Oral History of Judge Stephen F. Williams
May 5, 2013
Mr. Nuechterlein: It is May 5, 2013, and we are back after a five-year hiatus to resume Judge
Williams’ oral history interviews. Judge, we left off at the point in 1986
when you had been confirmed by the Senate and were preparing to take
office. Remind us of the events through your first days on the bench.
Judge Williams: I came back from Colorado. I interviewed a harvest of Scalia clerks.12 And
I had lunch with Pat Wald, who was very nice.
Mr. Nuechterlein: She was the Chief Judge?
Judge Williams: She was the incoming Chief Judge. It must have been—just before or just
after July 1.
Mr. Nuechterlein: Who was it before that?
Judge Williams: It was Spotts Robinson. I don’t know if Pat had actually taken office or
was about to. And she was very gracious. We chatted amiably. I am not
sure there were any special lessons I picked up in that discussion, but it
gave me a comfortable feeling. There was also a party for some judge on
the court, and it was at Pat’s apartment, I’m pretty sure. That was a very
Mr. Nuechterlein: And that included the rest of your colleagues?
Judge Williams: Yes.
Mr. Nuechterlein: Did you know any of them before you arrived at the court?
Judge Williams: Well, I knew two of them. I knew Nino [Scalia] because he had been head
of the Administrative Conference when I did a paper for them. And I knew
Jim Buckley, though not well at all, because he and I had been at a dinner
at Bill Buckley’s.
What I remember particularly about Jim was that he was flat on his back
with a backache, which he happily recovered from. Later, when I had
terrible backaches, I followed his procedure, and that enabled me to see
something [interesting] in the court’s robing room. Lying on my back in
the robing room, I noticed that there is sort of an entryway before you get
to the door into the courtroom, and at the very top of this sort of arch was
what looked like a track for a curtain. It was my guess — and it has been
neither confirmed nor rejected — that the track once held a curtain to
produce a certain muffling of sound. And that [would have been useful] if
12 Then-Judge Scalia had hired these clerks for D.C. Circuit clerkships before he was elevated to the Supreme Court.
the rumor was correct that Judges Bazelon and Burger13 used to scream at
each other in the robing room. As I say, that hypothesis has never been
confirmed or denied. But I certainly have never gotten any alternative
explanation of why this track was there. It has since been removed.
Mr. Nuechterlein: So tell me about your first sitting. What did it feel like being a law
professor for all those years to suddenly be a judge hearing cases? Do you
remember the cases?
Judge Williams: Let me see — I think one of those days, maybe all the days, I sat with
Larry Silberman, along with a district judge sitting by designation, an
Eisenhower appointee from the district of Montana. As you could guess,
he was quite old. There were a torts case and a Speedy Trial Act case;
those are the ones that are coming to mind.
Mr. Nuechterlein: Did you feel at all shy about asking questions?
Judge Williams: No. I think it is possible that the court [as a whole] has gotten more active
in asking questions since then. But no, I didn’t feel shy. Dick Posner has
observed somewhere that the transition from professor to appellate judge
is pretty easy. You are doing something very similar: you’re trying to
figure out what the rule means, what its application would be in different
circumstances, and what the arguments are for interpreting a case this way
and interpreting it that way. That is what professors and appellate judges
do. Opinion-writing is somewhat different.
Mr. Nuechterlein: Do you remember your first opinion?
Judge Williams: The first one to come out was in a labor law case. I remember feeling not
very comfortable with it because I had never had anything to do with labor
law at all.
I also remember that Judge [Ruth] Ginsburg sent me a nice note
congratulating me on my first opinion. So that was a heartwarming thing.
Incidentally, the Ginsburgs had us to supper that fall. Which was very
nice. That was my only exposure to Marty Ginsburg’s cooking, which was
fantastic. His charm was fantastic too.
Mr. Nuechterlein: So you had a full complement of clerks by the time court stared?
Judge Williams: Well, a full complement — the standard package then was three clerks and
two secretaries. It seemed to me at the outset that having two secretaries
was strange. And three clerks were plenty.
13 David Bazelon served on the D.C. Circuit from 1949 to 1985 and was Chief Judge from 1962 to 1978. Warren Burger
served on the D.C. Circuit from 1956 until his elevation to Chief Justice of the United States in 1969.
Mr. Nuechterlein: Presumably the two secretaries were a vestige from the days before
Judge Williams: Exactly. I may have been the first judge to have a regular desktop
computer with suitable word-processing software in it. In those early
years, you could see signs in the way [opinions were revised] that the
secretaries were clearly trying to avoid the horrors of trying to retype the
whole 70 pages. I think it’s fair to say that the court’s opinions have
somewhat shrunk in length since that era. Probably a healthy thing.
Mr. Nuechterlein: That’s interesting. If I remember correctly, by the time I ended up clerking
for you in 1990, you were using what was then considered an
unconventional word-processing program. It was Microsoft Word, which
no one used. I had to learn it [from scratch] because the industry standard
Judge Williams: Yes, exactly. And the federal judiciary, except for a few crazies like me,
pretty much stuck with WordPerfect.
Mr. Nuechterlein: Why were you using Microsoft Word?
Judge Williams: Because I was using it in Colorado.
Mr. Nuechterlein: You weren’t using it to strike out against the monopolist WordPerfect?
Judge Williams: No. [Laughter.]
Mr. Nuechterlein: So you had three clerks. Did you give any thought to how to use clerks
now that you finally had them?
Judge Williams: I intuitively was not drawn to the idea of bench memos.
Mr. Nuechterlein: Why is that?
Judge Williams: There never seemed much point in it. Surely I was going to have to read
the briefs. So if a clerk did a bench memo, that would mean more for me
to read. And it was a distinct likelihood that a lot of what I [would have] to
read [in the bench memos] would not be of great interest to me. After
looking at all three briefs in a standard case, I would have reached a pretty
firm conclusion on some of the issues. And on the issues I found
troublesome and needed more to be done on, there was no guarantee that
the clerk would necessarily have shared that view and would have gone
out and done the extra research [on those issues]. So [bench-memowriting]
just seemed to me a process that would delay things.
Mr. Nuechterlein: If I am remembering this correctly, instead of bench memos, you would
have meetings. Everyone would read the brief, and then you would get us
in a room and ask questions.
Judge Williams: Yes — that was roughly it. There has been a change since your era.
During my first time in the hospital for hip surgery, I knew I was going to
be convalescing for a period running up pretty close to the next sitting. So
I thought that to coordinate with the clerks, I should get all the reading of
the briefs done before the surgery and would leave the clerks notes giving
my initial take on all these cases. So I did that. That seemed to be a very
good idea. Since then I have usually done that [for all cases that haven’t
been] no-brainers. This approach starts a conversation with a clerk — and
sometimes it ends the conversation, or comes very close to ending it,
because, on the issue where I have indicated a leaning but noted some
doubts, the clerk comes back with something that resolves the doubts.
Mr. Nuechterlein: I am trying to remember if all three clerks read all the briefs in all the
cases for a sitting and then met with you collectively.
Judge Williams: It was more the clerk on the case and me. That is certainly the way it is
now. I don’t recall any shift. Sometimes, in a hot case, all the clerks get
very interested and start pursuing it.
Mr. Nuechterlein: You mentioned when we spoke informally before that Richard Posner had
offered advice on how to use clerks and how not to use clerks.
Judge Williams: He had said that you should write the opinion from scratch yourself and
not have a clerk draft it. In some cases I have done that, but I have to say
that they have been rare. The extent to which I change a clerk’s [draft]
opinion obviously varies a lot. What usually happens is this. I read over a
clerk’s draft and think, “that is pretty good.” And then I start all over again
on it, and I see some little thing that bothers me, and it snowballs, and I
see more and more things that bother me. So typically I think the draft
changes quite a lot. You might be a bad example in this because your
drafts were very good. It varies hugely depending, unsurprisingly, on the
draft that I get. I certainly like to think that, by the end of the process, the
opinion has the stamp of my thinking, that it reflects entirely what I think.
The litigants and the taxpayers deserve that.
Mr. Nuechterlein: Judge Posner is one of the few judges that doesn’t rely on his clerks to
Judge Williams: He and Easterbrook are the same.
Mr. Nuechterlein: And I think Justice Stevens. Give me your rationale for letting the clerks
do the drafts.
Judge Williams: The rudimentary thing is this. You come away from oral argument with a
vision of what has to be addressed, but that vision may be incomplete. So
having someone else’s take on it usefully highlights issues that ought to be
addressed and, to some extent, facts that ought to be included. The vision
one person comes away with is incomplete and in some cases inaccurate.
Having a draft [from a clerk] is certainly useful [to address that concern].
I know that Posner is very scornful of the following proposition, but I
think it is also useful to have a clerk place some raw facts out there — on
paper. Otherwise, I might have to spend quite a lot of time getting the facts
from the record to formulate them exactly. The same is true to some extent
for cases that ought to be cited. Typically I start with a clerk’s draft,
[which] has most of the cases that will end up in the final opinion and
most of the facts. [This process] seems timesaving.
Mr. Nuechterlein: I should add that, from the perspective of the clerk, it makes all the
difference if you feel like you are part of the process of writing the opinion
and not just cite checking or critiquing it. That makes the job that much
Judge Williams: I would think so.
Mr. Nuechterlein: I often think that one of the things you love the most about your job is the
opportunity to have what are essentially small seminars, with students of
your choosing, to talk about any number of issues under the sun.
Judge Williams: Yes — I think that is one of the joys of it. One little phrase of yours is
very important there — “students of your own choosing.”
That reminds me of a conversation I had with a student at the University
of Colorado in my last year there. I knew that my nomination was under
consideration in the Executive Branch, and I remember talking with a
student who had very firm thoughts about what should be and should not
be in an opinion. I was thinking, no, you are wrong — at least in the case
of my opinions, [the things you would omit] will be there. It involved the
issue of addressing possible vulnerabilities of the opinion — [whether to]
state them clearly and then try to refute them.
Mr. Nuechterlein: The student thought it would be better to write something closer to a brief,
where you hope that the reader doesn’t notice the arguments on the other
Judge Williams: Right.
Mr. Nuechterlein: Tell me a little about the culture of the court. Your first year was 1986–87.
Justice Scalia has just been elevated to the Supreme Court.
Judge Williams: Yes, he was around in the summer, but he was terribly busy on his
Mr. Nuechterlein: Justice Rehnquist had just been made Chief Justice?
Judge Williams: The clerks that year and I went up and heard part of the Senate debate on
that, I think.
Mr. Nuechterlein: As I recall, that was ideologically contentious. Tell me the extent to
which, during all your years on the court, the court has been affected or
not affected by partisan ideological divides in DC.
Judge Williams: I think certainly in the early years it was quite heavily affected by them.
There was some sense of entering warring camps. Actually, Harry
Edwards told me that, when he first arrived, Bazelon said to him, “Well,
you will be voting with us!” And Harry said, “When I agree with you!”
I think there was certainly some feeling like that. [During that period,]
Wald, Mikva, and Edwards sat together on a panel that, in one day,
produced two en bancs even though they heard only three cases that day.
That is a pretty striking ratio. And you can play it either way. It doesn’t
make any difference which team was ahead on the panel — just that there
was a deep split.
Mr. Nuechterlein: What were some of the issues in those days that were particularly divisive
along ideological lines?
Judge Williams: Well, certainly the whole question of affirmative action or racial
preferences, however you wish to frame that issue. I’m trying to think if
criminal law was divisive — maybe a little bit.
Mr. Nuechterlein: When I think of criminal law in the D.C. Circuit, I think of the old-style
disputes between Burger and Bazelon.
Judge Williams: Right — it wasn’t as extreme as that. I do recall some cases where it
seemed to me that there was a bending over backwards to generate a
retrial when there was no need for a retrial. In the beginning we didn’t
have sentencing appeals. I haven’t seen any numbers on this, but I would
say that the principal issue in half of our criminal cases is the sentencing,
given the complexities of the Sentencing Guidelines. And we are not out
of that [now even though,] since Booker,14 the Guidelines have been
optional or at least not mandatory, because they are still the first step in
calculating what a Guidelines sentence would be and then analyzing the
sentencing judge’s explanation of why he deviated from them. …
Mr. Nuechterlein: One set of issues you have not mentioned is justiciability — for example,
14 United States v. Booker, 543 U.S. 220 (2005).
Judge Williams: Yes, those, too, certainly were big issues. My reading is that those disputes
declined in significance after those early years because the Supreme Court
became much more open to finding standing than it was in the past. …
Mr. Nuechterlein: To what extent do you think these disputes about standing [in the 1980s
and 1990s] reflected a deep philosophical divide about justiciability issues
themselves as opposed to underlying disagreements about the relief sought
in particular cases?
Judge Williams: … A view that originated in the Progressive Era was that courts should not
get involved in anything that smacked of policy at all. That view was then
picked up by conservatives, largely as a response to the Brennan Court.
The progressives did not want the courts doing the things that progressives
saw the courts doing [during the Progressive Era], and the conservatives in
the Brennan era did not want the courts doing the things they saw those
courts doing. And in both cases, of course, the position was reinforced
with a broader ideological structure, which at least on its face was neutral.
And I thought, at the time [I arrived at the court], that the ideological
structure was sound.
I notice small signs that there may be a non-ideological weakening of that
view. One sign is the Fifth Circuit opinion in the Louisiana casket case,15
in which the court is confronting a Louisiana statute that limits the people
who can sell caskets. Basically, the statute creates two separate rules: one
for out-of-state caskets and one for in-state caskets. The one for out-ofstate
caskets has no particular limit on what the casket-maker must be
doing with the rest of his time. The one on in-state caskets says that the
casket-maker must be an undertaker, essentially providing undertaker
services. And the plaintiff is a group that, as one might say, has been
chosen by God for this purpose. It is a Benedictine monastery that is
following the edict of St. Benedict to make its own living. This one chose
to make its own living by making caskets for sale.
There is a brief discussion [in the opinion] of a completely flimsy
argument [by the state] as to some genuinely public-regarding purpose for
this rule. It is very flimsy because, if out-of-state caskets can be made by
people who are not undertakers, what is the consumer-protection element
in applying that rule to in-state casket makers? That is pretty quickly taken
care of. Then you come down to Louisiana’s flat-out honest defense of the
statute, which says it is simply to afford a benefit to the current incumbent
in-state casket makers by cutting down the competition. And the court
says that is not an acceptable purpose. The court does not go into a long
treatment of the history of [substantive] due process — or as I like to call
it, the rule against partial legislation. But it rejects [the state’s defense].
And it cites a Tenth Circuit case that flat-out says, “this is what
15 Abbey v. Castille, 712 F.3d 215 (5th Cir.), cert. denied, 134 S. Ct. 423 (2013).
legislatures do all the time, and it is perfectly OK constitutionally.” So the
Fifth Circuit very consciously, I assume, framed the issue in terms of a
circuit split. Whether the Supremes will bite, who knows?
Mr. Nuechterlein: So where do your sympathies lie?
Judge Williams: My sympathies lie entirely with the casket-makers.
Mr. Nuechterlein: The competitive casket-makers.
Judge Williams: Yes, the Benedictine Monks.
Mr. Nuechterlein: And what would be the constitutional basis for invalidating that state law?
Judge Williams: Well, that is obviously the weak point. The very fact that you have a
doctrine that was foolishly called “substantive due process” suggests
problems, because process is process and not substance, right? So that is a
problem. Even if you apply the more accurate name, “rule against partial
legislation,” you can fairly ask, “where does that rule come from?”
Mr. Nuechterlein: If you were a judge in that case, what tools would you seriously consider?
Judge Williams: I would do a lot of research.
Mr. Nuechterlein: Perhaps the rational basis test under the equal protection clause?
Judge Williams: The rational basis test can be rational basis with teeth, instead of toothless.
And we know that there are realms where rational basis does have teeth,
and maybe this should be one of them.
For a long time, I was persuaded that the strongest objection [to this
judicial role] was not really a doctrinal one but a practical one. And that is
the question: doesn’t this involve judges too much in policymaking?
Richard Epstein called my attention to the fact that we do have a
methodology in the negative commerce clause [context] that seems to
have worked without causing outrage — except perhaps on the part of
Justices Scalia and Thomas, and even then in a muted way.
Mr. Nuechterlein: On the other hand, the judicial rulings on the commerce clause can always
be overcome by Congress.
Judge Williams: That is true.
Mr. Nuechterlein: So there is a political check.
Judge Williams: Yes, that is fair enough. But have any [those rulings] been overturned by
Congress? I don’t think it’s common.
Mr. Nuechterlein: I have no idea. This discussion reminds me that, back in the ’80s, the
rallying cry of the conservative movement was “judicial restraint,” and the
liberals made fun of the doctrine of judicial restraint. It seems to me that
the battle lines have shifted since then. When you joined the court, did you
feel that “judicial restraint” described your [operative approach]?
Judge Williams: Yes, I did.
Mr. Nuechterlein: Although in this casket case, there is no readily available doctrine [for
countervailing the will of the people to suppress the ability of] the
Benedictine monks to manufacture their competing caskets. A good
majoritarian would say, well, that’s legal.
Judge Williams: Definitely an area where [my views have] changed. I guess, living in
Washington, you see more of the [crony] capitalism and all the forms of
anticompetitive manipulation that go on, and increasingly it seems like a
Mr. Nuechterlein: So are there other areas besides fostering competition where you think
courts are rightly vigilant against the majoritarian process, apart from the
obvious cases like restrictions on free speech?
Judge Williams: The big area is one kind of anticompetitive legislative activity or another. I
emphasize that area because a lot of the anticompetitive activities are not
in the form of restrictions on competition; they’re in the form of assistance
to a particular competitive group. Those are peculiarly unreachable [by
courts] because of our standing rules, particularly Massachusetts v. Mellon
and Frothingham v. Mellon.16
Mr. Nuechterlein: Those are the taxpayer standing cases?
Judge Williams: Right.
Mr. Nuechterlein: So you are talking about ostensibly pro-competitive initiatives that involve
giving money to competitors, as opposed to legislation that gives
competitors, say, access rights to the assets of incumbent market
participants. The latter obviously are subject to judicial review.
Judge Williams: Right. Of course that is purely statutory. It is a matter of interpreting the
Mr. Nuechterlein: Tell me a little bit about how the ideological divide in the court mellowed
over the years.
16 262 U.S. 447 (1923).
Judge Williams: I think the biggest single thing was the Microsoft case (which only
involved seven of us).17 Here was a very high-profile case; it is silly that it
appeared to the public in ideological terms, but I think it clearly did. At
least some people who self-identified as conservatives thought that
Microsoft should be left alone, and a lot of the people hostile to Microsoft
seemed to be liberal in some sense of the word. Yet here our court, which
ranged from me as a conservative to David Tatel as a liberal, agreed [on
When it was decided [that we would hear the case] en banc, I thought
there was a prospect that we would agree, and I thought that that prospect
would be enhanced if Tatel and I worked together in preparation for it. So
we did. We had a number of meetings, and our clerks worked very closely
together. And I think that did produce a shared basis of understanding. I
have always gotten along well with him, but that [experience] really
deepened our relationship a lot. [I don’t know if] there’s any colleague I
find it more easy to deal with. That does not mean that I can always bring
him around where he ought to be brought around, but I do feel he listens to
Mr. Nuechterlein: Do you share my sense that [the ideological divisions on the Court may
have] mellowed even before the Microsoft case?
Judge Williams: I think they did. The date that Harry Edwards became chief judge18 was a
big date because he clearly was setting out to mend fissures. I think
everyone recognized that and was ready to credit it.
Mr. Nuechterlein: How did he manifest that?
Judge Williams: Some of it was superficial. He organized an annual court dinner. He gave
people birthday presents on their birthdays. He conveyed very strongly the
sense that we were individually a reasonably distinguished court, but
maybe not as distinguished as a collective entity as we could be if we tried
to be more collegial and more respectful of each other’s opinions.
Mr. Nuechterlein: How do you think the public perception of the D.C. Circuit has changed?
Judge Williams: Most people have not heard of the D.C. Circuit; you should start with that.
When you say “the public,” you mean some very narrow set of the public.
Mr. Nuechterlein: Yes — so, the elites that pay attention to what courts do, how do you think
their perception of the court has changed in the last 25 years?
17 United States v. Microsoft, 253 F.3d 34 (D.C. Cir. 2001) (en banc).
18 Judge Edwards served as chief judge from September 1994 to July 2001.
Judge Williams: I think there is still a tendency to pigeon-hole the court. The history of
pigeon-holing that I see is: “D.C. Circuit wildly liberal,” “D.C. Circuit
violently divided,” “D.C. Circuit staunchly conservative.” And all of those
are oversimplifications, although they have an appearance of truth. … If
you look at all the cases where there is a dissent, it is extremely common,
probably a majority of cases, that the majority is composed of an
appointee of a Republican president and an appointee of a Democratic
president. And so it makes you think this is a strange “ideological”
division if it is an ideological division. And I think you can look at the
merits of the cases and see something more complicated than a simple
“ideological” division—at least in terms of the politics of the Senate or the
House or something like that.
Mr. Nuechterlein: The D.C. Circuit does seem much less ideologically divided than it was
before and also less divided than the current Supreme Court. There might
be a number of reasons for that. One is that the cases that you get are not
as often prone to strong ideological divisions as the cases that the Supreme
Judge Williams: Absolutely.
Mr. Nuechterlein: Another reason may be that the Republicans on the D.C. Circuit have
more of a libertarian streak than their counterparts on the Supreme Court.
Judge Williams: I think that is true. Of course, there is another big thing: everyone on the
D.C. Circuit knows that if an issue is highly ideologically divisive and of
any significance at all, we are not likely the people who are going to be
deciding it in the end. It is going to be decided by the Supreme Court, so
the stakes are much lower for us.
Mr. Nuechterlein: I want to close this discussion of the court’s culture by talking about your
opportunity, which I take it you declined in 2001, to become chief judge.
Give us a little background on your eligibility and why you decided to
pass that up.
Judge Williams: The rule is that you can be chief judge for a term of seven years, but you
can’t be chief judge beyond the age of 70. [And you can only start being
chief judge, basically, if you would have five years in that position before
turning 70.] When Harry’s chief judgeship was coming to an end, it was
clear just by looking at the calendar that I would fit in that window by a
matter of weeks.
I had seen chief judges over the years. I won’t say that any of them were
completely unhappy in the post, but I could see that they had quite a bit of
administrative work to do, which didn’t particularly appeal to me. And
then in the spring, before this critical moment, which had to be September
of 2001, Harry gave me a call and said, “Steve — I have this meeting two
weeks away, and I think that since you are going to be chief judge, you
should be in on it, or at least know about it and come if you’d like.” It was
a meeting in which the people in charge of the buildings on C Street
between 3rd and 6th Streets — I think Harry used the term “the
stakeholders” — were to meet to talk about the parking and congestion
problem. [That problem] was terrible then and is still terrible, because the
people using the Municipal Building do not want to pay for a parking
garage, and they do not want to have someone else come with them who
could drive around the block while they do their business. And therefore
they double-park, and there is quite a lot of traffic in addition, so the
traffic is often tremendously congested. So I was grateful to Harry for
alerting me to this, but I was very confident that I did not want to spend
much time working on that problem.
Mr. Nuechterlein: Do you think he was doing you a favor by showing you a day in the life of
a chief judge?
Judge Williams: It certainly made sense that, if I was going to assume these responsibilities
in September, I should get in on the background, and this would be a way,
at relatively low cost, of getting in on the background with respect to
parking on C Street. But I don’t think it was a plot to shift the chief
judgeship to Ginsburg.19
Mr. Nuechterlein: Do Chief Judges get paid more?
Judge Williams: No — I don’t think so — if they do, it would be $5000. Their secretaries
get more — Mary wanted me to be chief judge. One thing that has
reminded me about this episode time and again is discussions with
Russian judges. In Russia, being chief judge is a very, very big deal. They
have a control over the other judges that is completely foreign to us.
Mr. Nuechterlein: What does that mean?
Judge Williams: Shift them around from one place to another, disciplinary activities. The
Russian judiciary is like a civil service, and if you are boss of a
department, you have a lot of power that goes with that. And I think that is
the way in which the government controls the outcome of the cases it
cares about. Sometimes they have a more traditional “telephone justice”
line straight to the judge, but I think in a lot of cases they do it through the
Mr. Nuechterlein: Before we conclude this discussion of life on the court, I do want to ask
you about your secretaries. Mary [Matera] has been with you since the
19 Douglas Ginsburg was chief judge from 2001 to 2008.
Judge Williams: Right — she came from Pat Wald.
Mr. Nuechterlein: So that has been what — 25 years?
Judge Williams: 27 years.
Mr. Nuechterlein: Lindy was your secretary for a long time.
Judge Williams: Yes. The chief judgeship was also combined with the option to take senior
status, and senior status had some real appeal. At that point, assuming that
I was going to do a three-quarters load — which, from an early stage, I
contemplated — I couldn’t have three clerks and two secretaries. … Lindy
understood this, and so she started looking around. Before the actual
critical moment when I did take senior status, she saw it coming and
sensibly made arrangements. I am not sure what her first job was after
that, but pretty soon she got a job with a judge at the Tax Court. And that
has worked out very well for her.
Mr. Nuechterlein: I want to ask you about taking senior status. So the rule is that you have an
option to take three-quarters of your cases, or …
Judge Williams: You have an option in principle to take anywhere between 0 and 100.
Mr. Nuechterlein: How does your compensation change?
Judge Williams: Your compensation is completely unaffected by it. What is affected is
staff, and at three-quarters, you are entitled to three clerks and a secretary.
Mr. Nuechterlein: I guess it would be meaningless to take senior status and keep your full
complement of cases?
Judge Williams: Actually, Judge Sentelle is doing that. He says that the reason he is doing
it is that he has had two additional jobs besides being a judge in the D.C.
Circuit for the last several years. One was being Chairman of the Judicial
Conference, and the other was the Chief Judgeship itself.20 With those two
going away, with his taking senior status, he is like on vacation.
Mr. Nuechterlein: What does it mean if you are a circuit judge that takes senior status but
does not have your case load diminished?
Judge Williams: How does your life change?
Mr. Nuechterlein: Yes.
Judge Williams: At least on the D.C. Circuit, your ranking changes. You immediately
become junior. One thing I had not realized until I did it was that you can
20 David Sentelle was chief judge from 2008 to 2013.
become very junior. If the judges who previously took senior status were
appointed before you, they have seniority over you. So when I took senior
status, I think it was only Harry [Edwards] and Larry [Silberman] who
were already senior at that point, and they had seniority over me, just as
they did when I started in 1986.
Mr. Nuechterlein: What is the significance of seniority other than who gets to preside over
Judge Williams: [Opinion] assignment is really the only significance. There are cases that I
would have assigned to myself but that were not assigned to me. And vice
versa. Not a huge number. And people know that I enjoy not merely the
FERC cases, but also the other regulatory cases, and so they think rightly
that everyone will be happier if I do quite a lot of those.
Mr. Nuechterlein: Let’s begin talking about the substance of administrative law, which is
probably the most important substantive area of law at the D.C. Circuit
and certainly the area the court is best known for. At a very high level of
generality, I think of the principles of administrative law [as boiling down
to a core objective]: keeping the Executive Branch honest. Tell us how
you conceptualize the function of a court of appeals, and the D.C. Circuit
in particular, in scrutinizing the decisions of regulatory agencies.
Judge Williams: Well, your summary, at a high level of generality, is the essence of it.
What follows from that is making sure that the Executive Branch is
complying with the law, assuming it is a case where the plaintiff has
One might think that [this goal — ensuring that an agency complies with
the law —] does not justify a court of appeals in being bothered at all with
the facts. [Under that line of thinking,] the court of appeals could just
accept the facts as found by the agency. But when you think about that for
a minute, that doesn’t make complete sense because it would mean that an
agency that really wanted to get around the law could just find facts that
justify its outcome, and at that point the review on the law would become
meaningless. … But a court does have some duty to make sure that it is
not getting into decisions that are properly the agency’s.
Incidentally, once when I was in Russia, I had a lengthy discussion with
interpreters on the meaning of the word “deference,” and I found it
impossible to get across what was involved. I think it is partly because we
use “deference” in two different ways. One of these — I think the proper
[use of the term] — describes the lens through which we examine what the
agency has been doing. The other is a statement of the conclusion: “we
defer to this,” meaning “you win.” Although I am not sure if my opinions
have always rigorously done this, the word “defer” should not be in that
Mr. Nuechterlein: We have a lot of verbal locutions to describe the proper role of a court in
scrutinizing what agencies do. And deference in the more nuanced sense
you mentioned — the lens through which you view agency decisions —
obviously plays an enormous role in the life of the court. As a practical
matter, it requires a lot of subjective judgment to know when you have
gone far enough in second-guessing what an agency has done. … When
you analyze APA cases, how do you draw that line? Not you in particular
— the court in general. What is your sense of when the court is more or
less likely to second-guess a decision of an agency?
Judge Williams: Are we beyond the area where the language of the statute seems to require
or forbid [a particular outcome?]
Mr. Nuechterlein: I am talking more about the State Farm type of case,21 [where the question
is whether the agency has reasonably analyzed the record.]
Judge Williams: It seems to me there is inevitable subjectivity there. If you think about the
area being regulated with one outlook, then omissions or contradictions in
the agency opinion will look large or small depending upon how your
brain automatically fills in the blanks. … Harry [Edwards] has written
quite a few articles about this. One of his most intuitively correct findings
is that it is on those issues, the State Farm issues, that you see the biggest
divisions. In at least one of those articles, he pointed out that these are
problems for the agency, which the agency can solve, if there is a good
answer to the basis of the court’s remand.
Mr. Nuechterlein: The D.C. Circuit has a reputation for being more scrutinizing of agency
decisions than other federal courts of appeals.
Judge Williams: Yes. But I have heard people say that other courts sort of fluctuate more
between being totally submissive to the agency and just acting as if they
were the agency, while we fret and worry a great deal trying to avoid those
extremes. I can neither confirm nor deny this statement about other
Mr. Nuechterlein: There is this sense that courts are more likely to second guess an agency if
they understand the subject matter and find it interesting.
Judge Williams: Yes.
Mr. Nuechterlein: Often when you have complicated ad law cases in regional courts of
appeals, you get a panel with no prior exposure to the area of regulation,
and it finds the case mystifying, so the path of least resistance is just to
affirm. In the D.C. Circuit, you have judges who have a deep background
in certain substantive areas, and in some cases they were picked for the
21 Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Insur. Co., 463 U.S. 29 (1983).
court for that reason — because they are comfortable with analyzing the
regulatory state. You are thus more likely to get [second-guessing] and
less likely to get crazy remedies.
Judge Williams: Yes — I think both of those are true. Actually, the Federal Energy Bar
Association very kindly gave me an award — it is on the windowsill in my
chambers — and the price of getting the award was giving a little talk.
And I made the argument that a judge who knows something about an area
has sort of a plus and a minus for the parties involved. Exactly this: they
more likely understand what is going on, but on the other hand there is
some risk that they will be overconfident about their grasp of what is
going on in the area and will therefore not leave to the agency what ought
to be left to the agency. I think those are genuine problems.
Mr. Nuechterlein: Which is a reason for humility in the remedies: instead of saying “you
may not do this,” you say “have another look.”
Judge Williams: I think I may have mentioned to you that sometime, quite a while ago, I
asked a clerk to do a check of D.C. Circuit opinions to survey the
frequency of remand for want of reasoned decision-making. Somewhat as
I expected, I seem to be, in that particular sweepstakes, the winner.
Mr. Nuechterlein: We have been talking about the State Farm set of issues. There is also the
Chevron set of issues.22 I have heard several different justifications for the
practice of deferring to agency interpretations of ambiguous statutes.
Some of these rely on notions that Congress kind of intended for the
Executive Branch to be the one to resolve interpretive ambiguities in a
statute that has been committed to an agency.
Judge Williams: My guess is, as on so many issues that Congress addresses without a great
deal of focus, that [rationale] is largely a fiction to justify a conclusion
arrived at upon other grounds.
Mr. Nuechterlein: So what do you think the [appropriate] rationale is?
Judge Williams: I think it is the basic idea that there is a lot of policy involved in coming to
an interpretation of a statute, and that policy is properly for the agency, as
long as Congress doesn’t resolve the policy issue one way or the other.
Incidentally, coupled with that is the strong belief that one generation of
regulators, one cohort of regulators interpreting an ambiguous statute,
should not bind the others for all times, or at least until Congress manages
to get around to correcting the statute.
Mr. Nuechterlein: So what I am hearing is a sort of highly utilitarian explanation: “We
assume that Congress probably had some policy objective in mind. Who
22 Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).
better than the agency — which spends its whole day thinking about these
issues — to resolve the ambiguities Congress leaves in the statute? That
approach is more likely to produce a result that comports with public
policy than if generalist judges play that role.”
Judge Williams: I am not 100% sure about [whether the outcome would be] better or
worse. I guess [the key is the] flexibility of the agency, the ability to put
the learning of experience to work. If a court interprets a statute, and says
it means “this,” then it is very rare that [a litigant] will be able to come
back ten years later and say, “Look, history has shown that that has
produced terrible results.” The court is unlikely to dig its way out of the
hole it has put everybody in.
Mr. Nuechterlein: We operate as lawyers and judges in a world where we speak of statutes as
having “a meaning,” and it is left to the agency or a court to determine in
the first instance what that meaning is. Of course, that can’t possibly be
the right model for understanding statutes if we say that it is OK for the
agency to reverse course on what it thinks a statute means. So [courts do
not defer to agencies on the theory that agencies are somehow uniquely
qualified] to perform the epistemic function of determining what the
“true” underlying meaning of a statute is. I think we are just saying that
courts should [defer to agencies because, through gap-filling, agencies
rightly] play a role in legislating.
Judge Williams: I have always liked the metaphor used by Cardozo in his separate opinion
in Panama Refining.23 It makes sense to look at what Congress does as
creating a canal that has borders and has a body in between, and as long as
the agency is on the water, within the boundaries of the canal, it is OK. He
was using the metaphor for a different purpose, of course, but it still
Mr. Nuechterlein: Let’s go back to reasoned decision-making. When you remand for
reasoned decision-making, the instruction is essentially this: “you, agency,
may well have the authority to do this thing….”
Judge Williams: On the other hand, you might not.
Mr. Nuechterlein: Might not but you may well. We are not going to say that you don’t, but
we do want you to explain yourself. That is part of keeping an agency
I was just wondering: I have worked at federal agencies that may
sometimes, under some leaderships, view that sort of remand as essentially
a victory. If the remand is not accompanied by a vacatur remedy, the
agency can take a few more years before it has to cook up its next
23 Panama Refining Co. v. Ryan, 293 U.S. 388 (1935).
rationale for the thing it wanted to do, and its political decision is in full
force in the meantime. Perhaps a remand for reasoned decision-making is
highly valuable in an idealized political system, where agencies care about
meeting their responsibilities. Do you have any views on whether the
actual political system we live in works like that?
Judge Williams: That is very interesting. What you are suggesting is that the key variable
with a remand for reasoned decision-making is the application of vacatur.
… In deciding whether to vacate, we expressly apply the sort of balancing
test that we definitely, expressly apply when it’s an application for a stay.
Mr. Nuechterlein: Yes, you look at the likelihood that the agency will be able to reach the
same outcome on remand, after further deliberation, and you also look at
the disruptive consequences of vacating a rule that may later be reinstated.
So if you do vacate when remanding for reasoned decision-making, you’re
telegraphing a very strong sense that the agency will probably not be able
to reach the desired policy. Here, though, I’m referring to the other class
of cases where you are throwing up your hands and saying, “Agency, it’s
up to you guys. You are the expert agency, but you have not produced
something we can defer to yet, and you need to explain yourselves, but we
are not at the point where we believe you can’t [justify the same
outcome].” My question is, in those cases what do you think the practical
effect of a remand is?
Judge Williams: [The question is] how does such a remand actually work out? You
probably recall the [FCC] case involving the question of ISP reciprocal
compensation. There we did not vacate. …
Mr. Nuechterlein: You did not vacate. [You remanded twice but] didn’t vacate either the first
or second time. I think you wrote both of those opinions.24
Judge Williams: Then there was [a later] opinion by Merrick Garland.
Mr. Nuechterlein: That was the mandamus proceeding. Your second decision was in ,
and the Core mandamus opinion was in either 2007 or 2008.25 There was
some strong language in that opinion.
Judge Williams: That’s interesting. My sense in the first opinion was that the case was
really, really complicated, and we did not want to screw things up. And
the Commission in the next round came up with a justification that really
didn’t work. I and the panel were unanimously convinced that it didn’t
24 WorldCom, Inc. v. FCC, 288 F.3d 429 (D.C. Cir. 2002); Bell Atl. Tel. Cos. v. FCC, 206 F.3d 1 (D.C. Cir. 2000).
25 In re Core Comm., Inc., 531 F.3d 849 (D.C. Cir. 2008). The court directed the Commission, which had not responded to
the second remand for a half-dozen years, to come up with an acceptable rationale within six months or face automatic
work. [That suggested that the issue] was not only complicated in terms of
the regulatory economics and so forth, but also complicated legally, or
they would not have stumbled onto a solution such as the one they did. …
[That it took the Commission six years more to come up with a new
rationale] may suggest you are right — that [the remand without vacatur
penalty] doesn’t send a strong signal. On the other hand, taking the history
as a whole, I am not convinced that it would have been good to throw the
system into potential chaos.
Mr. Nuechterlein: What do you think is the main contribution to our regulatory system of
having this light-touch [remand-without-vacatur] remedy?
Judge Williams: I am enough of an optimist to think that [our demand for a] clearer
expression in an agency’s justifications has some sort of positive effect on
actual thinking. I understand there are some big gaps in that proposition.
Mr. Nuechterlein: So, over the long term, the fact that there is a court willing to hold an
agency accountable for [its failures of reasoned decision-making] …
Judge Williams: Yes — it has a healthy effect. And how big that effect is, I certainly
wouldn’t try to estimate.
Mr. Nuechterlein: Do you have a sense that remand-without-vacatur is more common now
than it was when you took office?
Judge Williams: That’s a good question. I don’t know. I honestly don’t know. I suppose
that Chevron might make it more common—reversals in the statutory
interpretation process at least theoretically would decline and therefore
leave more room for spotting [other] troubles in agency decisions. There
is also a paucity of empirical data on what Chevron’s effects have really
The most famous [study was by Peter Schuck and Donald Elliott, entitled]
To the Chevron Station. There have also been some more recent studies.
Tom Merrill has done a piece. [Bill] Eskridge, too, has a very recent
article. He accumulates the total volume of Supreme Court decisions
where Chevron was, in principle, in play — in the sense that the Court, if
it were dedicated to Chevron, would have necessarily gone through a
certain mental process in arriving at its conclusion. I think that, in a
majority of those cases, Chevron does not appear. Period.
Mr. Nuechterlein: In terms of the remand-without-vacatur remedy, there is some
disagreement on the D.C., Circuit about whether that is even permissible
under the APA. My first question is this: I don’t recall seeing the court
actually do an analysis of the words of [5 U.S.C.] section 706 to determine
whether it is actually permissible.
Judge Williams: I think that is true. The prime exponent of the view that the words require
vacatur is Judge Randolph. Certainly in the long period between his initial
opinion and his taking senior status, he has had plenty of opportunities to
call for an en banc on the issue. That would not have necessarily produced
an en banc, but it might have produced a whole set of opinions concurring
with and dissenting from the denial of en banc. I don’t believe that
happened. So I guess the unspoken understanding is that the words are not
all that clear.
Mr. Nuechterlein: So Judge Randolph’s view would in theory, though probably not in reality,
divest the D.C. Circuit of the humble remedy of sending a case back to an
agency to think about it some more without disrupting the status quo. But
in reality, as he has pointed out, and as have you, that the difference
[between his view and yours] does not amount to very much. That’s
because, [even when the court issues a vacatur order,] the agency can just
come and say, “don’t issue your mandate.”
Judge Williams: Yes.
Mr. Nuechterlein: Then you’d have presumably the same balancing of stay factors [that the
court already uses, under your approach, to decide whether to remand
without vacating]. So why does anyone bother with this issue?
Judge Williams: Well, I think the answer is that nobody bothers about it very much — Ray
is bothered about it a bit — but as I said not all that much.
Mr. Nuechterlein: Do you have a sense of whether certain agencies come to the D.C. Circuit
with a greater presumption of reasoned decision-making than others? Are
there certain types of agencies that are more likely to be subject to hardlook
Judge Williams: Do we view a decision from agency X differently than a decision from
agency Y because of a prior record of lesser competence? I honestly think
the answer to that is no. We look at the issues as raised by the petitioner,
we look at what the agency had to say, and we decide the case on the basis
of those things. So I don’t think an agency can get into an automatic losing
streak, or an automatic tendency to lose, on the basis of its record.
Actually, it would be sort of reprehensible if we did [consign agencies to
different categories on the basis of their past records] because, after all,
agency staffs change, and one hopes there is improvement. It would be
really peculiar to review a well-reasoned decision by a “bad” agency
partially. This is like the problem that investors have in dealing with
mutual funds (at least mutual funds other than index funds). Investors can
sometime [choose a fund] on the basis of its prior record, which of course
may be luck, and which also may [reflect the investment decisions] of
some genius fund manager who has then gone off [to some other job].
[This issue of treating agencies differently reminds me of a question I was
asked when] I was initially being vetted for the Tenth Circuit. It must have
been the FBI. I had not realized that the FBI not only checks to see if you
are smoking dope and doing other terrible things like that, but also seems
to try and answer questions that have more to do with judicial skill or
something. Anyway, the FBI agent talked with a lot of people in what is
considered the relevant community, and apparently some people in the
relevant community in the Tenth Circuit said, “he doesn’t really know the
district judges, and it is very useful to know them in ruling on their
I share your laughter, but mine is mixed with dismay about the attitude
that this seems to represent. And the suggestion that we would have a
fixed view of an agency — a fixed view that was relevant to [reviewing
the agency’s] decisions — falls somewhat in the same category.
Mr. Nuechterlein: Although some agencies are subject to political pressures in ways that
other agencies are not. So even though they may be just as smart as other
agencies, just as capable of producing well-reasoned, lawful opinions,
some agencies may be disproportionately subject to, for example,
congressional oversight pressures that other agencies are not.
Judge Williams: I am not sure if I am well plugged in enough to know which these
agencies would be.
Mr. Nuechterlein: Independent agencies are more subject to congressional pressure than
[Executive Branch agencies. Of course, Executive Branch agencies may
be more subject to White House pressure.] It may be just different types of
Judge Williams: Yes — that would be my guess.
Mr. Nuechterlein: After you have remanded a case to an agency for the first time, is there
any presumption either way about the soundness of what it has done on
Judge Williams: That reminds me of when I was first teaching administrative law. In
Chenery,26 you will recall, the Supreme Court’s first decision does
effectively remand for reasoned decision-making. And then the second
decision says [the agency order on remand] is OK. Justice Jackson,
however, remarks: “the more you explain, the less I understand it.” Of
course, he is talking about his colleagues on the Court, saying that their
26 SEC v. Chenery Corp., 332 U.S. 194 (1947); SEC v. Chenery Corp., 318 U.S. 80 (1943).
opinion doesn’t make much sense. Maybe a change of personnel on the
Court [explains the second] Chenery [decision].
Also, how long can [a court] keep it up? There are two possible ways that
a court can respond [to an agency that stays the course following a remand
without any satisfying explanation]. One is to think [the agency is] utterly
determined about this so it seems fruitless to keep trying. And the other is
[reflected in] an opinion [I wrote]. This was not a case of successive
remands of the same actual decision, but successive critiques of the
agency’s reasoning. The agency said, “we were looking for data that
would support this,” and the prior panel said, “all right — you can’t be
expected to have good data that early.” But then finally there came a time
when [the agency] came back to us, and they had by then years and years
to produce data, and yet they had produced no data. They did produce sort
of theoretical reasoning as to why this was a sensible rule, but the theories
did not seem to hold up logically at all, and so we did say: “No, you can’t
Mr. Nuechterlein: The presumption is that if the agency tried that hard to find some good
reason to produce an outcome, and if they could not come up with it even
on remand, they are probably even less likely to justify that outcome the
third time, and therefore you should vacate?
Judge Williams: That would be the lesson from that case. … If there is a discrete
proposition that appears not to have a basis, and then turns out not to have
a basis, that sounds like a case for drawing the line. …
Mr. Nuechterlein: Do you ever get approached by regulators about your decisions
informally? Do you ever run into someone at a party who says, “You
know, you reversed me like eight years ago, and I just want to tell you that
you misunderstood X and Y.”
Judge Williams: No, I haven’t — that is interesting. I would enjoy that. …
Mr. Nuechterlein: You mentioned discussions with Russian judges before. You’ve spent a lot
of time in Russia and a lot of time thinking about Russian legal reforms,
such as they are. In your discussions with judges from Russia and other
foreign countries, what is their view of the administrative law regime that
we have in the U.S., where agencies are quasi-lawmaking bodies, and
where courts supervise agencies but defer to them? Is there a counterpart
to that in other countries?
Judge Williams: The answer is I can’t be sure, but I think so. We know from Justice
Breyer’s books that the conseil d’état plays sort of a role like this [in
France]. But it’s less a court than a super-agency, so it’s not quite
analogous. In Russia, courts are deeply involved in looking at what
administrative agencies do, and interestingly enough, the complainants
there have, at least numerically, a very successful track record. I think the
[reason] is partly because these are very small-stakes cases, and the
higher-ups in Russia are not involved at all.
[Also, whereas a U.S.] court’s decision has an impact going forward, [the
Russian] system is either [ignores] precedent or views precedent in a very
special and limited way. So the government can lose a lot of socialsecurity-
type cases without any impact. In contrast, in the disability area,
the U.S. courts were highly interventionist and produced a significantly
different system that [because of precedent] has led to an enormous
increase in the number of people on disability. That may be a good thing
or a bad thing, but it certainly had a huge fiscal effect. Very little risk of
that in Russia. And my sense is that the idea of having a [deferential] lens
through which [a court] examines agency decisions is quite alien to
[Russian judges]. Also, they don’t write opinions of the sort that we do.
Although they have something called an “opinion,” it just has a lot of
“whereas” clauses and then it has an outcome.
Mr. Nuechterlein: Is there reasoning in the “whereas” clauses?
Judge Williams: I don’t think so.
Mr. Nuechterlein: That is discouraging. I want to ask you about the choice that we make in
our system to have generalist courts of appeals review agency decisions of
Judge Williams: Yes — exactly.
Mr. Nuechterlein: The D.C. Circuit seems to occupy a middle ground between the regional
courts of appeals on the one hand and the Federal Circuit on the other,
[which specializes in matters within its exclusive appellate jurisdiction].
… What are your thoughts about the extent to which it is useful to have
truly generalist judges reviewing decisions of administrative agencies
versus judges who really have (or acquire) a deep background in the
Judge Williams: It is self-serving, but I am quite comfortable with this intermediate ground
that you have identified. And I have heard the argument — which, it
seems to me, is probably correct — that a court such as the Federal Circuit
is likely to feel that it owns its area and therefore, if there is an area where
there is supposed to be any deference, it gets clouded and torn away. Now
assume that the court and the agency are of absolutely equal quality and
intellectual quality. The problem that I mentioned earlier, about the
interpretation of the law getting locked into a particular viewpoint — is
more likely to occur [if the court supplants the role of an agency, because
agencies can more easily change their positions].
Mr. Nuechterlein: Do you think anyone in Congress ever gives serious thought to when it is
appropriate to have a specialist court of appeals review something and
when it is not?
Judge Williams: Well, they do sometimes. Remember there was kind of jurisprudential
crisis in veteran’s benefits. … The V.A. did not particularly want 600
generalist federal judges getting into their stuff, but the veterans wanted
some kind of supervision of V.A. decisions because they assumed,
probably correctly, that the courts would be tilted pro-veteran, probably
even more than the V.A. The solution was the creation of a Court of
Veterans Appeals. My guess is that it is a pretty special set of
circumstances that makes Congress take a serious look at [consigning
particular topics to specialist courts].
Mr. Nuechterlein: Thanks, Judge — that concludes today’s session. We’ll continue in a few