Bowen v. Owens

LOCATION:Heath Residence

DOCKET NO.: 84-1905
DECIDED BY: Burger Court (1981-1986)

CITATION: 476 US 340 (1986)
ARGUED: Feb 26, 1986
DECIDED: May 19, 1986

Carolyn B. Kuhl – on behalf of the appellant
Gill Deford – on behalf of the appellees

Facts of the case


Audio Transcription for Oral Argument – February 26, 1986 in Bowen v. Owens

Warren E. Burger:

We will hear arguments next in Bowen, the Secretary of Health and Human Services, against Owens.

Ms. Kuhl, you may proceed whenever you are ready.

Carolyn B. Kuhl:

Thank you, Mr. Chief Justice, and may it please the Court:

This case presents two issues for decision, first, the constitutionality of certain provisions of the Social Security Act in effect between 1977 and 1983.

The provisions at issue permit payment of survivors’ benefits to widows and widowers to continue even after they remarry after age 60, but payment of survivors’ benefits to surviving divorced spouses ceases upon their marriage even if that event occurs after age 60.

This differing treatment of widowed spouses and divorced spouses is challenged under the equal protection component of the due process clause of the Fifth Amendment.

The second issue in the case concerns the propriety of the classification entered by the district court, and this issue requires the Court to consider yet again the requirements of Section 405-G of Title 42, jurisdictional provision for review of Social Security benefit claims.

Before getting into the merits of the argument, I would like to clarify just how the provisions in question work, because the terminology used is not always as illuminating as it could be.

A wage earner is entitled–

Byron R. White:

That’s a marvelous understatement.

Carolyn B. Kuhl:

–Thank you, Justice White.

I thought that was almost the hardest part of this case.

A wage earner is entitled to have his or her family receive certain types of benefits based on the wage earner’s earnings attempt.

The type of benefits involved here are called survivors benefits, and that means that they are benefits paid to certain members of the wage earner’s family because the wage earner has died.

Now, the wage earner’s widow or the wage earner’s widower receives benefits after the wage earner’s death, and they… these benefits are called surviving widowed spouse benefits, and they come rather automatically.

That is, the spouse is presumed to be dependent for purposes of the Act.

If the wage earner, however, was divorced before he or she dies, the wage earner’s former spouse receives benefits after the wage earner’s death if the marriage between the wage earner and the former spouse lasted at least ten years, and these types of benefits are called surviving divorced spouse benefits.

Although the rule used to be that survivors’ benefits terminated upon remarriage of the spouse, and this was called the remarriage rule, in 1977 Congress eliminated this remarriage rule for widowed spouses over the age of 60, but all other survivors’ benefits still continued to terminate on remarriage, and this was true not only for divorced spouses’ benefits but also for benefits to dependent parents of the wage earner and dependent children of the wage earner.

So, in 1977 Congress made a distinction for purposes of widows and widowers but it still kept in another category not only the divorced spouses but also the dependent children and dependent parents.

There are three basic reasons why the step Congress took in 1977 has a rational basis.

First, Congress has always treated married spouses different from divorced spouses, both because the entire structure of the Act is based on the family unit, and because Congress has assumed that divorced spouses depend less on each other for economic support than do couples who stay married.

It was rational for Congress to recognize these differences between married and divorced spouses on the event of remarriage because–

Harry A. Blackmun:

Ms. Kuhl, you say they have always treated them differently.

Is that correct?

Carolyn B. Kuhl:

–I believe it is, Justice Blackmun, until 1983 when… which is after the–

Harry A. Blackmun:

What about prior to 1977?

Carolyn B. Kuhl:

–Prior to 1977, yes, they were treated differently because initially, of course, widows and widowers… well, widows at least were getting benefits and divorced spouses were not getting benefits at all, and then Congress permitted divorced spouses to get benefits, but that divorced spouses had to be married for 20 years and to meet certain criteria of dependency before they could get benefits.

So, in other words they were treated differently because Congress had a dependency test all along with regard to divorced spouses but it did not have a dependency test with regard to widows and widowers.

To briefly, then, state the second reason why… the second rational basis for Congress’s action here, when Congress permitted some divorced spouses to receive benefits, Congress created a situation where there was a potential for two spouses to be receiving… one, a current spouse and one a former spouse, to be receiving benefits at the same time based on the one wage earner’s account, and it was–

John Paul Stevens:

May I go back a little?

John Paul Stevens:

I just want to be sure.

You were sort of interrupted.

What was your first reason again, that they were always–

Carolyn B. Kuhl:

–They had always been treated differently, both because there were different assumptions about dependency of divorced spouses, and because there was a… the whole system is based on the family unit.

John Paul Stevens:

–Our relevant period is after 1977?

Carolyn B. Kuhl:

That’s right.

John Paul Stevens:

And during the period after 1977, and confining our attention to survivors when the wage earner… after the wage earner’s death–

Carolyn B. Kuhl:


John Paul Stevens:

–Congress then didn’t treat them differently except for the one thing that’s in issue here, isn’t that right?

Carolyn B. Kuhl:

Well, they still treated them differently because there still was this ten year dependency test, if you will.

John Paul Stevens:

They had to have been married for ten years?

Carolyn B. Kuhl:

For ten years.

John Paul Stevens:

How was that… that’s why there are two different classes.

One is a divorce and the other is… the wife had to be married also.

Carolyn B. Kuhl:

She had to be married, but didn’t have to be married for ten years.

The ten year test is really a dependency test and it’s similar to dependency tests that Congress uses for dependent parents.

John Paul Stevens:

Is it a dependency test… if the divorced wife was a millionaire, wouldn’t she still get the… had been married for ten years, would she not be eligible?

Carolyn B. Kuhl:

Yes, but the fact that Congress may have been over-inclusive… I mean, this Court has recognized in many earlier cases that in a Social Security system Congress is forced to deal with sort of broad, generalized criteria.

It can’t have individualized tests.

Harry A. Blackmun:

Isn’t that a test of genuineness rather than dependency, as Justice Stevens points out?

Carolyn B. Kuhl:

I’m not certain I understand the question.

Harry A. Blackmun:

Well, so that there wasn’t a marriage just before death in order to qualify?

Carolyn B. Kuhl:

No, it wasn’t, because of the–

Harry A. Blackmun:

How can you say that?

Carolyn B. Kuhl:

–Well, first of all, this Court has, I believe in Califano versus Jobst, there’s sort of a chronology that the Court follows through in its opinion, but the way this ten-year test evolved, the way it evolved was, as I mentioned previously, there were no benefits for surviving… for divorced spouses, and then Congress had a two-pronged test of dependency.

The reason was because they were concerned about women in particular who had been married to a wage earner for a long period of time with perhaps having no job themselves and then were divorced, so the criteria that Congress set up were, number one, marriage for 20 years and number two, some… a more specific dependency test that had to do with whether the divorced spouse was receiving alimony while the wage earner was still alive.

Then, I believe in 1972, and again this is recounted in the Jobst opinion, Congress became a little bit concerned about the more specific criteria of dependency because Congress realized that there were states that did not permit alimony under state law.

And so, in response to that problem, Congress said, all right, we’ll be even more broadly inclusive and we won’t have the specific criteria of dependency but we’ll continue in effect the 20-year marriage requirement, subsequently reduced to a ten-year marriage requirement.

If the test… I think if the test was, was it really a sham marriage, it would be a much shorter period of time.

Harry A. Blackmun:

As between the divorced spouse who satisfied the ten-year period, and the widow, no difference?

Carolyn B. Kuhl:

No, not as of 1977, with regard to survivors’ benefits.

But as we know, and I believe it’s footnote 2 of our reply brief, divorced spouses were still being treated differently for other purposes; that is, with regard to what are called spouses’ benefits which are benefits that a divorced spouse might receive when the wage earner is still alive, the remarriage rule continued, even after 1977.

So that, in essence, Congress in 1977 was treating divorced spouses more alike and widowed spouses more alike.

Before discussing these points in a little bit more detail and going over two more of the rational bases on which Congress acted, I’d like to say a word about the legal standard to be applied in judging the constitutionality of Congress’s actions here.

It seems clear that the standard should be the rational basis test, and district court in this case in fact applied the rational basis test.

In the context of the Social Security Act, this Court has measured classifications based on divorce, in Matthews versus De Castro, and based on remarriage in Califano versus Jobst, according to a rational basis test, and there’s no reason to do otherwise here.

The test, then, is simply whether Congress’s action can be said to manifest a patently arbitrary classification utterly lacking in rational justification, and I think the district court’s error in applying this test is fairly easy to pinpoint, and it goes to the rationale that we have been discussing previously, rationale of differences in dependency between divorced spouses and widowed spouses.

The district court seemed to believe that when Congress gave survivors’ benefits to divorced spouses, it committed itself to the position that there was no basis whatsoever for distinguishing between divorced spouses and widowed spouses.

But what the district court failed to recognize is what I’ve stated previously, that Congress has always treated, even in 1977, has treated divorced and widowed spouses differently, and the fact that Congress chose to treat them the same for purposes of initial eligibility for survivors’ benefits does not mean that Congress was requiring them to treat them the same for all purposes.

It was rational for Congress to give effect to he differences between the two groups at the time of remarriage.

I’ve discussed previously in answer to questions the chronology of Congress’s first giving no benefits to divorced spouses, thereby recognizing the reality of the legal severance of relationship at the time of divorce, and then giving benefits to divorced spouses but using two criteria of dependency, 20-year duration marriage requirement plus some more specific criteria, and Congress then being concerned that the more specific criteria were not working well and retaining only the 20-year marriage requirement and then reducing the 20-year marriage requirement to a ten-year marriage requirement but still retaining that as a criteria of dependency.

But all of this suggests that the differences between divorced spouses and widowed spouses have been at least as apparent to Congress as their similarities.

Thus, when Congress considered changing the rule that ended survivors’ benefits on remarriage, and when Congress chose as it did to proceed a step at a time with regard to the elimination of this remarriage rule, Congress acted quite reasonably in eliminating the remarriage rule, first for the one category of beneficiaries for whom there had never been a dependency test, and that was widows and widowers.

Congress at the same time then continued, in effect, the remarriage rule for all other categories of beneficiaries, those categories being categories where there were dependency tests attached which was, as I’ve stated, not only the surviving divorced spouses but also dependent parents for whom there was a dependency test, and dependent children for whom there is a dependency test.

Put another way, Congress could rationally assume that its concerns that had been manifested in prior legislation about divorced spouses who had not worked during the years when they were married, could be set aside on the event of the remarriage of the divorced spouse.

I would like to briefly mention two other things that can be said about the rationality of the distinction that Congress chose here in 1977.

First of all, as I have alluded to very briefly earlier, when a divorced spouse receives survivors’ benefits there is a possibility that another spouse may also be receiving benefits based on the one wage earner’s account.

Benefits to divorced spouses are not counted against the ceiling amount that may be paid based on one single wage earner’s account.

Thus, Congress’s decision to pay benefits to divorced spouses creates the potential for… and I don’t mean this pejoratively but just descriptively… as sort of a double dipping based on the one wage earner’s account, and this is a special kind of drain on the trust fund.

John Paul Stevens:

Is that possible in the other situation where the widow’s benefit… could the deceased wage earner owe a divorced… maybe was married ten or 15 years and divorced and then had a second wife, could he have double obligation?

I mean, could the account have that double–

Carolyn B. Kuhl:


If I understand your question correctly, that would be–

John Paul Stevens:

–So that either the widow or the divorced wife survivor could be a double dipper?

Carolyn B. Kuhl:

–That’s correct, although I think as your description of the problem indicates, it really is best to find by the situation of looking where there’s been a divorce.

I think that is a–

John Paul Stevens:

There’s got to be at least one divorce?

Carolyn B. Kuhl:


John Paul Stevens:

Both categories?

Carolyn B. Kuhl:

That, I think, is a better rational classification.

Carolyn B. Kuhl:

Of course we don’t have to… I think the Court does not have to look to whether it’s better or not, but I think it’s a rational classification to look to the situation where there has been a divorce because I think that even creates the probability, not the certainty but the probability, of this, what I’ve called double dipping.

And it seems clear that it would be rational for Congress to pick the event of the divorced spouse’s remarriage and this special rain on the account because at that point Congress… again, a new family unit was formed and a new basis for economic dependency accrued with the new marriage.

Finally, despite the appellee’s claim to the contrary, Congress clearly was motivated, certainly could have been motivated by fiscal concerns when it acted in 1977.

Harry A. Blackmun:

I’m interested in your saying it was, and then changing to the possible “could have been”.

What in the legislative history supports your argument?

Carolyn B. Kuhl:

Well, first of all I do not believe that the rational basis test requires us to discern Congress’s actual motivation, so long as Congress could have rationally acted with this purpose, and in this case, first of all, I think as to dependency, Congress’s purpose can be discerned from the progress of its concerns over the years with regard to widowed spouses.

With regard to its fiscal concerns, what we have is kind of… is the progress of the legislation itself.

The House would have eliminated the remarriage rule entirely and the legislative history indicates that that would have cost $1.3 billion per year.

The Senate bill had no provision in it with regard to the marriage rule.

Harry A. Blackmun:

Well, I take it you’re saying, although you are hesitant to express it, that there is nothing in so many words in the legislative history.

You have to speculate on the fiscal considerations and what Congress might have done.

The legislative history is very sparse.

Carolyn B. Kuhl:

Well, yes, it is.

There is a House Committee report.

The Senate did not consider the issue, and the Conference Committee acted really without an explanation in the conference report, but this Court has never held that an actual motivation must be discerned in order for the Court to find that Congress acted on a rational basis.

William H. Rehnquist:

How did the bill which passed the Senate differ from the bill which passed the House?

Carolyn B. Kuhl:

I couldn’t tell you in all the particulars, but the bill which passed the Senate had no provision in whatsoever addressing the remarriage rule.

William H. Rehnquist:

And did the Senate consider it after the House?

Carolyn B. Kuhl:

I’m not certain.

In sum, to just draw a conclusion with regard to this fiscal concern, I think that the appellees have agreed that the… by choosing the categories that it did in 1977, Congress saved about $17 million per year and this Court should be unwilling to assume that $17 million per year is simply too little for Congress rationally to be concerned about.

The second issue in this case relates to the class definition.

The district court lacked jurisdiction over two groups of individuals that it purported to include in the class.

First, it lacked jurisdiction over persons who did not exhaust their administrative remedies, and second, it lacked jurisdiction over persons whose claims had already become final when the case was filed.

In this respect the district court erroneously included some persons who had received a final decision of the Secretary, more than 60 days before the case was filed.

First, perhaps I should say something about whether the Court needs to reach the class certification issue if it decides the constitutional issue in favor of the government.

The Court in this case does have jurisdiction, indisputably, to decide the constitutional issue because the individual plaintiffs in this case did meet the 405-G requirements.

In similar situations, that is where there is jurisdiction with regard to some plaintiffs in the case but there is a question with regard to the overall class certification, the Court has in some cases ruled in favor of the Government and failed to decide the class certification issues, and it did that in Califano versus Boles.

But in other cases such as Matthews versus Diaz, the Court has in the very same circumstances gone on to decide the class certification issue.

And if I may suggest, the Court may wish to give guidance to the lower courts on this issue, related issues with respect to class certification arise in the next case to the argued, the City of New York case, and the Court of Appeals in this situation here would be required to decide the class issues because it would need to define the collateral estoppel effect of the judgment that it was entering.

As I have stated, both class issues relate to 405-G which requires, first of all, that there be a final decision of the Secretary on a claim for benefits, and secondly that the civil action be commenced within 60 days of notice of the Secretary’s final decision to the claimant.

Carolyn B. Kuhl:

This Court has held that these requirements must be satisfied with regard to each individual class member, and the Court has so stated in Califano versus Yamasaki.

In that case the Court emphasized that class relief is consistent with the need for case by case adjudication, that is evident in the structure of 405-G, only if every member of the class can meet individually the 405-G requirements.

The first requirement that was not met here by some of the class members was the requirement of a final decision of the Secretary.

The district court did not limit the class to those who had exhausted the entire administrative process.

This Court in Weinberger versus Salfi has held that the exhaustion requirement is statutorily required and that it cannot be dispensed with by a judicial conclusion of finality.

The Secretary himself can make exceptions to the exhaustion requirement and the Secretary did so in this case with regard to the individual claimants, but did so because the Secretary had reached the stage of the administrative process where the facts had been found, the issues of law had been determined insofar as the Secretary could determine them, and the only thing that was left to be decided was the constitutional issue which the Secretary could not rule on.

Now, the class members, however, had not been through this process and I think that the importance of exhaustion can be seen in this case by looking at what happened after the district court decided the constitutional issue.

At that point, the agency was required to go through the files of the individual claimant and decide who would qualify for the relief that the Court granted.

This is just the kind of fact finding that is supposed to go on before the case goes to court.

Finally, the remaining issue concerns the requirement that a claimant file suit within 60 days after the final decision of the Secretary.

I think that Califano versus Sanders holding, that motions to reopen are unreviewable–

John Paul Stevens:

May I go back to your first point about the fact finding issue.

Is it not correct that if you prevail on the merits you’d never have to engage in all that fact finding, whereas under your view you’d have to do it regardless of who wins?

Under your view, as I understand it, they can’t have one decision binding on everyone, you’ve got to go through all this extra fact finding regardless of what the general ruling would be on the basic proposition?

Carolyn B. Kuhl:

–Well, it would not be one decision.

John Paul Stevens:

If you win, everybody in the class loses, and if you have all these hearings you’ll find out if there a another reason why they might–

Carolyn B. Kuhl:

Well, I’m not sure that everybody in the class loses.

Insofar as the Court had no jurisdiction… insofar as the class was improperly certified and the Court had no jurisdiction… I’m not sure what the collateral estoppel effect would be as to those persons who were improperly included in the class.

John Paul Stevens:

–It’s not collateral estoppel.

They just lose on the merits, if you’re right on the basic–

Carolyn B. Kuhl:

What I’m saying is that some of those people who were nominally included might be able to come back and make the same claim later if collateral estoppel did not bind them, and I think maybe more importantly, the Secretary may well have to go through the fact finding in each of these claims anyway.

The persons may have other bases on which they would claim benefits and the Secretary may have other matters to adjudicate with regard to these individuals.

If I may, I would like to reserve the remainder of my time.

Warren E. Burger:

Mr. Deford.

Gill Deford:

Thank you, Mr. Chief Justice, and may it please the Court:

I’d like to correct one statement that my opponent made.

There is no double dipping at issue in this case.

No individual with a surviving divorced spouse or a surviving spouse who is not divorced will receive any extra benefits as a result of this case.

Warren E. Burger:

If the rule were otherwise, would it be possible in some cases?

Gill Deford:

If the rule that’s at issue in this case were otherwise?

Warren E. Burger:


Gill Deford:

As far as I know, there’ no way that a Social Security recipient could receive more than he or she is entitled to under the most liberal possible benefit scheme that he or she is entitled to.

If she is entitled to more than one benefit, she will receive only the highest of those two benefits, not a combination of the two benefits.

John Paul Stevens:

I don’t think her point was that any beneficiary might receive double benefits.

Rather, a given wage earner’s account might have to pay two different people benefits?

Gill Deford:

That’s correct, Justice Stevens.

I thought there might be some confusion in the Court about that because of the use of the word, “double dipping”.

The constitutional issue of this case is the validity of the Social Security Act scheme which treats two classifications of survivor beneficiaries differently for a reason unrelated to the overall purposes of the scheme.

In so doing, the discrimination contradicts two crucial ingredients on which Congress has premised its distribution of survivors’ benefits.

Those two factors, dependency on the deceased wage earner and the encouragement of survivors to remarry, are applicable in all relevant respects to all widowed spouses, both non-divorced and divorced.

The challenged legislation, however, treats them differently upon remarriage, despite the fact that both classifications of widows, again non-divorced and divorced, are entitled to survivors’ benefits in the absence of a remarriage.

It is appellees’ contention that Congress’s disfavored treatment of divorced widows at remarriage was not the product of reasoned legislative analysis but was instead an unthinking and stereotyped response spawned by the urge to reach a compromise.

Now, there are two–

William H. Rehnquist:

You can certainly say that about a lot of legislation, can’t you, that was spawned by the urge to reach a compromise?

Gill Deford:

–That may be an overstatement on my part, a generalization, Your Honor.

I think it is true that this instance more than most indicates that Congress wanted to make a deal quickly and was looking for some way to make that deal, and rather than carefully investigate who the various classifications were and what the reason, what the rationale for that deal was, simply selected a political group that was handy, a discrete group that was handy, to cut the deal.

I think that was clearer in this instance than in most other Social Security Act situations we have where one classification is disfavored as opposed to another.

William H. Rehnquist:

Supposing that the legislative history in this case were perhaps clearer than it is, so that it was apparent that the Senate’s balking at the inclusion of the divorced and remarried… was purely financial, that their view was that it’s going to cost us more money than we have now, we may be ready to do it next year but we can’t do it this year, we just don’t have the money now, so we’ll not give benefits to as many people as the House wanted to, in your view is that irrational?

Gill Deford:

I would think that would probably be irrational because–

William H. Rehnquist:


Gill Deford:

–Irrational, because of the amount of money at stake.

It seems to me that the more that Congress indicates, the Senate or the House or both, that it’s planning ahead, this one step at a time analysis, and planning to take care of everything when it can, the more clearly that is set out and the more bases it gives for doing that, the more likely the scheme is to be rational.

But in the situation you describe where all the Senate is doing is simply saying, there’s a few dollars more to be spent here and we don’t want to spend it at this time, I don’t think that that–

William H. Rehnquist:

How much would have to be involved for the Senate’s action to be rational?

Gill Deford:

–Obviously, I’m not in a position to draw that line, but we pointed out in our briefs, the amount of money at stake here at the absolute most was less than one one-hundredth of a percent of the annual outlay of the Social Security system.

William H. Rehnquist:

What was it, $17 million?

Gill Deford:

It was less than $17 million a year.

We do not know the exact figure.

William H. Rehnquist:

What, $15 million?

Gill Deford:

It’s unclear from the legislative history how much less than $17 million, but that’s the maximum that it could have–

William H. Rehnquist:

But your argument is, anyway, that that small an amount, as you put it, makes it irrational?

Gill Deford:

–No, not that that amount alone makes it irrational, but in the context of what happened between the House and Senate in this situation, I do not think it is reasonable to assume that fiscal concerns motivated Congress in this case.

William H. Rehnquist:

So you say that with that small an amount, it is irrational to think that the Senate would be motivated by concern for saving money?

Gill Deford:

Well, you hypothesize that the Senate set that out in its legislative history, which of course is not the case we have here.

William H. Rehnquist:

Yes, I’m hypothesizing, that the Senate said, we are concerned about this $17 million?

Gill Deford:

I would say that even if the Senate had said that, without anything else to explain why it selected this particular discrete classification of divorced surviving spouses, that it’s inappropriate for the legislature simply to say, they’re not going to get the benefits but people in a virtually identical situation are.

There has to be a rationale that’s either set out in the legislative history which we can reasonably infer to have motivated Congress to have selected that particular classification.

William H. Rehnquist:

Other than a desire not to spend as much money as the proponents wanted?

Gill Deford:

I think just fiscal concerns alone cannot provide the basis for discriminating against a particular classification.

This Court in the Fritz decision, in U.S. Railroad Retirement Board v. Fritz, rested in large part on the fact that Congress was seeking to protect the Railroad Retirement Fund, but it still carefully determined what the rational basis was for selecting the particular group of railroaders whose benefits would be cut.

William H. Rehnquist:

Of course, there, there were people being cut.

Here it’s a question of people getting an additional benefit.

Gill Deford:

I don’t think there’s any indication that I’m aware of, in this Court’s decision, that that distinction should make any difference for purposes of the importance of fiscal concerns.

Fiscal concerns are, of course, always relevant but I think there has to be some basis to draw the line, some rational basis other than fiscal concerns, to draw the line between various classifications.

As I said, there are two consistent themes in the development of the Social Security survivors’ programs, and both of those are raised in this case.

One of those themes is Congress’s recognition repeatedly over the years of various classifications whose dependency on a deceased wage earner renders them entitled to benefits under the survivors’ program.

And the other theme which runs parallel to that is Congress’s persistent desire to eliminate the remarriage penalty.

The remarriage penalty, of course, operates so that if a survivor beneficiary remarries someone else, the general rule was, that person would lose her Social Security benefits.

But Congress has gradually eliminated the remarriage penalty in various circumstances, and it is indicated that the general remarriage rule, the remarriage penalty, should essentially be eliminated altogether.

So, those two themes, I think are crucial in understanding appellees’ contention in this case.

By 1965 Congress had recognized that divorced surviving spouses’ dependency on the deceased wage earner was sufficient that they should be entitled to survivors’ benefits.

This rendered them effectively the same in all relevant respects as non-divorced surviving spouses.

The remarriage penalty, however, was still in effect and Congress recognized that it created a serious dilemma for many widows.

In 1977, therefore, Congress set out to remove the remarriage barrier, but as we have noted, removed it only for one classification of surviving spouses, those who were not divorced, what we normally call widows and widowers.

I think it is very important to understand what happened in 1977.

The House voted to remove the remarriage penalty for both surviving divorced spouses and for surviving spouses who were not divorced.

The Senate said nothing about this issue and indicated it had no intention of removing the remarriage barrier for either classification.

The issue then went to conference committee, and that conference committee as we have noted, essentially split the difference and decided to–

John Paul Stevens:

May I ask you a question.

Your opponent makes the point that the remarriage rule applies not just to the spouses but also surviving parents and children, I guess.

John Paul Stevens:

In the House version would they have removed it entirely, or just in the spousal category?

Gill Deford:

–I believe the House version would have removed the remarriage penalty for virtually every classification, including the ones you’ve mentioned, and that’s why the $1.3 billion figure which has been mentioned is higher than we’re talking about because it includes other classifications as well.

Now, when the Conference Committee decided to split the difference, as I’ve called it, and remove the remarriage penalty only for widows, not for divorced surviving spouses, it provided absolutely no explanation for this decision.

It was clear at that time that Congress as a general rule did want to eliminate the remarriage penalty, but no explanation was provided as to why that penalty should continue for one group of surviving spouses and not for the other group of surviving spouses.

The irony is that far from providing a reason to treat these two classifications differently, remarriage if anything should have provided an incentive, as the district court observed, to treat them the same.

Congress had therefore by 1977 carefully reasoned one two crucial factors.

One, why survivors’ benefits should be available to surviving divorced women, namely, their dependency on the deceased wage earner and secondly, why the remarriage penalty should be eliminated as a general rule.

Congress provided by that time no reasoning whatsoever about why the remarriage penalty should continue for this one discrete classification of surviving spouses.

Congress based its reasoning, to the extent there was any reasoning at all, solely on divorce but it did not explain why divorce in that context was a relevant characteristic to discriminate between two classifications.

Consequently the government throughout this case in the district court and the briefs for this Court, today in oral argument, has been forced to offer a series of after the fact rationales in an effort to explain what Congress might have had in mind when it treated these two classifications differently.

Warren E. Burger:

Well, is it your position that Congress has got to announce its reasons, or is it true that there must be discernible irrational basis for what they’ve done?

Gill Deford:

Mr. Chief Justice, it’s clear that… to this Court’s rulings, that if the lowest standard of review, the rational basis test is used, that the actual reasons do not have to be set out in the legislative history.

It’s also clear, however, as this Court has indicated in various opinions, that it certainly would prefer if the reasons were set out, and that if the reasons are not set out, the Court should at least attempt to determine if the reasons advanced by the government could reasonably be presumed to have been the reasons which motivated Congress.

The discrimination in this case, we submit, simply does not serve the purposes of the Social Security Act, especially as explained by the survivors’ program and by the treatment of the remarriage penalty.

Last term this Court in another equal protection case, Williams v. Vermont, observed that the distinction between the classifications bears no relation to the statutory purpose, and I think the same thing holds true here.

The overriding purpose of the Social Security Act Survivors’ Program has been to recognize dependency on the deceased wage earner and to encourage a survivor to remarry.

These factors are not carried out by the discrimination.

Sandra Day O’Connor:

I’m not sure that’s quite sufficient to describe it.

Hasn’t Congress’s obligation in Social Security been to distribute a finite amount of money among beneficiaries that Congress determines should receive that finite amount?

Gill Deford:

Justice O’Connor, I think we’re talking here about a portion of the Social Security program, the survivors’ program, and the two factors that I have set out, the question of dependency and the remarriage penalty–

Sandra Day O’Connor:

But overall, isn’t it a fact that Congress just has not had unlimited funds in the Social Security Fund to distribute, and it’s had to make some very hard choices, how to distribute those, and what people should receive as benefits?

Gill Deford:

–I think that’s true, as a general rule.

Sandra Day O’Connor:

This is just a part of the overall scheme?

Gill Deford:

This is a part, and a very small part, of the overall scheme, but I think it’s also true that when Congress selects one classification to receive benefits and another classification to be disfavored and not to receive benefits, there has to be some discernible distinction between those classifications which explains Congress’s actions.

Sandra Day O’Connor:

Historically, Congress has made some differences applicable or attributable to divorce status.

Maybe it’s moving away from that today, but I’m sure that you have to concede that over a period of time Congress has opted to make some decisions turn on divorce status.

Gill Deford:

In the past, Congress definitely made some decisions based on divorce status.

But in this case, by 1965 Congress had already determined that divorced surviving spouses were also entitled to survivors’ benefits, essentially equivalent to those received by non-divorced surviving spouses.

And, I think that determination by Congress indicated essentially an equivalence in dependency and therefore–

Sandra Day O’Connor:

Well, do you think that just tied Congress’s hands forevermore and that was the end of it?

Gill Deford:

–No, but I think it indicated that Congress saw so few distinctions between the two classifications that if it were to distinguish between them at a later point, it had to provide, or we had to be able to discern the valid reason for that distinction.

And it’s particularly ironic here that the remarriage, the instance of remarriage, is a basis on which Congress uses to distinguish between them when prior to remarriage it does not distinguish between them.

Both sets of spouses, divorced and non-divorced, are entitled to survivors’ benefits.

It seems logical as the District Court concluded, that if Congress were going to distinguish between the two classifications it would have done so at the death of the wage earner, not somewhere down the line when remarriage takes place.

Just to conclude on this point, we think in this instance Congress was not discriminating between the classifications in order to further a particular goal.

It was discriminating because it needed some group to single out in order to forge a compromise, and it selected the divorced because they were a group which were handy, and the amount of money at stake was relatively small, and a compromise could be made.

We submit that in light of the overall goal of encouraging survivors to remarry, it was totally irrational for Congress to make remarriage the basis on which to discriminate between these two classifications.

The other–

John Paul Stevens:

Before you get to your class action, would you agree it was rational for them to apply the remarriage rule to the parents and the children, not to the widows?

Gill Deford:

–Justice Stevens, I have to confess I haven’t given that any thought.

I believe in 1983 the remarriage penalty was eliminated for parents and children as well.

John Paul Stevens:

I suppose one could argue that there also should be an incentive in those groups, not to discourage–

Gill Deford:

Certainly there’s an incentive, but the reason we brought the case and briefed the case as it is, is because of the similarity between the two groups of spouses, and I’m certain many distinctions could be drawn between parents and children as opposed to spouses, which might make it rational to distinguish between those classifications.

John Paul Stevens:

–I may have it wrong, but as I understand it in ’77 they decided to put the divorced survivors either with the non-divorced survivors, widows, in other words, or with the children and their parents.

Gill Deford:

I think in 1977… I think all that happened in 1977 was that Congress decided to eliminate the remarriage penalty for widows alone, widows and widowers, but not for divorced surviving spouses, parents and children.

John Paul Stevens:

But see, the effect of that was to treat the divorced people in the same way they treated the children and parents.

Gill Deford:

That’s correct, but as I indicated, I think the distinctions between children and parents are such that it’s not logical to compare those classifications.

The other issue in the case, of course, is the district court’s class certification, but I think the real issue and the overriding question involving that is whether the Secretary of Health and Human Services can restrict the size of classes and therefore as a practical matter the number of those who can receive relief through the use of strict, literal and essentially impractical interpretations of provisions of the Social Security Act which were designed for entirely different purposes.

In light of the remedial goals of the Social Security Act, and especially the nature of this claim, constitutional claim, the Court should not permit the Secretary to deflect legitimate constitutional contentions through this mechanism.

Now, the first issue of course is what is known under the rubric of exhaustion, whether an individual claimant has to go entirely through the administrative process before he or she can move into district court and make the contention there.

In this case the Secretary formally and in writing waived full exhaustion for the named plaintiffs.

The Secretary determined that there was a final decision on their claim in the administrative process.

The Secretary determined that there were no factual issues in dispute, and the Secretary determined that there was no statutory interpretation in dispute.

The only claim raised is the one the Secretary was not competent to resolve, namely the constitutionality of the challenged scheme.

This Court in prior cases has recognized waivers by the Secretary of full exhaustion in other constitutional cases.

In Califano v. Goldfarb, in Matthews v. Diaz, Weinberger v. Southey, the Court has recognized the Secretary can waive full exhaustion.

William H. Rehnquist:

No one is challenging the standing of the people with whom the Secretary… with respect to whom the Secretary has waived, are they?

Gill Deford:

No, Justice Rehnquist.

William H. Rehnquist:

This is a question of whether dissimilar people can be included in the class with respect to whom the Secretary hasn’t waived?

Gill Deford:

Well, the crucial issue is whether or not they’re dissimilar, the term you use.

William H. Rehnquist:

Well, certainly they’re dissimilar to the extent that in one group the Secretary has waived and in the other they haven’t.

Gill Deford:

The Secretary has formally waived for one group.

The question is whether a district court can infer from that waiver, and from the nature of the case, that as a practical matter the Secretary has also made a final decision for the other members of the class who are, a defined by the class–

Sandra Day O’Connor:

But Secretaries issue regulations that define when the Secretary is to be deemed to have waived.

Why should a court be permitted to ignore those administrative regulations?

Gill Deford:

–The court is not ignoring those, nor is the court overruling the Secretary.

The district court in this case… and our argument is that a district court, as this Court has done in other instances, can infer that a final decision has been entered by the Secretary.

In the Diaz case the Secretary herself moved to dismiss the complaint for failure to exhaust.

The district court and then this Court held that despite that, as a practical matter the Secretary had to determine that there was a final decision and that no further exhaustion was required.

So, what we are talking here for the most part is not asking the Court to overrule the Secretary but for the Court to analyze as a practical matter what the Secretary has done with respect to whether or not a final decision has been entered in the administrative process.

We think it’s logical for a district court, based on the determination that named plaintiffs have no factual disputes.

It’s logical for a district court to infer that final decisions for the class members are also appropriate.

For the class members as well as for the named plaintiffs there are no factual or legal disputes with the Secretary, nor is there any ability of the Secretary to take any action with respect to the claim at issue which is a constitutional claim.

We think it is very important also to recognize what Justice Stevens was hinting at in his questions to my opponent, that there is no practical purpose served by forcing the class members to go to exhaustion when the named plaintiffs have already received a final decision from the Secretary.

The effect of forcing exhaustion in that context is not only wasteful but it’s counterproductive.

If the Secretary is ultimately successful at whatever level, district, court of appeals or Supreme Court, then all the factual exhaustion which has taken place prior to the filing of the complaint will have been totally wasted because the class members for the most part are not going to be eligible for benefits anyhow, as the Court has held that the statute is constitutional.

So, not only is a waste of administrative resources, but it’s actually counterproductive to the entire system.

We think also that the Court should keep in mind the practical reason why the Secretary is pressing exhaustion in this context, and we think he’s doing it because it’s a shield to protect him from significant relief when plaintiffs are successful in district court.

There is absolutely no practical reason to require exhaustion for named plaintiffs.

Once they receive an initial decision saying they are ineligible for benefits because their facts are such that they are not eligible under the statute, no purpose is then served by requiring them to go through an administrative process.

Once the Secretary determines that a constitutional claim exists, which the Secretary himself has no competence to resolve, the only logical conclusion to be drawn is that a final decision has been reached, so that no further exhaustion is required.

Now, the other component of the class action issue is what is known as the 60-day rule and that is the question of whether class members moved untimely through the administrative process and in the district court.

Now, it is our contention that the notice filed by the named plaintiff, Mrs. Owens, in October 1982 asking for reconsideration of her initial decision was a notice to the Secretary that a constitutional claim was being raised and that it therefore told the administrative and judicial process, this is a conventional principle, it’s been used in many other contexts.

When notice is given, the time is told.

It is commonly used in class actions as a way of protecting the rights of class members.

It’s regularly used in Title 7 cases.

William H. Rehnquist:

Has it ever been applied against the government in any case that you know of?

Gill Deford:

I believe the Honda case, Honda v. Clark.

William H. Rehnquist:

By this Court?

Gill Deford:

That was a Supreme Court decision, yes, Your Honor.

Gill Deford:

The Honda v. Clark holding was used to protect the right of the class members.

William H. Rehnquist:

Ordinarily the rule that several cases in this Court laid out was that when you deal with the government, you turn square corners, and that lots of rules that apply to a private defendant don’t apply to the government.

Gill Deford:

Well, this Court of course has never dealt with this express issue under the Social Security statute.

I would point out also that the briefs of the parties are somewhat misleading on one crucial point, and that is that for 99 percent of the class members the totalling at issue here is not what we might call judicial totalling, it’s administrative totalling.

By filing a notice with the Secretary, by putting the Secretary on notice, what Ms. Owens did was to stop the running of the time for individuals to move through the administrative process.

She was not… there might have been some individuals who already received a final decision of the Secretary, although I’m doubtful of that fact.

It also would have told the time for them to proceed to district court.

But I think it’s important for the Court to recognize, for most of the class members, what was told was the administrative process, until such time as Mrs. Owens had herself received a final decision through the Secretary’s ruling on the reconsideration request.

Now, there are excellent reasons, we think, to apply this general totalling principle in this context.

The remedial purposes of the Social Security Act are one, and the nature of the issue of course is another crucial component.

This is a constitutional challenge and it is imperative, as this Court has indicated on many occasions including the Sanders case, the Social Security recipients have an opportunity to bring constitutional challenges before this Court.

As a practical matter if totalling is not permitted, 99 percent or 90 percent of most class actions will be prohibited from raising a constitutional challenge.

They will never seriously have an opportunity to bring their case to court.

The Court, this Court has construed other aspects of the statute which allows individuals to go to district court in a flexible way under the venue provision, the exhaustion provision, the question of injunctive relief, the question of class action in general.

This Court has indicated that the literal terms of 42 U.S.C. Section 405-G do not necessarily control.

There has to be a practical way of guaranteeing that classes can be certified and the individuals can bring their cases in a realistic way.

The class members in this case are not trying to avoid compliance with the 50-A rule.

They are simply recognizing that as a normal principle, when the Secretary is put on notice, the period for proceeding through the administrative process is told.

No purpose, no practical purpose is served here by a literal application of the 50-day rule.

The only purposes served is that of the Secretary which artificially prevents some claimants from obtaining relief.

Thank you.

Warren E. Burger:

Do you have anything else, Ms. Kuhl?

Carolyn B. Kuhl:

Just briefly, Mr. Chief Justice.

With regard to Honda versus Clark, I’d like to correct the statement that was made with regard to totalling against the Government there.

The Court specifically said in that case that because the Government was acting as a stakeholder with regard to property in the Office of Alien Property, that the court would allow totalling.

There was no potential drain upon th Treasury there and it was an exception to the usual rule with regard to the Government which is not at all applicable here.

Appellees state that the Secretary is trying to restrict the size of the class, but with all due respect it is Congress that has set forth the scheme under Section 405-G and this Court has recognized in Yamasaki and Ringer that Congress wants these benefits determinations to go forward on an individual basis and wants the legal issues to be raised in an individual context.

John Paul Stevens:

May I ask, in connection with both that point and the totalling point, the Government disagrees with the date that they picked for the 60-day rule, the date that the named plaintiff received notice.

Do you agree, though, that if the date had been advanced to the date of filing suit, that there would be totalling for all the class, and you redefined the class with reference to that date, the totalling would be proper?

Carolyn B. Kuhl:

It would not… we would not consider it a question of totalling.

Carolyn B. Kuhl:

If someone has a final decision of the Secretary within 60 days before the suit is filed, it’s not really a question of totalling but rather that in lieu of filing a separate action they may be considered to have filed vis-a-vis that action, and that’s the basis.

John Paul Stevens:

In other words, they can be properly included within the class even though they don’t file individual actions?

Carolyn B. Kuhl:


We have not raised anything to the contrary in this case.

The point here on the exhaustion and 60-day principles is that we’re dealing with individuals who did have an opportunity to raise the constitutional issue in the administrative process but didn’t.

They let their claims become final.

It was res judicata, and what appellees are asking for here is that this Court recognize the principle that allows continuous correction process to occur with regard to the system, and that is not at all what Congress had in mind.

These individuals gave no indication that they wanted further review.

At each stage of the four-stage administrative process there is a notice sent stating, if you wish further review file within 60 days.

The people who are included in the class here who should not have been did not take advantage of that opportunity.

The rule that they are asking for here is almost as if they are asking that when a court of appeals creates a new rule of law, district court people who didn’t think to raise the argument and who have… had their cases decided be allowed to re-open them over and over.

Thank you.

Warren E. Burger:

Thank you, counsel.

The case is submitted.