Friday, October 3, 2008

One Nation Under God?

The Supreme Court's own Oath of Office makes for some Very Interesting Reading, and Who's this "God" fellow, anyway?
by Tom Van Dyke

Well, there's been a big hubbub and some decent discussion around here lately about the Pledge of Allegiance, especially the "under God" part that was added in 1954, apparently American society's response to the growing godless worldwide Communist hegemony. Not an unreasonable fear. The Commies took over a helluva lot of countries. By ideology, and by force.

Me, I could live without the Pledge today in 2008---it seems the squabbling about it diminishes any utility it might have once had. Seems like a cudgel to beat "God" out of our public square, the exact opposite of its original intent.

And it's not good to drag our kids into our adult conversations and controversies, since the kids who opt out of "under God" might get the fisheye from their religious classmates---or vice-versa if they put it back in, depending on how "cosmopolitan" the school is. It puts in a dimension of social coercion and conformity that even most religionists should be uncomfortable with. Takes all the unum out of our pluribus, which is the whole point of the Pledge in the first place.

But in principle, I don't give a damn that some atheist finds hearing "under God" an offense to his sensibilities, or feels "diminished" as a patriotic American when he stays silent during that part. This nation has always made room for conscientious objectors, going back to the debate on the Second Amendment when it was thought the militia clause might obligate Quakers to bear arms, which is against their religion. They don't bear arms to this day, and similarly, it's entirely proper to "opt out" of the "under God" part.

Right now, Michael Newdow enjoys an appellate victory from the Ninth Circuit, somewhat putting the Pledge on ice. But the Ninth Circuit is America's most radical one, the one that's overturned most often at the top level. It's likely the Pledge will end up in front of the Supreme Court somehow someday.

And judging by the decisions in similar cases over the past 50 years, I think the odds are on Newdow's side. But mebbe somebody will ask the members of the Court to look at their own oath of office, written in the first days of the very Founding of our nation in the Judiciary Act of 1789:

"I, A. B., do solemnly swear or affirm, that I will administer justice without respect to persons, and do equal right to the poor and to the rich, and that I will faithfully and impartially discharge and perform all the duties incumbent on me as [blank], according to the best of my abilities and understanding, agreeably to the constitution, and laws of the United States.

So help me God.

Huh? What was that? "So help me God?"

One of my blogbrothers spent a considerable amount of time working for or with Michael Newdow to track down whether George Washington added "So help me God" when he took the office of president. [Yeoman's work here, and proving to near-certainty that Washington did not say it. Well done, Ray.]

But whether Washington informally added "So help me God" is small potatoes compared to Congress formally putting it into a Supreme Court justice's oath, wouldn't you think?

And the previous section 7, requiring the same oath of judicial clerks, allows them to "opt out" of the "God" part---

"Which words, so help me God, shall be omitted in all cases where an affirmation is admitted instead of an oath."

So, there should be no problem here. A norm, a custom and practice is established among the very first acts of the new Congress, the trap door left open for freedom of conscience. No coercion---no harm, no foul.

Opt out. Conscientious objection. Freedom of conscience is preserved.

Perhaps in our era of the "living Constitution," the Justices will someday see fit to rewrite their own oaths while they're consigning "under God" to the ashcan of history.

Because they should rewrite their own oath of office (as written by Congress) and trash the "under God" thing (as written by Congress) at the same time, because they are one in the same.

Then we'll establish once and for all who's in charge of this here country. Not the elected president, and not the elected Congress. The courts. Because if and when they decide in 2009 or 2012 or 2020 that one of the very first acts of Congress in 1789 is unconstitutional, then everything's up for grabs.

Everything, folks, no lie. So help me God.


Ray Soller said...

Tom Van Dyke wrote, "Right now, Michael Newdow enjoys an appellate victory from the Ninth Circuit, somewhat putting the Pledge on ice. But the Ninth Circuit is America's most radical one, the one that's overturned most often at the top level. It's likely the Pledge will end up in front of the Supreme Court somehow someday."

The claim that the 9th Circuit is our nation's "most radical one" is more a matter of perception managed by the mass media than actual fact. In the book, Taking on the Pledge of Allegiance, Ronald Bishop explains, "[A]ccording to [Duke University law professor Edwin] Chermerinsky, reporters, perhaps caught up in the desire to prove that they and their employers are not liberal, mischaracterized the Ninth Circuit's makekup and rulings. Take the claim about the frequency of Supreme Court reversals: during the high court's 2002 term, it reversed decisions from the nation's federal appeals court about thee-fourths (74 per cent) of the time. The justices reversed the Ninth Circuit exactly thee-fourths of the time. But you wouldn't know that by reading coverage of the Newdow ruling -- reporters were too busy casting the appeals panel, like Newdow, as a threat to our values. Judge Goodwin and Judge Reinhardt were portrayed as Newdow's accomplices in his push to bring down God, mom, large gas-guzzling cars, celebrity worship, baseball, and apple pie." (pgs. 51-52)

Ray Soller said...

A synopsis of Ronald Bishop's book, "Taking on the Pledge of Allegiance: The News Media and Michael Newdow's Constitutional Challenge," is available at Reviews Written by Explicit Atheist.

Matt Huisman said...

I'm not sure that the % of cases overturned is the most relevant statistic. Given that SCOTUS gets to select which cases it reviews, the number of cases overturned (or possibly the % of cases granted certiori) would be the better indicator of a circuit court's 'radicalness'.

Brian Tubbs said...

Just for the has NOT been proven that GW didn't say "so help me God." It's practically impossible to prove a negative anyway.

I will grant that Ray and others have cast DOUBT on the traditional assumption that GW said those words. But they have not proven (nor can they) that he didn't say them.

And even if they could, it's more or less irrelevant, because GW uses his Inaugural Address to make abundantly clear his belief in and dependence on God. In fact, he calls on all of America to depend on and "adore" the "Invisible Hand" - i.e., God.

As for as George Washington is concerned, the United States is and should be "one nation under God."

On that, I stand with Washington.

Tom Van Dyke said...

Mr. Huisman properly points out that it would not be the percentage, but the numper of cases that the Ninth Circuit has overruled that is probative. Chermerinsky's argument is (not atypically) sophistic, as I believe the stats hold up.

But this is a side issue. My thesis stands unmolested, then, although I'm surprised that nobody expresses surprise at the fact that the Supreme Court oath itself formally includes "So help me God," a little-discussed fact in all this Pledge and presidential oath furor. I believe it's a fact that cannot be ignored.

Eric Alan Isaacson said...

Thank you, Ray, for pointing out that the Ninth Circuit by no means deserves the reputation that right-wing commentary has assigned to it.

Here’s a link to my friend Erwin Chemerinsky’s excellent article, The Myth of the Liberal Ninth Circuit.

I ran into Erwin in Cincinnati on September 17, where I saw him argue an appeal before the United States Court of Appeals for the Sixth Circuit. I had an argument in the same courtroom the next morning, and had gone over to scope things out. I really wasn’t expecting to see Erwin there. Lucky for me: He was brilliant.

I also was gratified to see that the courtroom featured a large portrait of the Sixth Circuit’s most famous judge, William Howard Taft, as my own argument relied in part on an opinion written by Taft.

And no, Tom, my briefs did not mention that in addition to becoming President and Chief Justice of the United States, Taft also served his denomination as Vice-President of the American Unitarian Association, from 1916 to 1922, and as president of the General Conference of Unitarian and other Christian Churches, from 1915 to 1925 – when it was absorbed by the American Unitarian Association (which itself would merge with the Universalist Church in America in 1961, to become the Unitarian Universalist Association of Congregations). See generally Vernon B. Hampton, Religious Background of the White House 262, 340 (Boston: Christopher Publ. House, 1932) (“Taft gave of himself to his church unstintingly, much more so than any other of our occupants of the highest office in the land.”).

Tom Van Dyke said...

Eric, if you have a counterargument to Mr. Huisman about the number of cases, please post it.

And no, Tom, my briefs did not mention that in addition to becoming President and Chief Justice of the United States, Taft also served his denomination as Vice-President of the American Unitarian Association...

Huh? I know you have an agenda to promote your church, Eric, but a little subtlety, please, man. That one gave me whiplash.

But I guess if there are no substantive objections to my main thesis about "So help me God" being in the Supreme Court oath of office, court is adjourned, and as it was in the time of the Founding, this space may now be used for religious services. [Another little-discussed fact.]

Eric Alan Isaacson said...

Why would I advance a "counterargument" to Mr. Huisman's point that one should consider what percentage of the Ninth Circuit's cases the Supreme Court chooses to review?

With 27 active judges, and 22 senior judges, the Ninth Circuit is the nation's largest circuit, covering the states of California, Arizona, Nevada, Oregon, Washington, Idaho, Montana, Alaska, and Hawaii. It decides more than six thousand appeals a year.

Compare the Ninth Circuit's 27 active judges and 22 senior judges to the First Circuit (covering Massachusetts, Rhode Island, New Hampshire, and Maine) with five active judgesand three senior judges, or to the Second Circuit (covering New York, Connecticut, and Vermont), with 12 active judges and 10 senior judges, and you begin to get an idea of how much larger it is than its sister circuits.

You'd expect the Supreme Court to review (and to reverse) more decisions from the Ninth Circuit than it does from any other court, simply because the Ninth Circuit decides so many more cases than the other circuits do.

The fact is, Supreme Court grants certiorari to review a only tiny fraction of the thousands of decisions that the Ninth Circuit issues each year.

Tom Van Dyke said...

You have data, then, Eric, but none of it is usable. Very well.

But I completely withdraw the remark about the Ninth, anyway. I shouldn't have added it, as it just gave fodder for getting hung up on a side point. In fact, I explicitly wrote that at some point I found it likely Newdow or someone like him would win in the Supreme Court itself.

My core argument stands unmolested and indeed, unaddressed.

Matt Huisman said...

Not to make too big of a deal on this, but here are some statistics from 2003 (9th Circuit/All Appellate):

Appellate cases reviewed - 25/68 (37%)

Appellate cases reversed/vacated - 19/54 (35%)

Appellate cases unanimously reversed - 10/23 (43%)

9th Circuit % of total appellate cases - 19%

There may be reasons for disputing the liberalness of the 9th circuit, but I have not seen statistics (which I'm not crazy about anyway) that are able to make the argument.

Explicit Atheist said...

"But whether Washington informally added "So help me God" is small potatoes compared to Congress formally putting it into a Supreme Court justice's oath, wouldn't you think?"

Yet the Joint Congressional Committee on Inaugural Ceremonies apparantly thinks the "fact" that "everyone" (all presidents) appended that phrase is important enough to justify a "so Help me God" video on its web site featuring a Senate Historical Office staff member asserting this "fact" along with the written assertion "First Inauguration; precedents set include the phrase, "So help me God," and kissing the Bible after taking the oath.". When we point out that there is insufficient evidence to show that GW or any other president to at least the Civil War appended that phrase during their oath of office recitation the Senate Historical Office refuses to respond with any evidence to the contrary. A Senate Rules Committee staffer told me the Senate Historical Office has no obligation to provide the source for their claims because they work for the Senate not the public. So why, if this isn't important, does the U.S. Congress keep publically asserting this as historical fact on their official web site without the proper evidence?

Brad Hart said...

Very interesting post, TVD. I was unaware that the Supreme Court's oath included "So Help Me God." Very interesting indeed!

You also state in your post the following:

"Me, I could live without the Pledge today in 2008---it seems the squabbling about it diminishes any utility it might have once had. Seems like a cudgel to beat "God" out of our public square, the exact opposite of its original intent...

...But in principle, I don't give a damn that some atheist finds hearing "under God" an offense to his sensibilities, or feels "diminished" as a patriotic American when he stays silent during that part."

For the most part, I am in agreement with you here. I see the "So Help Me God," "In God we Trust," "One Nation Under God" arguments as being more divisive than anything. It's like listening to the hosts of "The View" argue over mundane issues that have no real bearing on a person's life. For me, I really could care less if these slogans/phrases were put into law or done away with entirely. I simply don't believe it is all that important to begin with.

With that said, if I was forced to take a side I would probably go with the Christian apologists on this one. After all, those who disagree with these mottos can simply refrain from them. However, removing them completely would deny believers the opportunity to express these sentiments.

Again, I go back to Benjamin Franklin and his "public religion." As was the case for most founders, Franklin believed in a neutral public religion, which was inclusive of all belief systems and creeds. This public religion also stressed the importance of a secular government that would also remain neutral in the religious area. With this in mind, I don't see how "In God We Trust," "One Nation Under God" etc. is a problem. In these cases, "GOD" can be seen as a neutral and all-encompassing being and NOT exclusively the Christian God. If our coinage or our Pledge of Allegiance were to say something like, "In Jesus Christ we Trust," or "One Nation under our Savior Jesus" then I would have a major problem with it.

In short, these slogans/mottos are fine as they are…at least…in my opinion. Good post!

Eric Alan Isaacson said...

Mr. Huisman's link for statistics concludes that in 2003 "9th Circuit rulings constituted about 32 percent of both the cases [reviewed] (25 of 79) and the reversals (19 of 59) the Supreme Court decided by written slip opinions this term."

In other words, the Supreme Court in 2003 reversed Ninth Circuit opinions at about the same rate as the other courts that it reviewed.

Explicit Atheist said...

Yet the mass media, particularly television, kept reporting over and over again that the Ninth District is the most liberal federal District during their coverage of the Newdow decision. Prior to the Newdow decision, it appears that none of the newspapers or radio or television stations reported that the 9th was the most liberal District. Furthermore, when the media reports on other 9th District decisions the media doesn't repeat this claim that the 9th is the most liberal District. Apparantly, this false "fact" acquires some special significance and relevance only for the Newdow decision.

Explicit Atheist said...

Maybe the fact that a higher precentage of 9th Circuit decisions were reviewed by the Supreme Court in 2003 is evidence that the 9th Circuit is more conservative or more libertarian or less libertarian or more federalist or more anti-federalist or ... than the Supreme Court. By itself that statistic doesn't support any one of those particular conclusions. All that statistic proves is that in 2003 the Supreme Court thought the 9th Circuit made more decisions that addressed an important and/or interesting issue of constitutional or other federal law that was unsettled.

Ray Soller said...

Brad Hart wrote, "As was the case for most founders, Franklin believed in a neutral public religion, which was inclusive of all belief systems and creeds." That is most likely true as an historical observation, but please explain how you see them as having codified their personal religious belief as part of the Constitution.

Brad Hart said...

I don't think the founders codified their personal religious beliefs as part of the Constitution. In fact, I believe the exact opposite. The Constitution was clearly written as a secular document.

Tom Van Dyke said...

EA identifies the true mathematical point here, which Mr. Isaacson and Prof. Chermerinsky miss. Their counterargument is not mathematically relevant: If the Ninth were within statistical paramenters, its review rate [37%] would roughly equal its percentage of appellate cases handled [19%]. But it's double!

But EA also correctly points out that the double review rate doesn't necessarily seal the deal. However, that also means double the number of reversals than statistics would predict. I wrote "the one that's overturned most often at the top level" accurately, then, even when we figure in the Ninth's bigger workload than other circuits. [Another argument Eric offered.]

But to the key point, I used the word "radical," not liberal. The Ninth's reversals by unanimous Supreme Court decision [43%] is way off the charts, indicating that "radical" is not an unfair term, as everybody from Stevens to Scalia agreed the Ninth was off base in those cases. A unanimous Supreme Court should be a fair standard for the opposite of "radical."

That the Ninth handles 19% of all US appellate cases but represents 43% of the unanimous SC reversals suggests more than mere statistical anomaly, it indicates something fishy.

Unfortunately, I doubt seeing everything in black and white will make any difference to most people. Damn lies and statistics. Prof. Chermerinsky's faulty statistical counterarguments will continue to be trotted out, much like David Barton's bogus quotes from the Founding will.

But we tried, and my thanks to Mr. Huisman for providing usable data, and to Mr. Hart for actually addressing what my post was about in the first place. I notice these things, and they keep me going.

Explicit Atheist said...

Tom, those numbers also result in double the number of agreements than statistics would predict. We don't appear to have the number for how many of those agreeing decisions were unanimous, but unanimous decisions are too infrequent to make such a statistic from only one year's data meaningfull because of the large error factor from such a small sample size

Tom Van Dyke said...

Yes, EA, more data are always helpful. [I hate using data as a plural, even if it's technically correct.] Everyone except Mr. Huisman is too lazy to look any up. Meself included, but that's because I regret bringing it up in the first place, as it has absolutely zero to do with my actual point.

May I repeat, absolutely zero with my actual point.

But 10 unanimous reversals of Ninth Circuit decisions out of 68 total cases reviewed from all circuits in the country is statistically significant, and can not be waved away or buried, even by principled and accurate objections like yours. Which I hope you, EA, appreciate I've read, identified and acknowledged.

The search for truth must be a two-way street, between men of good will. If 10 nukings out of 68 cases doesn't at least raise your or Eric's intellectual curiosity, I cannot credit either of you with being anything but immovable objects, which only a fool would try to move. And Momma didn't raise no fools.

Explicit Atheist said...

The statistic is 10 of the 23 unanimous reversals, not out of 68 cases. I don't think a reputable polling organization or statistician would interpret a 43% result from a sample size of 23 versus an expected result of 37% as being a statistically significant difference.

Eric Alan Isaacson said...

It's not clear to me why data from 2003 alone should be the critical test.

In any event, the Ninth Circuit decided 5,295 appeals on the merits in 2003. Thus, if the Supreme Court reviewed 25 Ninth Circuit decisions in its October 2003 term, the Ninth Circuit's "review rate" appears to be something less than one-half of one percent. And the reversal rate was even lower.

These are only rough measures -- you'd expect the Supreme Court's October 2003 term to include review of appellate decisions from 2002, when the Ninth Circuit decided only 5,190 cases on the merits. Still, it should be clear that few Ninth Circuit decisions rated Supreme Court review, let alone reversal.

The court's current Chief Judge, by the way, is the Honorable Alex Kozinski, who used to do volunteer work for the libertarian-conservative Pacific Legal Foundation and was appointed to the Ninth Circuit Court of Appeals by Ronald Reagan.

Jonathan Rowe said...
This comment has been removed by the author.
Jonathan Rowe said...

Kozinski is one of my favorite guys. He was a mentor to his former clerk Eugene Volokh, one of my favorite bloggers and sharpest legal minds, and the Pacific Legal Foundation employs my blogfather Timothy Sandefur. Plus Judge Kozinski has a great, kinky sense of humor (which got him in semi-trouble in recent times).

Eric Alan Isaacson said...

I too think quite a lot of Judge Kozinski. He is both principled and brilliant.

Tom Van Dyke said...

The statistic is 10 of the 23 unanimous reversals, not out of 68 cases.

Both are accurate figures, EA. We need an impartial mathematician around here on retainer.

Mr. Isaacson, you continue to elide the statistical analysis. I do not know if you're being disingenuous or you honestly miss the math. Everytime your statistical arguments are answered, you move on to another one without acknowledging the time and effort people took to answer and engage you honestly on your own terms.

I don't believe you're playing fair. I'm also coming to believe you don't care.

Eric Alan Isaacson said...
This comment has been removed by the author.
Eric Alan Isaacson said...

Checking the Harvard Law Review’s statistics for the Supreme Court’s October 2003 Term, I find that the Supreme Court reversed or vacated the opinions below in 100% of the decisions reviewed from the Second Circuit (two of two), Fifth Circuit (six of six), Tenth Circuit (three of three), and Federal Circuit (one of one). Yet that does not suggest that these circuits, with their aggregate reversal rate of 100% for the term, are more “liberal” than the Ninth Circuit.

When I begin to review the term's unanimous “nukings” of Ninth Circuit opinions, I find cases like National Archives and Records v. Favish, 541 U.S. 157 (2004), in which the Ninth Circuit agreed with a lawyer for Accuracy in Media the “conservative watchdog group,” who sought release of photographs from Vince Foster’s death scene. Two of the Ninth Circuit’s Ronald Reagan appointees, Judges Noonan and O’Scannlain agreed with AIM’s lawyer, over the dissent of Judge Pregerson, a Jimmy Carter appointee (who is regarded by many as one of the Ninth Circuit’s more liberal judges). See Favish v. Office of Independent Counsel, 217 F.3d 1168 (9th Cir. 2000). But the liberal dissenter was right, for the Supreme Court unanimously reversed, ruling against the right-wing organization.

So, it appears that even Ninth Circuit rulings that favor conservatives are counted to demonstrate that the court is “too liberal.”

Explicit Atheist said...

I took graduate courses in probablity and statistics. The sample size for a statistic doesn't include any items that don't appear in the equation that produces that statistic. The equation is 43%=10/23. There is no 68.

Tom Van Dyke said...

The SC spent 14% of its time [10 cases] unanimously overturning the Ninth, which accounted for only 19% of all appellate cases [68]. This argues against your [very good] objection that perhaps the SC takes more Ninth cases because they're "interesting."

And I see once again that Mr. Isaacson emits a fog where clarity is needed. The reversal rate was never argued except by you, Eric, and Prof. Chermerinsky, and it's a faulty argument to begin with.

Since you didn't respond to the principled replies and rebuttals of your previous inadequate arguments, I can only assume your emphasis and style is on quantity of arguments, not the quality of them, so that the other fellow eventually gives up.

Perhaps that works for you in the courts. It works with me. It's not worth the trouble to dissect your latest attempt. [You opened the door to the problem of en banc decisions, for instance, which undercuts one of your own previous arguments.]

And since I withdrew my remark about the Ninth many comments ago, I can only assume you're content with diddling with the irrelevant details you think you can win, rather than the actual topic of the post.

I'm OK with that, Eric, except that we are obviously not searching for truth, but playing Debate Club or "Law & Order." I have to get better at declining to play such games.

Eric Alan Isaacson said...

Your insults offend me Tom.

Please retract them.

Tom Van Dyke said...

Eric, I'm not going to play the "offended" game with you either. You don't respond to the carefully-written rebuttals of your points, you just throw new ones out ad infinitum. That's insulting.

Your ignoring my main thesis and continuing to niggle at a side issue---even after I withdrew the remark for the sake of discussion---was insulting.

And my initial ire was directed at Prof. Chermerinsky's faulty statistical arguments. When you stood behind them, you simply got caught in the crossfire. But that was your choice.

Charles said...

Notwithstanding that Tom is correct that the "radical" issue is irrelevant to his post, the statistical question is much more interesting to me than the question of oaths. So, I'll belatedly chime in.

Quite apart from the statistical details, there seems to be a logical inconsistency in the general concept. Presumably, most of those who consider the 9th C to be "activist" (broadly meaning, I assume, that its decisions have a high probability of being "in error") also consider SCOTUS to be activist. If so, using SCOTUS reversals as a measure by which to judge lower court activism seems inappropriate. Eg, if SCOTUS decisions are in error with probability greater than 1/2, decisions from the lower court with the "most" (measured however one chooses) reversals are less likely to be incorrect, and thus that lower court would actually be the least activist.

Which suggests (as do many legal scholars) that the whole concept is rather silly. And to take the silliness to the extreme, one might argue that since 100% of overruled SCOTUS decisions come from itself, by the "reversal standard" SCOTUS is the ultimate activist court.

- Charles

Tom Van Dyke said...

But the end result is that you seem to be saying there is no sensible way to say any court is any more radical than any other. This strikes me as nihilistic or postmodern or something.

Charles said...

Since I'm not claiming any court is activist, radical, whatever, it's not my responsibility to come up with a way of demonstrating it. I was just pointing out a possible flaw in the proposed method of doing so.

According to the definition I use, I am in fact a nihilist, but it's not clear why that is relevant. And "postmodern" seems to be one of those words which is bandied about as an insult by people who may or may not have a clear idea - if any - what they mean. (I don't.)

- Charles

Tom Van Dyke said...

Me either, which is why I used "or something." But it indicates a belief that all truth is subjective.

Your objections are not unreasonable, BTW, and I certainly think being overturned 5-4 proves little or nothing. But I think the Ninth's sky-high unanimous reversal rate indicates something, and I do not think that skepticism should be ceded all wins. Since both you and EA have poked holes only in the side you disfavor, it's clear you're not exactly above the fray. I do think each side bears a burden of proof or at least the duty to provide argument and counterargument.

For instance, I do not surrender to the creationists just because they are correct there are huge holes in the fossil record's proof for evolution. [But that's another discussion, and I'm unprepared to argue their side of it. And I've yet to come up with a credible spokesman from their side on whom I could lean. But you see what I mean, and I hope you'll kindly stipulate the principle without us getting into the particulars.]

Charles said...

"you and EA have poked holes only in the side you disfavor"

As I tried to indicate in my comment, because of a relatively extensive background in probability theory, I am always interested in statistics problems just as an intellectual activity. I don't really think I have a "side" in this issue.

It is true that I consider the general concept of "activist" courts and judges to be somewhat dubious. But I was attacking this specific issue as a technician, not as a partisan. I even found Mr. Huisman's numbers highly suggestive until I saw the flaw. I won't rehash the point, but IMO Prof Chermerinsky and Mr. Isaacson's analysis is correct. And using their approach, I get the stats for unanimous reversals to be:

all = 34%, 9th C = 40%

Given the small sample size, I don't find that a convincing difference. But having no vested interest in the conclusion, I'll defer to someone who cares to interpret. I'm interested in "just the facts, ma'am".

- Charles

Tom Van Dyke said...

I respect your studies, Charles, and am willing to admit I'm wrong if you can show the error of my ways. However, if I'm on an assembly line and 20% of my widgets are flagged for inspection as possibly defective, and half of them are found to indeed be defective [a nominal defect rate of 50%], I have a failure rate of 10%.

If the next fellow on the assembly line has only 10% of his work flagged for inspection and has the same nominal defect rate as me [50%], his failure rate is only 5%.

This is the rebuttal to Chermerinsky's argument, as he examines only the "nominal defect rate." He says since they're all equal, case closed. I do not see how 5 = 10.

Now, you made some additional arguments about "interesting" cases being more likely to be "flagged," and I stipulated that was a sound counterargument. However, we cannot simply negate or ignore unanimous reversals, for reasons I already gave.

If the discussion is not to sink into total negation---nihilism, if you will---a court that represents 43% of all assembly line defects [unanimous reversals] but only does 19% of the widget-making deserves more than a shrug of the shoulders.

If you want to deconstruct the term "radical" to meaninglessness, or discard the whole deal for its small sample size, that's fine, but your graduate education in statistics is not brought to bear, then.

Charles said...

OK, if you insist. I hadn't actually thought it through carefully, but once I did I found that neither Prof. Chemerinsky nor Mr. Isaacson nor I had used quite the right numbers. But the results were pretty close.

Start with Mr. Huisman's numbers:

Appellate cases reviewed - 25/68 (37%)

Appellate cases reversed/vacated - 19/54 (35%)

Appellate cases unanimously reversed - 10/23 (43%)

9th Circuit % of total appellate cases - 19%

As previously noted, a disproportional number of cases reviewed came from the 9th C (19% of the total cases, 37% of those reviewed). We don't know why that is true, but it doesn't matter since the calculations involve only the 68 cases that actually made it to SCOTUS.

First, consider the reversal rate among all cases reviewed by SCOTUS:

Among the 68 cases that SCOTUS reviewed, how does the reversal rate among cases from the 9th C compare to the reversal rate among cases from the other circuits?

The reversal rate for cases from the 9th C is 19/25 = 76%. The reversal rate for cases from all other circuits is (54-19)/(68-25) = 35/43 = 81%. Ie, the 9th C is slightly better.

Next, consider the unanimous reversal rate among all cases reviewed by SCOTUS:

Among the 68 cases that SCOTUS reviewed, how does the unanimous reversal rate among cases from the 9th C compare to the unanimous reversal rate among cases from the other circuits?

The unanimous reversal rate for cases from the 9th C is 10/25 = 40%. The unanimous reversal rate for cases from all other circuits is (23-10)/(68-25) = 13/43 = 30%. Ie, the rate for the 9th C is one third more than the rate for the other circuits, probably statistically significant (but not as dramatic as the comparison of 19% and 43%, neither of which is relevant).

Finally, consider an alternative approach to unanimous reversals:

Among the 54 cases that SCOTUS reviewed and reversed, how does the unanimous reversal rate among cases from the 9th C compare to the unanimous reversal rate among cases from the other circuits?

The unanimous reversal rate for reversed cases from the 9th C is 10/19 = 53%. The unanimous reversal rate for reversed cases from all other circuits is (23-10)/(54-19) = 13/35 = 37%. Ie, the rate for the 9th C is 43% more than the rate for the other circuits.

I'm not sure which of the last two calculations is more meaningful. Either way, it does appear that although the overall reversal rate for the 9th C is in line with the rate for the other circuits, unanimous reversals are more likely for them by a statistically significant amount.

- Charles

Matt Huisman said...

Some comments:

1) The percentage of cases reversed/vacated that were granted certiori is the least relevant statistic cited here. The appellate cases reversed/vacated - 19/54 (35%) represents a hypothetical 'perfect' SCOTUS selection rate.

2) The percentage of unanimous reversals from a circuit's total reversals is not particularly significant here either. The reason for using the unanimous reversal standard is to 'clean up' the sample (hopefully removing any uncertainty due to inherent SCOTUS issues as opposed to the 9th Circuit).

Again, I don't like relying on these statistics to make the charge of radical against the 9th, but I like them even less in their defense.

Matt Huisman said...

I should have added in #2 above that the 10/23 (43%) unanimous reversals is a valid metric (along the lines of #1) because it allows comparison among a more consistently applied selection standard.

Charles said...

Mr. Huisman:

Just to be sure we are on the same playing field, let me repeat yet again: I'm not defending anything, I'm trying to work a math problem and to the extent it's relevant, understand the SCOTUS procedures (IANAL).

To those limited ends, I would like to understand your points. Re item 1, how does cert get into the statistical analysis. My understanding is that cert is the process by which a case gets to SCOTUS, so that once one starts computing using the 68 cases it becomes irrelevant. Am I missing a subtlety? Also, what does "a hypothetical 'perfect' selection rate" mean? And last, why do you consider the various percentages computed by dividing 9th C numbers by total numbers relevant since we have established that a disproportionate number of cases came from the 9th C?

I don't know to whom item 2 was directed. Since I didn't introduce the idea of using unanimous reversals and was just "running the numbers", I assume it wasn't me. If that assumption is wrong, please elaborate since I missed your point. (And if it is right, "never mind".)

Since I started composing, your second comment appeared. Again I don't understand your terminology, specifically "a more consistently applied selection standard". I infer that you are addressing some issue relating to the selection of cases which are granted cert, but I don't understand what it is. Does the fact that a disproportionate number of cases came from the 9th C tell you something that you are trying to "add" to the raw numbers?

Thanks - Charles

Matt Huisman said...


My understanding is that we are trying to come up with a measurement to determine which circuit is 'most overturned' at the top level. We've all agreed to adjust (weight) for case volume.

The key here is that we're looking for cases that were overturned. SCOTUS does a nice job of eliminating thousands of cases from our potential sample, but the fact that their selections don't guarantee us a 100% overturn rate should not stop us from simply removing the poor selections from our sample. The same logic applies to the unanimous reversals, which we use only to reduce the risk that a given SCOTUS decision is a 'mistake'.

So when we find that 35% of the cases overturned came from a circuit that only represented 19% of the cases adjudicated, we have a weighted measure of who is 'most overturned'.

I don't see why the fact that a disproportionate number of cases were selected from the 9th Circuit should not be held against them. SCOTUS does not use random sampling techniques to select cases, they use a method that (for the most part) looks for problems. My assumption is that they apply a consistent standard uniformally across all circuits.

If the Circuit A had a 75% overturn rate on 100 cases, and the Circuit B had a 100% overturn rate on 2 cases, and their overall workload was identical, who would you say is more out of step?

Charles said...


Re assembly lines and 75% of 100 vs 100% of 2, I don't find verbal approaches very reliable in statistical games because there are usually implicit, and hence often hidden, assumptions. If you are familiar with the Monty Hall problem, you can appreciate this. If not, check it out at wikipedia. It has fooled many, many well-versed people (and me, for what that's worth) who made their first guess based on "common sense". So, for better or worse, the following is my best semi-formal argument.

You make three key assertions:

1 - SCOTUS does not use random sampling techniques to select cases

2 - they use a method that (for the most part) looks for problems

3 - they apply a consistent standard uniformally across all circuits."

First, let's agree that #1 is obviously true. We also hope #3 - being basically an assumption of judicial integrity - is also, and hope it applies across cases as well. So, the interesting assertion is #2.

That assertion seems to suggest that the cert-granting justices are primarily looking for cases that they consider to have been wrongly decided, which in turn seems to be equivalent to assuming a priori bias. According to the "umpire" speeches of recent Justices in their confirmation hearings, that isn't supposed to happen. But even in the most unlikely event that it does, implicit in their doing so is an assumption on their part that the a priori probability of the selected cases from the 9th C being wrongly decided is higher than that of the other circuits. But if that is true and your third assertion is correct, IMO that implies that the a posteriori probability of being wrong and thus reversed should also be higher. But the limited data we have does not suport that conclusion in general although it may in the special case of unanimous decisions ("may" because of the small sample size).

That "IMO" is based on the following:

Let P{Ci=O} be the probability that case Ci from the ith circuit is reversed, let P{Ci=O|Ci=X} be the a posteriori (AKA conditional) probability of case Ci being reversed given that it was wrongly decided below, and let P{Ci=X} be the a priori probability that case Ci was wrongly decided. Then

P{Ci=O} = P{Ci=O|Ci=X}P{Ci=X}

Your assertion 3 is, I believe, equivalent to the conditional probability being constant across cases and circuits. Hence, if the a priori probability of a case being wrong is greater for one circuit than for others, so should be the a posteriori probability of being reversed.

In violation of my diatribe against verbal arguments, I will offer a "limiting case" observation, which doesn't prove anything but I think is suggestive. Suppose the cert-granting justices are absolutely convinced that in the current year one circuit is wrong so much more of the time than the others that they decide to pick cases from only that circuit. Then all of your ratio statistics are 100%. But if the reversal rate actually turns out to be consistent with the historical norm, what should one conclude?

- Charles

Tom Van Dyke said...

Your equation is really cool-looking, but the point can be made without it. What you overlook is another "limiting factor," the number of cases heard. If the SC reviewed only Ninth C cases, if it heard 25 cases, we might look for a reversal rate approximating the historical average. But if it reviewed 250, we would expect the rate to plummet. The Ninth isn't wrong all the time, nor consistently wrong across all its cases, most of which are relatively uncontroversial. That more of its cases are generally flagged a priori is no sign of bias by the SCOTUS, only of something fishy in the Ninth's decision; indeed your own argument would dictate that the a posteriori reversal rate would fall below historical averages if the Ninth were being unfairly reviewed at its greater rate. Mr. Huisman's argument stands.

Unfortunately, it's not available online, but Judge Richard Posner [admittedly, a conservative] did what's credited as the most thorough breakdown of the Ninth, and concluded its sky-high "dunk rate," as they call it [unanimous and supermajority reversals] cannot be explained away by statistical or other means.

I always enjoy Edward Lazurus' writing on the SCOTUS, although he's an unapologetic gentleman of the left with whose conclusions I frequently disagree.

In this very good 2002 article, Lazurus concedes the results of the Posner study, and freely admits that "[t]his Circuit, unlike so many others, has a decidedly more liberal bent than the current Supreme Court."

But Lazurus doesn't see that as a problem. He argues that the Ninth is simply visionary, ahead of the curve, and to a gentleman of the left, that's a very good thing. My use of "radical" may be a bit perjorative, and "visionary" would be an approving way of putting it, but Lazurus and I are acknowledging the same thing.

The article touches briefly on the Pledge decision, and argues nothing in conflict with my original post or in this comments section. Neither do I have any problem with Lazurus' dissent from my views. People of good conscience can disagree about stuff. 40-odd comments later, we're back where we started, except perhaps for some math to tidy up.

[Note Scalia's prescience about a slippery slope. Heh. The man's no fool.]

Charles said...


re my equations: "cool" or not, they and the associated assumptions constitute my argument. If one disputes any of those, I'll consider the objections. Otherwise, I'm done with the issue.

"Judge Richard Posner [admittedly, a conservative]"

I have never understood why people say this. Judge Posner is one of my very few contemporary "idols", and it's because I consider him to be essentially a logic machine (recall that I'm a materialistic nihilist). I have read many of his books (LD&P multiple times and counting) and have never found him to be ideological. I assume the "conservative" label has been affixed because he occasionally comes up with conclusions that don't fit the "liberal agenda" (eg, justifying Bush v Gore and Clinton's impeachment), but I've never found his reasoning to be blatantly biased. I am a regular reader of the Becker-Posner blog, although I'm less impressed by Becker, whom I probably would label an ideological conservative if I knew what that meant.

Lazurus's article was OK, but as I've tried to convey, I'm not really interested in the "activist" court issue per se, just the statistical aspect.

"Note Scalia's prescience about a slippery slope."

I see his "if A then B" arguments not as slippery slopes, just logical consistencies. If school prayer=unconstitutional, then "under God" as well; if homosexual rights OK, then gay marriage OK. In each case, Scalia has a problem with A because it logically leads to B, and he fancies neither A nor B. He's not a fool, but he is unquestionably biased (as, of course, we all are, but some honestly admit it).

- Charles

Matt Huisman said...

Hence, if the a priori probability of a case being wrong is greater for one circuit than for others, so should be the a posteriori probability of being reversed.

No. You're dealing with adverse selection here. SCOTUS has removed all the decisions it does not care about, and applied it's a priori standard. At this point, the likelihood of a cert-worthy case being reversed is based on the likelihood that SCOTUS reverses the cases it selects. (The circuit of origin is almost irrelevant, SCOTUS does not care. This is why you find that the 9th C reversal rate is remarkably in line with the overall reversal rate.)

Your method does not weight for the total workload of a circuit. You have reduced the denominator sample size to the cert-worthy cases, which seems highly unfair to a circuit that gets far fewer cases selected (or reversed) than the 9th does relative to its workload.

Charles said...

OK, I think I'm beginning to see your perspective. And to the extent that I have it right, my reaction is that most of this discussion is irrelevant since your argument reduces to the single fact that the cert rate per workload case for the 9th C is roughly twice that for all circuits. (Not to mention that the charge that initiated this subthread - that the 9th C's "reversal rate" is much higher than that of "other circuits" - is quite misleading in at least two ways). Thus, the only question of interest is "why so?". I have no idea and don't really care - that's not a statistical question but a legal/procedural one which I'll gladly leave to others to address.

I will repeat my previous observation that if Tom is correct and it's because the cert-granters think the cases coming out of the 9th C are "fishy", it has to be because of some a priori bias on their part, which seems in itself a very strange posture for impartial "umpires".

But thanks for the exchange anyway. It did force me to learn a little about adverse selection, which has of late come up regularly but never before in a context that motivated me to pursue it.

- Charles

Matt Huisman said...

If the "reversal rate" tag is misleading, you can go back and re-read the comments to find out whom to complain to; neither Tom nor I had anything to do with it.

As for "Tom's observation" that cert-granters are biased because they think the cases coming out of the 9th are "fishy", you'd better re-read his comments because you have entirely missed his meaning.

And as for whether my argument "reduces" to cert rates per workload case, you can go back and re-read my comments on that too (although at least this time you're close).

Charles said...


I read and reread comments carefully before replying. If I'm still missing something, I doubt that the problem is entirely with my reading comprehension skills, which I have good reason to believe are quite adequate to the level of material under discussion here. If you'd like to correct my misunderstandings, please do. If not, thanks again for the discussion.

- Charles

Tom Van Dyke said...

Oh, let's close it now. It was a very good discussion and I for one learned stuff. Thx to all involved, especially you, Charles, for your ace arguments and expressed openness to counterarguments, and to Mr. H for serving as the author's ombudsman.