involved, although in 1980--
Unknown Speaker: Okay.
Mr. Rifkind: --the explanation involved which involved imputing the existence of people and their numbers by a very crude device was challenged in Orr v. Baldrige in a dispute between Florida and Oklahoma because it shifted a Congressman, and the Census Bureau prevailed.
Unknown Speaker: Okay, but we have a different--
--What was the estimation?
I'm not sure.
What was that imputation?
Mr. Rifkind: My recollection is that in... that the process there involved is when a house was seen to be... you couldn't find the people in the house and the neighbors didn't know how many people were there and the postman didn't know how many people were there, they said, we know it's not vacant, so we'll attribute the numbers in this house to be equal to the numbers of people in the adjacent house.
And with that device, which I describe as crude, but better than nothing, they added, as I recall, 790,000 people to the census which shifted a Congressman from Indiana to Florida.
Unknown Speaker: Okay.
Mr. Rifkind: And in previous decades, they've added millions of people by similar estimating processes.
Unknown Speaker: Now, may I go back to the comparisons of these two estimations, one of which is more ambitious than the other?
If we assume that the two figures are properly comparable so that we can say one is better than the other, we have to assume something about the judgments that are made in deciding how to make the preferable estimate.
We have to say, yes, those value choices were good value choices.
Isn't it the case then that you have a problem in your argument, just taking it on its own terms, by virtue of the fact that, if I recall correctly, the head of the Census Bureau said, I wouldn't have done it this way, but the Secretary's choice is within the realm of reason?
In other words, it could... a reasonable person could make a different set of value choices from the ones which I personally prefer.
It would still be within the realm of reason.
Isn't that a difficulty for your case?
Mr. Rifkind: I can see it.
Dr. Bryant said, reasonable men can differ, reasonable persons can differ.
Unknown Speaker: And that's why you have to insist on the heightened scrutiny standard because otherwise you lose.
Mr. Rifkind: Well, I think there's a risk of that.
Unknown Speaker: And a high risk.
Maybe a high risk, yes.
Mr. Rifkind: --I must say I don't envy the position Dr. Bryant was in when she had to utter that statement.
She was standing next to her boss, and it's a common thing for non lawyers to say when they're standing next to their bosses.
Unknown Speaker: Okay, but we don't have any finding from the people... the judge who heard the testimony that in fact it's not to be believed.
Mr. Rifkind: I agree with that, but I think if you read the report... and it's in the joint appendix of Dr. Bryant... you read the report of the Census Committee, you read those sorts of materials, you come away with the feel that she felt very emphatically that they had solved the problem.
And with good reason because--
Unknown Speaker: Well, she felt very emphatically that they had come to one, and indeed in her view preferable obviously, solution to an assumed problem of undercounting.
But to say that they had solved the problem is truly to say they had come up with one solution to a problem and there were other solutions, and it was within the realm of reason to say that in fact the cure might be worse than the disease.
Mr. Rifkind: --No one at the trial and not the Secretary said that the enumeration was better, whatever that means, than the correction.
What the Secretary said in effect was because sampling is involved, because sampling always brings in an element of what the statisticians call sampling error, random error, there's an element I don't know quite as much about, and because I don't know, I'm going to prefer what is on its face the worst mechanism.
That's essentially the ground rules for decision that the Secretary put out.
And I think whether that's an appropriate ground rule for decision where constitutional rights are involved is an appropriate inquiry for a court.
Unknown Speaker: Mr. Rifkind, isn't that quite an exaggeration to say I'm going to stay with the worst method?
He said, I'm going to stay with the method we have unless I am convinced that something is better.
And even Judge McLaughlin didn't say the Secretary was wrong.
What he said was if I were the decision maker, I would probably... not certainly... I would probably have ordered the adjustment.
That's the furthest he goes.
Mr. Rifkind: No.
I think he goes further actually, Justice Ginsburg.
He says the court is satisfied that for most purposes... that is appendix 59... for most purposes the PES resulted in a more accurate count than the original census.
And he says, the Secretary has conceded that the objective criteria used to measure the adjusted counts show a greater numeric accuracy at the national level and that the Census Bureau's estimates of distributive accuracy marginally favor the adjusted--
Unknown Speaker: For most purposes, but that's the crucial phrase there.
What the Secretary did not concede was that this adjusted method was more accurate for the central purpose of distributing Representatives among the States more accurately.
Mr. Rifkind: --That's just--
Unknown Speaker: For that purpose, he thought that the actual enumeration was better without statistical adjustment.
Isn't that... wasn't that in essence what he said?
Mr. Rifkind: --My understanding of what he said is that he never actually says that the adjustment is not better, or to put it differently, that the enumeration is better for that purpose too, Justice Scalia, than the adjusted counts.
What he says is, since I can't tell--
Unknown Speaker: That the other one is better.
Mr. Rifkind: --then my rule of decision tells me I rule that way.
Unknown Speaker: Is that unreasonable?
That's... I mean, suppose there are two things causing undercount, A and B, and California has all A and New York has all B. And what we do is we adjust on A and we don't adjust on B. We might discover all the seats go to California, New York loses.
But the reality is that they shouldn't shift at all.
And so, obviously, the simple fact that you get a more accurate count doesn't solve the apportionment problem.
Indeed, all statistics has that problem, doesn't it?
You have to figure out how to draw the boxes and what your causal theory is before you do the sampling.
And as soon as you get into that, you know it or you don't know it.
If you know it, why not just send out the census takers?
If you don't know it, it can be manipulated.
Mr. Rifkind: I think it's unreasonable for at least three or four reasons.
First... and I think this is something that courts are quite as capable of addressing as the Secretary of Commerce.
Everyone knew going into this that what was going to go on was the substitution to some extent of random error, the error derived from sampling, as against what the statisticians call but we also call bias.
And the Secretary has decided here, among other things, that he prefers bias to random error.
And I think the courts are entitled to say that's something one ought to look at very closely because the bias that the Secretary preferred over random error is consistently, persistently decade after decade falling in the same places.
It doesn't float around and it has a very distinct sense of unfairness.
There are a lot of respects in which we use random methods.
When we pick a jury or when we pick a draft, we use random methods because there is a heightened sense of equity in it, and I think that's a relative consideration.
But secondly, the Secretary had agreed with us in the stipulation which this case produced on a mechanism that was going to be employed as part of the census.
At the time he did that, he had before him the machinery that the Census Bureau had prepared.
The decision he finally came to, after... and everyone agrees that the post enumeration survey was excellently conducted.
The Secretary doesn't dispute that.
So, the decision he finally came to was a decision he could have come to before he started because he knew that a measurable, foreseeable amount of random error was going to be interjected into the affair.
I think that's inconsistent with, if you like, it's not in good faith with, the stipulation that he entered into at the beginning saying we're going to conduct this post enumeration survey and we're going to conduct it as part of the census under the ground rules, under the tests, under the standards that the Bureau has set forth.
At the end of the day, it met all those tests.
Suddenly there was a change of grounds of decision, a change of the rules of decision, if you like, and I think that is suspect and requires further attention from a court, or at least I think that's what the Second Circuit was saying.
Unknown Speaker: Mr. Rifkind, are you... you are relying at bottom on a constitutional claim.
Mr. Rifkind: Absolutely.
Unknown Speaker: Under the census clauses of the Constitution or the Fifth Amendment as well?
It's not clear to me.
Mr. Rifkind: Well, maybe that's because it's not clear to me, but I think the answer is both, Your Honor.
I think we rely on what underpins the line of decisions, Wesberry and everything since Baker v. Carr, and I think those come both from the... according to their respective numbers--
Unknown Speaker: What is the specific harm claimed?
The change in congressional reapportionment?
Mr. Rifkind: --It certainly includes that.
There was... I should be clear because the Solicitor General questioned this in his brief.
There's extensive testimony in the record uncontested that congressional district by congressional district... and we used the example of California... the number of people uncounted varied directly with the number of minority in the congressional district and, to be frank, varied directly with the number of Democrats registered in the district.
So, there is a close... and the regression analyses that were done on it were very meticulous... a very close and overwhelming--
Unknown Speaker: Is there some claim of harm because of the intrastate districting that results ultimately?
Mr. Rifkind: --Absolutely.
Unknown Speaker: You claim that as well.
Mr. Rifkind: I who live in New York say that whether or not New York is entitled to an additional Congressman, New York City is entitled to a larger share of the electoral power of the State of New York both in Congress and in Albany.
And the same thing is said by my colleagues from Los Angeles who are here, although California--
Unknown Speaker: But, Mr. Rifkind, then we get back to the point that I made before.
That's not because the Federal authorities are compelling that.
That's because the States, so far as this record shows, have chosen that that's the way they're going to do their interstate.
There's nothing that shows that the States could not one by one have picked what you say is the better way of doing it.
They have the numbers.
They could do it.
Mr. Rifkind: --I think in my view in principle that ought to be right.
I think in practice, with all respect, I think it's wholly unrealistic.
Unknown Speaker: But practice... how does the practice rise to the level of a constitutional decision?
We're saying we know that States could do this, but they won't.
Politically they won't.
Mr. Rifkind: No, it's not only politically.
They're not equipped to do it as a general matter, and Congress has required--
Unknown Speaker: --Well, thought you agreed that you got the numbers for them and those numbers, I take it, are not going to be any less available in the future.
Mr. Rifkind: --But they are numbers that until this day had been under the great cloud that the Secretary of Commerce put them under, and that's why I say it would be a courageous State Attorney General in California or New York who said, well, we're going to... the Secretary of Commerce rejected that, but that's what we're going to use in New York.
I think that takes a temerity--
Unknown Speaker: Because New York would violate one person/one vote if it used those?
Do you think that anyone could mount such a challenge?
Mr. Rifkind: --The New York Constitution requires.
Now, you say the New York Constitution could be amended, and I agree with that.
All I'm saying is as a practical matter... and I can't get further than that... it's unrealistic.
I think to put it differently, the claim of the New York City dweller or the Los Angeles... or of Los Angeles or of the Los Angeles dweller is fairly traceable to the action taken by the Secretary of Commerce.
Chief Justice Rehnquist: Thank you, Mr. Rifkind.
The case is submitted.
Argument of Chief Justice Rehnquist
Mr. Rehnquist: I have the opinion of the Court to announce in No. 94-1614, Wisconsin versus the City of New York and companion cases.
In preparing to conduct the 1990 Census, the Secretary of Commerce determined that past censuses which were essentially headcount consistently had undercounted the actual population and that some minorities had been undercounted at a greater rate than the rest of the population.
As part of an effort to improve the quality of the 1990 Census, the Secretary considered the possibility of using a statistical adjustment of the actual count in order to correct for the undercount.
But he decided against that possibility after deciding that he could not be certain but the statistical adjustment would improve the quality of the Census.
Several cities and individuals sued arguing that Secretary’s decision not to use the statistical adjustment was unconstitutional.
The District Court rejected their argument but a divided panel of the Court of Appeals for the Second Circuit reversed the District Court.
It noted that the Constitution requires Census data to be used to apportion members of the house of representative and then looking at this Court’s decisions requiring States to adhere that the principle of one person one vote decided that the Secretary’s decision not to use the adjustment infringe on the rights of individuals to have their votes counted equally.
In an opinion filed with the Clerk today, we reverse.
We hold that the reasoning underlying our one person one vote cases does not apply to the Federal Government’s decisions regarding the conduct of the Census.
The Constitution vests Congress with virtually unlimited discretion over the conduct of the Census and through the Census Act, Congress has delegated its broad authority over the Census to the Secretary of Commerce.
The Secretary’s conduct of the census is constitutional so long as there is a reasonable relationship to the accomplishment of an actual count of the population.
Here the Secretary based his decision against statistical adjustment on three conclusions: First, he decided it was more important for the Census to accurately measure the proportion of people in different areas than it was for the Census to accurately measure the total number or people in the United States.
Second, he found that statistical adjustment of the preliminary Census data was something that had never been tried before, and finally he determined that statistical adjustment would not improve the distributive accuracy of the Census.
We found the first state conclusion consistent with the constitutional purpose of the Census, and we find the second and third conclusions to be reasonable.
Therefore we hold the Secretary’s decision not statistically adjust the 1990 Census is constitutional.
The opinion is unanimous.