Filibusters Again

by hilzoy

Steve Benen, Matt Yglesias, and Kevin Drum are all calling for reform of the filibuster. I agree. I am of two minds on the question of eliminating it entirely. (To anyone who thinks it's just obvious that the filibuster should be eliminated, I have three words for you: Janice Rogers Brown.) But what seems absolutely clear is that if it is kept around, it ought to be transformed back into a tool that is actually painful for the minority to use, and that they will therefore use only when they feel very, very strongly. As Kevin wrote:

"The filibuster was never intended to become a routine requirement that all legislation needs 60% of the vote in the Senate to pass.  But that's what it's become.  It's time for reform."

I don't think it's enough, though, to say that Senators who want to filibuster should be made to actually stand up and speak for hours on end. To see why, consider this:

"While a filibuster would seem to be more taxing on the side doing the talking, that isn't necessarily the case. The filibusterers need only one person in the Senate chamber at any one time, prattling away. The other side must make sure a quorum -- a majority of all senators -- is on hand, a constitutional requirement for the Senate to conduct business. If there's no quorum after a senator has demanded a quorum call, the Senate must adjourn, giving those leading the filibuster time to go home, sleep, and delay things even more. To ensure a quorum during the rancorous civil rights filibusters, cots were set up in Senate anterooms, and majority senators presented themselves in bathrobes during early-morning quorum calls.

Those seeking a quorum can even demand that the Senate's sergeant at arms arrest senators who aren't present and drag them into the Senate chamber, a measure that has led to absent senators playing hide-and-seek with police officers around Capitol Hill. As recently as 1988, officers physically carried Sen. Robert Packwood onto the Senate floor at the behest of then-Majority Leader Byrd."

This means that the pain of filibusters falls disproportionately on the side that is trying to end debate, not on the side that is mounting the filibuster. Senators do not like to hang around the Senate all night long. Sometimes, they would rather catch up on their sleep, go to fundraisers, fly back to their districts, or do whatever else they feel like doing. As things stand now, only one Senator from the side mounting a filibuster has to give up the rest of his or her life in order to be present in the Senate. The rest of them can just catch up on their beauty sleep. The side that is trying to end the filibuster, by contrast, has to keep almost all its members around in case of quorum calls.

If we're going to reform the filibuster, this has to change. The Senate might make cloture votes require 60% of the votes of those who are present and voting, for instance. That would mean that the side that was mounting a filibuster would have to keep all its members around for the duration. Alternately, the Senate might adopt a rule that said that during filibusters, if a quorum was not present, the Senator who was speaking could decide to go on speaking or to allow a vote on cloture, to be decided by a majority of those present and voting. If s/he decided to go on speaking, s/he could do so, but no other Senate business could be conducted until the next business day. If s/he opted for the cloture vote, it would take place.

There might, for all I know, be problems with either of these proposals. And there's probably an even better proposal out there. But what has to change, I think, is the fact that Senators can now declare their intention to filibuster and either have their way (if no one forces an actual filibuster), or visit considerable inconvenience on their opponents (if a filibuster is forced), without having to suffer the same inconveniences themselves. 

That's an incredibly perverse set of incentives. It might have been designed to create the idiotic situation we have now, in which a party that has seventeen fewer Senators than the other is nonetheless in a position to dictate what the majority party can pass, not just on issues on which they feel very strongly, but as a matter of course.

Comments

[Justin]

I've always been kicking around an idea that there should be some sort of "increased price" for the filibuster--each time you use it, it would require more senators to prevent cloture (perhaps it could be more gradual).

The obvious problems I see are that while the filibuster is typically used by the minority in the current setup, this doesn't have to be the case--there have been times when there was less intra-party discipline. In such a case, you'd have to assign responsibility, and that could be tricky.

[gregspolitics]

Actually, I think that simply requiring a real filibuster would go a long way towards solving the problem. 60 votes has become the threshhold for passage because Harry Reid has allowed it. Thus, reporting becomes failure to obtain enough votes, rather than that the GOP is filibustering. If they had to really filibuster, then politically the filibusters would often be hard to sustain as Obama would go around the country saying I am reaching out but all they want to do is oppose. If the stimulus had not required 60 votes because of the budget rules, I think the GOP would have had a hard time sustaining a filibuster because of public reaction. In that regard, while the stimulus required 60 votes, passage of the Obama budget will not. The reconciliation process can be used to prevent a filibuster and there is no 60 vote requirement so long as no program creating a net deficit extends beyond 10 years. This is why Dems who wanted the stimulus to include more should not be so upset. Much that should be done in 2010, 11 and 12 can be done in the budget - the stimulus probably puts as much into 2009 and early 2010 as could practically be accmplished anyway. More investment funding can be put in the 2010 budget and contiued into the following few years, even if it temporarily increases the deficit. Even a health care plan could be passed through the reconciliation process so long as any cost of the plan to the govt in the out years was paid for, leaving room for transitional subsidies in the early years without the necessity of $ for $ taxes or spending cuts. This, by the way, is why the Bush tax cuts expire at the end of 2010. Because there was no offset against them increasing the post-2010 deficit, they could only last for 10 years when passed as a budget reconciliation in 2001.

[Dan Miller]

Hilzoy, I don't think Janice Rogers Brown does as much work as you think she does in this example. Sure, there have been good uses of the filibuster on occasion. But against that you have to weigh universal health care or (more historically) a decades-long delay in civil rights laws. We could go back and forth naming counter-examples all day, but if you look at counter-examples, every other major democracy has managed to get along without a supermajority requirement to pass all controversial legislation (let alone a supermajority requirement in such a misapportioned branch as the Senate, where Vermont=Texas). There would be some bad outcomes if we eliminated the filibuster, but it's a lot more common-sensical and a lot more conducive to ending misuse to just take the thing out entirely.

[tomtom]

Great post! Now I know why the Dems allow the Republicans to filibuster without actually doing the work of reading from the phone book.

"The Senate might make cloture votes require 60% of the votes of those who are present and voting, for instance. That would mean that the side that was mounting a filibuster would have to keep all its members around for the duration"

This might be too extreme. The burden should be on the party that is filibusteringbut requiring ALL of the filibustering party to be present seems extreme. It would be come virtually impossible to sustain a filibuster more than a day or so, and I think we want them to be sustainable until the nation really takes notice and weighs in. That could take a week.

Hizoy's second more complex proposal seems very good. Suppose the filibustering party quickly brought in all their members in the middle of the night so the filibustering senator could call a majority vote on cloture. So long as the cloture vote was a majority vote the party trying to break the filibuster would only need to quickly round up most (not all) of their votes to win cloture and end the filibuster.

This seems sustainanable for weeks at very high personal cost for Senators of both parties. Isn't this what we want?

[Bernard Yomtov]

I would rather just have some votes require a super-majority and do without filibusters.

The most obvious candidates are judicial confirmations, though there could be others. Unlike legislation, they cannot be amended or modified by subsequent action, and unlike other appointments they have no fixed term, and no supervision from an elected branch. "Repeal" via impeachment is rare and difficult. So we're pretty well stuck with whoever we get.

One advantage of this approach is that it's a lot more honest than the current method.

[Sebastian]

I'm pretty sure that if you made Republicans really filibuster and then play TV commercials of him reading from the phone book or whatever, that it would stop being used so trivially after just one painful two or three days.

[John Miller]

Simple, make it a rule that any Senator who votes against cloture must remain in the Senate chamber until such time as a vote for cloture passes (allowing for appropriate bathroom breaks).

[Nell]

Bernard Yomtov: I would rather just have some votes require a super-majority and do without filibusters.

But that's already the case with a majority of votes, because the Democratic leadership, unwilling to force a real filibuster, agrees to a rule that requires 60 votes if a Republian so much as threatens to block unanimous consent.

[OCSteve]

Take control of the Census, bring back the fairness doctrine, eliminate the filibuster… I voted Democrat yes, but I didn’t intend it to be a vote for permanent one-party rule…

[Scott P.]

Simple, make it a rule that any Senator who votes against cloture must remain in the Senate chamber until such time as a vote for cloture passes (allowing for appropriate bathroom breaks).

So no more than one day of debate on any issue, no matter how important?

[Nell]

Who exactly is calling for the return of the Fairness Doctrine, OCSteve?

The party in power at the time of the census always controls the census. Please explain what's different now.

As to the elimination of the filibuster: As far as I can tell, this is Hilzoy exploring the issue, rather than anyone in a position to do so changing the rules of the Senate (which is a very arduous undertaking).

So please drop the whining about 'permanent one-party rule' and address the issue at hand with a substantive point.

[Harald Korneliussen]

"The filibuster was never intended to become a routine requirement that all legislation needs 60% of the vote in the Senate to pass."

I don't know why a lot of weird characters were tagged onto the end of that sentence, it should obviously read "The filibuster was never intended to be". That the legal loophole that is filibustering is useful to correct the other mess-ups in the system (life time appointed high priests of the constitution) doesn't make it a designed feature.

[d'd'd'dave]

Why not just adopt a rule that the democrats can do whatever they want?

[KCinDC]

OCSteve, if you can get the Republicans to go back to the way things were before the 110th Congress, when they brought in the idea that filibusters should be threatened for essentially everything, rather than being reserved for extraordinary circumstances -- the new rules which the media innocuously describe as a 60-vote requirement for passing anything in the Senate, as if it's written into the Constitution -- then I'm sure the sentiment in favor of eliminating the filibuster will die down.

I don't understand the point of voting for the Democrats if you're perfectly happy to have 41 Republicans prevent them from passing anything. I guess the ideal government for you is one that takes no action and never fixes anything resulting from previous actions.

[Jesurgislac]

Invoking the rule that a filibuster has to be an actual filibuster seems simple enough, without changing any actual rule of operation.

Though Republicans will doubtless whine about any rule change that stops their party getting its way against the will of the electorate and the majority vote in their house of government. Whiny Republicans are even more annoying than whiny small children, and should get what was my usual response to whiny small children when they started going "B-u-u-u-t I w-a-a-a-n-t it...":

"Tough."

Repeated often enough, this got the other kids giggling at the whiny kid, and eventually the whiny kid would get the giggles too, proving that even a five-year-old brat is more grown-up than a Republican...

[OCSteve]

On the census – it was clearly a trial balloon. Should it political or accurate? On the FD…

Obama was unequivocal about it during the campaign:

"Sen. Obama does not support reimposing the Fairness Doctrine on broadcasters," press secretary Michael Ortiz said in an e-mail to B&C late Wednesday.

Now?

"I'm going to leave that issue to Julius Genachowski, our new head of the FCC, to, and the president, to discuss," Axelrod said. "So I don't have an answer for you now."

We’ve had a couple of high profile mentions of it in the last couple of weeks including by Bill Clinton.

KC: I don't understand the point of voting for the Democrats if you're perfectly happy to have 41 Republicans prevent them from passing anything.

Not just anything… I just wasn’t happy about how the biggest spending bill in history went down…

Roll back Bush’s tax cuts; get out of Iraq and even Afghanistan, make torture and extraordinary rendition 100% illegal, scale back defense spending a bit, and most anything else you want. You’ll have my support. Hell, raise the cap on FICA so I can believe it may be there for me and I’ll quit bitching about SS (and calling it a ponzi scheme).

You know what? I’m going to shut up and give the dude his traditional 100 days before I say anything else. So far I’m not impressed by his priorities. But the dude deserves a decent chance. And I’ll admit to coming down on him too hard too soon.

[Gary Farber]

"Actually, I think that simply requiring a real filibuster would go a long way towards solving the problem."

This entire comment seems 100% unresponsive to Hilzoy's post. You address exactly 0% of her points regarding why your proposal isn't a solution.

Maybe you could try addressing the actual post?

"...but if you look at counter-examples, every other major democracy has managed to get along without a supermajority requirement to pass all controversial legislation"

Sure, and when we revise our form of government so that the government, both Executive and Legislative, falls whenever they fail a vote of confidence, this might make sense.

Absent that revision, you'll notice that our form of government isn't at all like a parliamentary system, and therefore can't work the same way, because it doesn't. So absent wholesale reform to a parliamentary system, this sort of piecemeal reform proposal just doesn't make sense.

What it would mean is that there is no check whatever (other than the Supreme Court) on a government consisting of the same party holding both the presidency and the Congress.

Mind, it's true that we had almost that situation during the Bush Presidency, but at least it wasn't solidified into the structure of our government, and was dependent on the wimpiness of Democrats; your proposal, absent more wholesale reform, would surrender all power to whomever the current rulers are, and that's just a complete destruction of the whole concept of "checks and balances," and split government.

Thanks, but no thanks.

"Take control of the Census, bring back the fairness doctrine, eliminate the filibuster… "

For god's sakes, who on earth is calling for bringing back the Fairness Doctrine, other than insane Republicans who keep ranting about it for absolutely no reason. It's 100% hallucination. (But apparently they're just that desperate to find something to alarm their followers about, no matter that it has no connection whatever to reality.)

Also, cite please on what "take control of the Census"?

[John Thullen]

"why not just adopt a rule that the democrats can do whatever they want?"

Done.

[Gary Farber]

"Why not just adopt a rule that the democrats can do whatever they want?"

Too close to how the Republicans ran the Congress from 1994 through 2006. You might want to look into that before accusing the opposing party of making rules remotely as one-sided.

[Decline and Fall]

If it were simply required that all Senators engaging in a filibuster be either in Senate chambers or in their offices it would be an improvement.

In late 1994 (before November, when the GOP took over both houses of Congress) I lived next door to Sen. Jess Helms in Arlington. I remember watching some Republican filibuster on C-SPAN and looking out my side window to see my neighbor washing dishes with his wife. Then there was a roll call and his car pulled up, he walked out with a necktie in his hand, and appeared on television maybe 20 minutes later.

[Bernard Yomtov]

Bernard Yomtov: I would rather just have some votes require a super-majority and do without filibusters.

But that's already the case with a majority of votes, because the Democratic leadership, unwilling to force a real filibuster, agrees to a rule that requires 60 votes if a Republian so much as threatens to block unanimous consent.

Nell,

True. My suggestion is to specify a few things that simply require a super-majority, and do away with the filibuster. Aside from all else, this would make the procedures clearer to the public. I'd also do something about "holds" and "blue slips" and whatever else.

I don't really see what purpose all that serves.

[KCinDC]

Gary, as I understood it, Dan's proposal was to eliminate the filibuster, not to convert the US government into some sort of partially parliamentary system. That may or may not be a good idea, but I don't see what it has to do with removing constitutional checks and balances, since those are between branches of government, not between parties within the legislative branch, and neither filibusters nor parties are in the Constitutional.

[KCinDC]

... or the Constitution, even.

[KCinDC]

Roll back Bush’s tax cuts; get out of Iraq and even Afghanistan, make torture and extraordinary rendition 100% illegal, scale back defense spending a bit, and most anything else you want. You’ll have my support. Hell, raise the cap on FICA so I can believe it may be there for me and I’ll quit bitching about SS (and calling it a ponzi scheme).

And none of those things will happen if the Republicans in the Senate block cloture for the legislation required to accomplish them, especially if Coleman continues to delay the seating of Franken and Kennedy isn't replaced.

[Gary Farber]

"I don't really see what purpose all that serves."

George Washington:

The earliest known written version of this story appears in an 1871 letter from constitutional law professor Francis Lieber to Ohio Representative and later President James A. Garfield. Lieber recounted a story he had heard about Thomas Jefferson's visit to Mount Vernon where Jefferson disagreed with Gen. George Washington over the need for a bicameral legislature, and Washington's response:

"You, yourself," said the General, "have proved the excellence of two houses this very moment."

"I," said Jefferson. "How is that, General?"

"You have," replied the heroic sage, "turned your hot tea from the cup into the saucer, to get it cool. It is the same thing we desire of the two houses."

The Washington-Jefferson dialogue drew further attention in the writings of the late 19th century American historian Moncure D. Conway, who altered the language and the beverage:

There is a tradition that on his return from France, Jefferson called Washington to account at the breakfast table for having agreed to a second chamber.

"Why," asked Washington, "did you pour that coffee into the saucer? Why did you do that?"

"To cool it," answered Jefferson.

"Even so," said Washington, "we pour legislation into the senatorial saucer to cool it."

Robert Byrd has a lot more to say there, but I'll quote only this:

[...] Senators and other close observers of the institution have grappled with their own ideas about the Senate seeking to highlight its unique and enduring attributes, and to explain its role in the American system of checks and balances. What is it? What is it? What is it that makes the Senate stand apart from other legislative bodies? What is it? What is it that makes the Senate stand apart from other legislative bodies?

Why have its seemingly arcane rules and traditions survived, and what purpose do they serve?

[...]

As a deliberative body, the Senate has been hailed as a place for second thoughts, as a continuing body, and as an institution that values its traditions. The form of Senate elections, changed by constitutional amendment, and the rules for unlimited debate and cloture have been adjusted over the years, but the Senate still differs in fundamental ways from the House of Representatives. It stands out, the Senate does--the Senate stands out as a body of individuals with peculiar folkways that have fostered what has been described as the "Senate type."

A body of equals among individuals and among States, the Senate has been a difficult institution to lead. Its deliberations have frustrated impatient Presidents. Well, who cares? Senators don't care if they frustrate Presidents. Presidents come and go. Senators may stay on and on and on.

Its deliberations have frustrated impatient Presidents, leaders of the House, and even, yes, leaders of the Senate who seek speedy enactment: Let's get it done. We are in a hurry. Let's get it done. Do it now.

Remember that TV advertisement which said, "Do it now, do it here; do it now, do it now?"

There have been many efforts to modernize the Senate in order to meet new challenges. I have been here a long time. I have seen these efforts on the part of Senators. Some of them come over from the House of Representatives. They want to make this body into another House--let's get it done. Get it done; do it now; do it here; fast.

The point of the Senate is to have lots of ways of delaying things.

I'm fine with doing something to change the filibuster back to an extreme measure. What the Republicans have done in turning it into a normal tool of legislation, requiring 60 votes on any important issue, is simply wrong. We need to change the filibuster back to something that can only be used for the most vital of issues (such as a lifetime appointment to the Federal courts, or at least to the courts of appeals).

But I'd be very cautious about sweeping away everything that makes the Senate different than the House of Representatives. Why not just eliminate the Senate entirely, then? And there will be some who like that idea, but I say that if we do that, we should just switch entirely to a parliamentary system, rather than merely eliminate the Senate, and any significant check on the President; a President in such a situation would be a virtual dictator. I shudder at the thought.

[Nell]

Bush's tax cuts will expire unless Copngress takes action to maintain them. They had a sunset provision. So that campaign promise should be easy to keep.

But I'm not done thrashing OCSteve about the Fairness Doctrine.

Politico:

...no member of Congress has scheduled hearings, there is no Fairness Doctrine legislation being introduced, and the long-dormant broadcast law is likely to stay that way.

This Michael Calderone piece, together with subsequent ones, is a perfect example of both mocking and drumming up "it's out there" hysteria.

OCSteve, it saddens me to have you responding like a lemming to right-wing hate radio and Limbaugh talking points. Even if there were some plausible reason to think that the Obama administration were going to make a push to restore the Fairness Doctrine, which there is not, how in hell would that have anything to do with "permanent one-party rule"?

You've already walked it back with your '100 days' comment, but it's going to be hard to overlook the ditto-head nature of your reaction.

[hilzoy]

OCSteve: it is worth noting that reforming the filibuster is something I favor, not one of Obama's priorities. If he were to push for it, it would surprise me a lot, in large part because I think he has tried to show some respect for the Congress' role, which surely includes setting its own rules.

[hilzoy]

Plus, what I'm advocating is just making the filibuster something that requires some sacrifice for the filibusterers, as opposed to a cost-free way to impose a supermajority requirement. That would allow it to be used whenever any Senators felt strongly about it, and could rustle up 39 other Senators to join him or her.

[Model 62]

Drum et al's concern about abuse of the filibuster appears to have been prompted by the recent wrangling over Obama's stimulus bill. The Stimulus Bill, however, wasn't filibustered, and filibuster reform (or cloture vote reform) wouldn't have eliminated the 60 vote hurdle because Pay As You Go rules were in effect. PAYGO in the Senate requires a point of order vote (3/5ths majority -- 60 votes -- to carry) if the cost of a proposed bill will not be matched by spending cuts elsewhere or tax increases.

[Bernard Yomtov]

Gary,

Yes, yes. I understand that the Senate is supposed to be more deliberative, and so on. That's the reason for the six-year term, of course.

But that doesn't mean that everything that can delay legislation is a good idea, or that rules that seem sensible can't be misused.

As to the Senate itself, I am among those who think the two-senators per state rule is a mistake, a big flaw in the Constitution. That doesn't mean I think we need to abolish it or go to a pure parliamentary system. It does mean I think the Constitution is not perfect.

[OCSteve]

Gary – nice to see you commenting. I was beginning to worry about your absence.

For god's sakes, who on earth is calling for bringing back the Fairness Doctrine, other than insane Republicans who keep ranting about it for absolutely no reason. It's 100% hallucination.

Jeeze dude – there were many high profile supportive comments during the election. Since then, we have Bill and now Axelrod refusing to say it isn’t going to happen, refusing to confirm Obama’s election position on the topic.

Also, cite please on what "take control of the Census"?

What else do you call it? As I said, I think it was a trial balloon and it got shot down. (Hopefully)

KC: And none of those things will happen if the Republicans in the Senate block cloture for the legislation required to accomplish them…

So I agree with you in this case – but support taking away the filibuster for all time? I mean, I support Democrats now because they are the (obviously) lesser of two evils. But for all time? No way dude. Republicans are finally starting to get some balls back as far as spending goes. Yes – it is hypocritical, and pathetic even - but I’ll take what I can get.

Nell: …it saddens me to have you responding like a lemming to right-wing hate radio and Limbaugh talking points.

If I’ve ever listened to Limbaugh or RW radio at all it was 15 seconds at a time searching for a channel on a long drive. I just don’t. I find talk radio of any variety extremely boring. That is the case for Limbaugh or Air America. I want tunage. Some Cure, or AC/DC, or Aerosmith, or Zep. Some Fray for modern stuff. Hell, anything classic rock in a pinch. I’ll take the frick’n Spice Girls over talk radio…

You guys are all or nothing. You say Republicans enforce their ideology on the members of the club?

Here is where I agree with you and would support Obama and Democrats (not all inclusive – just off the top of my head):

-Roll back Bush’s tax cuts
-Get out of Iraq and even Afghanistan
-Make torture and extraordinary rendition 100% illegal
-Scale back defense spending a bit
-Raise the cap on FICA
-Repeal DADT.
-Constitutional amendment to legalize gay marriage.
-Swap the estate tax for fixing AMT.
-Fund contraceptive programs (straight out, not including them in an emergency stimulus bill).

Again – that is not all inclusive.

I’m just the stereotypical radical wing-nut huh?

[russell]

For god's sakes, who on earth is calling for bringing back the Fairness Doctrine, other than insane Republicans who keep ranting about it for absolutely no reason. It's 100% hallucination.

Jeeze dude – there were many high profile supportive comments during the election. Since then, we have Bill and now Axelrod refusing to say it isn’t going to happen, refusing to confirm Obama’s election position on the topic.

And all these years I thought I was the only person left on god's green earth who was interested in bringing back the Fairness Doctrine.

It's been lonely. You're making my day.

I’m just the stereotypical radical wing-nut huh?

No, OC, you're not. It's just the hat you're wearing here on ObWi these days.

[John Miller]

OCSteve, no you are not the stereotypical right wing wing-nut, we have other commenters here that handle that role.

But you do tend to over-react at times. And this census thing is one of those. You want to provide a cite to an actual trial balloon? Look, the WH is always involved to some degree in the census, so it being involved is no big step toward one party rule.

Also, don’t talk about this stimulus being rammed down our throats. Republicans were involved all along. The fact that they chose to play obstructionist is not Obama’s fault. And the fact they lie about is also not Obama’s fault.

And I am not so sure they have gotten their balls back as far as spending goes. It’s spending by Democrats they object to. It is productive spending they object to.

And they still are not showing any sign of actual fiscal responsibility.

So I guess what I want to say, is if you have a problem with what is happening, don’t just throw a charge out there and walk away. Defend the accusation. You have a voice most of here want to hear, even if we disagree with you. But it is because you at least try to explain your reasoning most of the time.

[hilzoy]

Fwiw, my understanding on the census thingo was basically this:

(a) Republicans and Democrats have a long-standing disagreement on census methodology. Democrats believe that the census seriously undercounts people in cities who are homeless or change address a lot. Republicans disagree. This is very much like the disagreement on voter fraud. Insofar as I have followed this debate, I think that it's an issue that genuinely has two sides, though I think the Democrats have the better argument, which you might take as bias, even though it seems to me (as no doubt it would) that I have tried to weigh the arguments evenhandedly.

(b) Given this history, and given that Gregg in particular opposes the Dem. position, and has moreover tried to block census funding on a number of occasions, and given that the census is a politically fraught issue, it would have been a very hard sell to give him control of the census. (Imagine, for the sake of comparison, Obama putting a Republican/ Bush putting Democrat in charge of designing policies to combat voter fraud nationwide.

(c) Thus, the decision to have the person put in charge of the census report to the WH. That person would have been nominated anyways, would have to be confirmed anyways, etc. In a lot of ways, nothing would have changed. Gregg would just not have been in charge of him, and that matters to a lot of Democrats.

It matters to know that people on both sides are sincere about this. (I mean, obviously not all of them, but neither side has ludicrous views.) To a lot of Democrats, it really does seem as though doing the census as the Republicans want will undercount a whole lot of people, for the purpose of drawing congressional districts, allocating various kinds of funds, etc.

(b) Gregg

[hilzoy]

I should say: one of the points I was trying to make with my last comment is: I don't think this is about one party trying to take control of the census for nefarious political reasons. I wouldn't have thought that about GHW Bush, had he done something similar; I might have thought it about GWBush, but only because Karl Rove would probably have been involved, and I think Karl Rove does almost everything for such reasons.

It's because there is a genuine disagreement that tends to sort along party lines. No doubt part of the reason it does so is that each party tends to favor the view that would advance its interests. But I really don't think that's all of it, at all.

The difference, basically, comes down to this: given that there are a number of people whom a count is likely to miss, and given the massive costs of employing not just enough people to go around with the census forms, but the further number of people who would be required in order e.g. to track down the homeless, what to do? The two answers are: (Republicans) just count the actual people you can count, and (Democrats) pick certain representative areas, hire those extra people, and extrapolate. (Note that there are quite sophisticated ways of extrapolating, making sure you've got your samples right, etc., which are used in a whole lot of non-political areas. It would not be as though Democrats got to decide what counts as a good sample, or a decent extrapolation, de novo.)

Accepting the Democrats' answer requires some familiarity with, and acceptance of, statistical techniques. One reason (surely not the only one, I don't mean to suggest that) would be serious distrust of those techniques, either because you don't understand them, or because you just don't care for them.

I don't think I'd want to rest a whole lot of weight on this, but I think that the frame of mind in which one might think: if you can't find someone, just don't count him, whatever you might find in "similar" areas -- only actual literal people, not statistical extrapolations, get counted in our census -- might be, for non-census-related reasons, more common among Republicans.

(For instance, it seems to me to be kin to the attitude that gives us: the literal meaning of the Bible is all that counts; airy-fairy interpretations that "explain" why God did not actually mean X when the Bible says X are just attempts to wriggle out of stuff. -- Back when I was Christian, and used to argue theology with people, I encountered fundamentalists who seemed to think that making an argument like this even about what seemed to me to be obvious cases would just be the entering edge of the wedge, and would lead inexorably to "interpreting away" the Bible's plain message that e.g. God exists and we should worship him.

(Example of what seemed to me an obvious case: the writer of a Psalm, which is plainly not written in the voice of God, recommends something; does it follow that God is in favor of it? Is there nothing that could show that, in a given case, He does not? Even in the case of Psalm 137 ("Blessed be he who takes your children and bashes their heads against the rocks"), in which taking the Psalms as literal guides to what God likes would not just be dreadful, but would also conflict with other things in the Bible? The idea that in virtue of taking the Bible very seriously, I was somehow committed to accepting that always struck me as just bizarre -- some parts of the Bible are plainly written in the voices of people other than God -- this is certainly true in the case of those parts that are explicitly attributed to other people (when the Bible says "Judas said X", it would be nuts to infer that you had to accept X, since the Bible says X, ignoring the "Judas said" part), and I couldn't see why it should not also be true of e.g. the Psalms.

Truly, though, the fundamentalists I argued seemed to think that accepting this was a horrible, horrible danger, and that if they accepted even this, some sort of sweeping textual nihilism loomed.

As I said, I think this is at least a relative of the view I cited about statistical methods.)

[liberal japonicus]

Just to fill in that dangling (b), I thought this, via 538, was interesting

According to Emanuel, Gregg's decision was not unexpected -- Gregg had called Emanuel to express "second thoughts" about the post three days ago. "You could tell by Monday, it was the sense that, you know, he must've over the weekend been kind of noodling on something and he just kind of became uncomfortable with what this was."

During the initial conversation with Emanuel, Gregg requested to speak to the President. Obama and Gregg met yesterday at the White House, but it was unclear whether that was where the final decision was made.

Gregg had wanted to wait on the announcement, until possibly after the stimulus bill had passed, but his withdrawal had leaked to a few of his Republican Senate colleagues this afternoon and he felt he couldn't wait.

When pressed about the seeming surprise in the White House press office about the timing of the announcement, Emanuel remained gracious, in contrast to the blunt release put out earlier. "I think you've got to take what (Gregg) said at his word."

"There's not hard feelings. I want to repeat it to you," said Emanuel.

Emanuel said the idea for Gregg as Commerce Secretary had come to the White House through Harry Reid, that it was not the White House's original idea.

Moreover, while Emanuel didn't want "to play psychologist, to get into (Gregg's) head," he insisted that the Census was not the root cause. "Trust me, the Census was not" the issue, he insisted.

Emanuel admitted the situation with the withdrawal of a second Commerce Secretary, sandwiched around Tom Daschle's aborted bid for HHS Secretary, was "unfortunate" and "disappointing" for the White House, but that was as far as he went, and continued to strike a gracious tone.

"I think (Gregg's) a very principled person. And he takes his role seriously. I believe he seriously wanted to do the Commerce job, he went into it eyes open, and realized after awhile it was just not going to be, as he said, the right fit. Full stop. I don’t think you can go back from that."

[OCSteve]

John: So I guess what I want to say, is if you have a problem with what is happening, don’t just throw a charge out there and walk away. Defend the accusation. You have a voice most of here want to hear, even if we disagree with you. But it is because you at least try to explain your reasoning most of the time.

I think you’ve nailed it. I’m kind of lost and fumbling around here. Politically I really don’t have a home. I’m happy and pissed and everything in between. I’m lost. I don’t know where the hell I am… I admit that I am flailing out at you guys. Sorry…

[hilzoy]

People: I have a request.

I think it is always a good idea to bear in mind that people you know re not just this or that remark; that you generally know a whole lot more about them than you're reacting to at any given instance; and to act accordingly. This can, of course, cut either way: some people have such a track record of vileness that you would get suspicious if they said hello to you on the street. But by the same token, some people have more than earned the benefit of the doubt.

I think OCSteve has. He has been incredibly open to ideas he didn't start out accepting, and incredibly willing to stick up for his views but to admit when he's wrong.

I have no problem with anyone disagreeing with him, and I don't ask anyone to accept my view of him. But I do think it's worth remembering who he is, and what he's like -- and probably also what it's like to be one of the few conservatives commenting here. I used to talk about this with Andy: he would say something conservative, and even if everyone was completely polite and decent, the fact that 100 people made the same reasonable point in response felt kind of like being jumped on.

Neither of us could see any way around that. But it's worth bearing in mind.

[Ugh]

some people have more than earned the benefit of the doubt.

I think OCSteve has. He has been incredibly open to ideas he didn't start out accepting, and incredibly willing to stick up for his views but to admit when he's wrong.

Fnck that noise!

OCSteve -- you are a muslamonazibuudistfascisticliberconservo-hater-lover!!!1!!! Plus, I'm sure you have B.O. and like the Cowboys, or something. Mumble mumble... Also, you said you're for repealing DADT, which means you hate the gay people and are gay and also are against repealing DADT!!

And muslim stuff.

[OCSteve]

Thanks Hilzoy. But I think I may be just lashing out. I don’t know why… Phil and Nell and John and others are right. I’m not contributing anything useful here and I’m just generally being a pain in the ass. It’s dumb and what the hell is the point?

I’ve strived not to do drive-by’s – but lately it seems that’s all I do…

Apologies to all and it wasn’t my intent to be disruptive. I just have a knack for that I guess. ;)

[hilzoy]

OCS: "I’m not contributing anything useful here "

I disagree. With all due respect, I disagree strongly.

[Nell]

@OCSteve:

I've been very harsh lately, and especially on you. Please don't go away, at least not for long.

If it's any comfort, I'm flailing too -- angry and occasionally happy at what's happening, but more afraid than anything else.

What the people in this administration and Congress are willing to do isn't going to be anywhere near the kind of change that's needed. It's so close to too late on so many fronts -- climate change, Israel/Palestine, Afghanistan/Pakistan, the hollowed-out house of cards economy -- that it's more than usually hard to have sympathy and patience for people for whom what's happening is way too much.

[publius]

Steve - i hope you stay to. I have this theory that people are often hostile on paper, but genuinely enjoy the exchange. for instance, i may snark a lot in comments, but i actually really enjoy reading pushback b/c that's what helps me think sharper.

now all that said, people should be respectful. but i think people genuinely appreciate having the diversity of views in the comments (and would be disappointed if it stopped)

[Gary Farber]

Gary – nice to see you commenting. I was beginning to worry about your absence."

I've been suffering a tremendous amount of this, along with the pain keeping me unable to sleep despite Ambien, along with also some gout pain and some toothache pain; it's a trifecta.

And the clinic I've been trying to get seen at for more than six months called again on Friday to tell me that my long awaited appointment for Tuesday was cancelled, as they wouldn't have any doctors in, as it turns out.

Prior to that, I was distracted by Facebook for a couple of weeks, but now that my account has been mysteriously disabled, that's not an issue at the moment. And then I had a problem with my virus protection program for some days, keeping the internet completely inaccessible unless I completely deinstalled it. And now, tonight, I just mostly can't seem to connect to the internet at all. I spent an hour trying to get a prior comment about torture to be posted, and it got eaten, and I give up trying to reconstruct it, especially since I still can't seem to connect to most anywhere on the internet tonight. No idea if I can get this posted.

[Brett Bellmore]

"Alternately, the Senate might adopt a rule that said that during filibusters, if a quorum was not present, the Senator who was speaking could decide to go on speaking or to allow a vote on cloture, to be decided by a majority of those present and voting."

Wouldn't that constitute a violation of the quorum clause, by demanding that the Senate conduct business without a quorum present? Normally the Senate and House circumvent that Constitutional requirement by avoiding any action which might reveal the absence of a quorum, (That's why they have so many voice votes: Generally a voice vote indicates they don't have a quorum present to do business, but are doing it anyway.) but your proposal would pretty much demand that they put the absence of a quorum on record, and then act despite the constitutional demand that they have a quorum present to conduct business.

[Sebastian]

"Republicans and Democrats have a long-standing disagreement on census methodology. Democrats believe that the census seriously undercounts people in cities who are homeless or change address a lot."

The problem isn't so much that there is an undercount, it is what Democrats want to do to 'remedy' the undercount. They want to use statistical sampling methods with disturbingly high error potential to estimate the count. If the main thrust was to use improved techniques to actually find and actually count the people, that would be fine as that is just refining the normal operations of the census. But the Democrats want to transform the census from its historical actual count to a statistical approximation of a count. This of course introduces the possibility of all sorts of political gamesmanship in selecting the sampling method, controling the statistical weights, and shaping the calculation heuristics that aren't available in a straight up physical count of human beings.

Opening up the census to more manipulation a la gerrymandering (which is a political evil that no US party while in power cares about) strikes me as profoundly unwise. Any attempt to do so should be done very much above board, and with a super-majority.

Now if we want to use statistical methods to estimate homelessness, or transient people, or other hard to count people *for purposes other than that of assigning a number of representatives* that might be a different thing entirely, because while it would still be open to political manipulation, the stakes for cheating wouldn't be as high.

[KCinDC]

Sebastian, what you're saying is that you're willing to accept the certainty of an undercount in order to avoid the possibility of an overcount.

[Sebastian]

I'm willing to accept the relative certainty/difficulty of gaming a physical count to avoid introducing yet another politically gameable element into the system.

If you want to actually reduce undercounting by beefing up the physical counting side, I'm all for it. If you want to theoretically reduce undercounting by introducing a whole new venue for political gameplaying and manipulation, I'm not.

[John Miller]

KCinDC, This is so different (as hilzoy alludes to) than the wole voter fraud issue.

On the surface, at least, it would appear the difference is that Republicans would rather a legitimate voter not be allowed to vote rather than an illegitimate voter get to vote. Democrats, on the other hand, are willing to take the risk that an illegitimate voter be allowed to vote in order to avoid disenfranchisement of eligible voters.

There are actually arguments for both sides, although I definitely fall into the latter camp.

[Brett Bellmore]

"If you want to actually reduce undercounting by beefing up the physical counting side, I'm all for it. If you want to theoretically reduce undercounting by introducing a whole new venue for political gameplaying and manipulation, I'm not. "

Yeah, that pretty much sums up my concern: While it's certainly possible to apply statistical techniques to enhance the accuracy of the census, "The census takers didn't visit neighborhood X" is a heck of a lot simpler and easier to prove than" "The census adjusters used insert mind blowingly complex techno-babble where they should have used insert different mind blowingly complex techno-babble."

The concern here is that, with statistical sampling, you could rig the count in a way you could never explain to the average Joe. And so, you'd get away with it.

And the other concern is that it says "enumeration", and that means counting, blast it. Not estimating.

Double the census budget, and forget statistical sampling.

[jdkbrown]

"But the Democrats want to transform the census from its historical actual count to a statistical approximation of a count."

Correction: Some Democrats, in accord with a recommendation of the National Academy of Sciences, want to transform the census from an actual count of only some of residents of the U.S. to a more accurate assessment of total residency.

[hilzoy]

See what I mean? *smiles*

[Catsy]

OT, but the funny thing is that the Census debate is one on which I have sympathy for both (generally defined) sides of the argument.

Hilzoy made an observation that I think is particularly insightful here: the disagreement between liberals and conservatives about the Census can be traced back to the same general difference of principles as "voter fraud" (scare quotes used advisedly here)--and a score of other civil rights and law enforcement issues.

Consider: these issues fundamentally draw from how each ideology weighs the conflict between punishing the guilty and preventing crime vs. protecting individual rights and avoiding the punishment of the innocent. I am going to generalize a bit here, but I think that by and large these are accurate statements.

On voting fraud:
- Liberals: prioritize the assumption of legitimacy and the protection of individual franchise, accepting that some fraud will occur as a result of erring on the side of preserving franchise.
- Conservatives: prioritize fraud prevention and the strict scrutiny of legitimacy, accepting that some disenfranchisement of individuals will occur as a result of this vigilance.

On the Census:
- L: prioritize the goal of accounting for those who are underrepresented by traditional counting methods, accepting some loss of precision in exchange for improving representation of the disadvantaged.
- C: prioritize the goal of only counting those who can be explicitly counted and verified, accepting that some will go uncounted and unrepresented in exchange for precision.

On law enforcement:
- L: prioritize avoiding the punishment of the innocent or excesses in punishing the guilty, accepting some loss of deterrent and that some guilty will go free in exchange for preserving individual liberty.
- C: prioritize deterrence through punitive measures and punishment of the guilty, accepting some collateral damage to the innocent in exchange for the preservation of public safety and order.

You will find this line of thought running through many of the places where liberals and conservatives actually disagree in good faith about policy issues that do not have an easy right/wrong answer.

[John]

Consider: these issues fundamentally draw from how each ideology weighs the conflict between punishing the guilty and preventing crime vs. protecting individual rights and avoiding the punishment of the innocent.

Seriously? And this just happens to track with the politics of it?

Republicans want a strict count because it helps them politically. Democrats want to use statistical sampling because it helps them. They can gussy it up with ideological justification, but that's the basic issue. Arguments about process are always arguments about substance.

[Catsy]

Seriously? And this just happens to track with the politics of it?

Yes. That's not to say that political considerations don't enter into it, but the consistency of the ideological underpinnings between issues like this is hard to dismiss out of hand.

Republicans want a strict count because it helps them politically. Democrats want to use statistical sampling because it helps them.

While there is some truth here, these blanket assertions are far too simplistic, bordering on reduction to absurdity. Human motivation is a complex system.

Let's walk it through. Sampling would have the effect of improving representation of disadvantaged demographics. These demographics lean Democratic. They lean Democratic because, broadly speaking, liberal priorities and solutions tend to favor the disadvantaged more than do conservative ideas. Liberal ideas tend to, by design, favor the disadvantaged because urban and low-income individuals are disproportionately affected by "the conflict between punishing the guilty and preventing crime vs. protecting individual rights and avoiding the punishment of the innocent".

These things are all connected. It's not as simple as saying this side favors X because X has political benefits for them. There are reasons why this is so. Many people enjoy drinking wine. Wine is alleged to have certain health benefits. Believing that it has health benefits may make someone who already likes wine more comfortable with their tastes, or more likely to have some, but even though it may factor into their motivations, it's not accurate to say they're drinking wine because it's good for them if they already liked it and would drink it anyway.

[Phil]

I don’t know why… Phil and Nell and John and others are right. I’m not contributing anything useful here and I’m just generally being a pain in the ass.

Not only did I not say that, I don't think it. It's frustrating for me when I know you've seen well-reasoned arguments that flatly contradict shibboleths that you keep repeating, and act as if you haven't seen them. It's like, now that the Democrats actually are in power, you've reverted to GOP groupthink (on some topics) that you know doesn't cut the mustard, but can't seem to help it.

Sebastian:The problem isn't so much that there is an undercount, it is what Democrats want to do to 'remedy' the undercount. They want to use statistical sampling methods with disturbingly high error potential to estimate the count.

Can you break this out a bit? What sampling methods are we talking about, what are the error bars, and why are they "disturbingly high?" Especially in comparison to whatever the error bars are for the currently-practiced actual counting method?

LOL at Brett's "Double the census budget." Why should we give the Mafia any more money to bring more citizens under their thumb?

[Sebastian]

Catsy your system is interesting. It fits well with the backlash against the liberal law enforcement ideas of the 1960s and 1970s. Not only did you lose the conservative-types off the bat, but many of the more Democratic-party oriented communities saw that changes in the "punishing the guilty" part of the equation were hurting them with dramatically higher crime rates.

The census is an interesting interaction case. Democrats seem to want to count 'uncountable' census people because they believe there are more of them in cities. This would result in city-dense states getting more representatives. This is not however likely to cause more VOTERS to be represented, because voters are not generally among the uncountable people we are talking about. If they were voting, you could for the most part find them and count them.

The voter fraud case is interesting too. My propensity is mostly on the conservative side as you probably know. But the history of investigation in the last few years (see especially the recent Supreme Court case on the Indiana voter ID law and the Bush era investigation into voter fraud) suggests that voter ID laws don't particularly disenfranchise (despite huge motivation to find people disenfranchised, the people opposing the Indiana law couldn't find such people) and that not much wrongful voting from people without voting rights was happening (despite huge motivation to find such people the Bush-era law enforcement could find only a handful of such cases).

But for the most part, people in the debate choose to only recognize exactly one of those two facts. It is as if we want to cling so much to our ideological priors that when it turns out we are both right (such fraud happens at such a low rate and such disenfranchisement is nearly non existant that taking either path turns out to have virtually no consequence) we still can't accept it.

[Gary Farber]

"It fits well with the backlash against the liberal law enforcement ideas of the 1960s and 1970s. Not only did you lose the conservative-types off the bat, but many of the more Democratic-party oriented communities saw that changes in the 'punishing the guilty' part of the equation were hurting them with dramatically higher crime rates."

I'd like to know more specifics about this, Sebastian: can you specify some of the "liberal law enforcement ideas" you have in mind, and how, exactly, they led to "dramatically higher crime rates"?

Thanks!

[Sebastian]

Sure, more lenient punishments for violent crimes and a rehabilitation concept that permeated the organizing concepts in the 1970s and 1980s, especially in big cities (See New York and LA) reversed by the late 1980s.

That is a good start, and you're an excellent researcher, so I'm sure you can find more.

You're Welcome!

[liberal japonicus]

liberal law enforcement ideas of the 1960s and 1970s

I think there is a bit of historical context involved in that period.

[Jeff]

I just wasn’t happy about how the biggest spending bill in history went down…

I wasn't either, but blame the Repubs, who i think behaved like slime. Could you clarify why you weren't happy?

[Gary Farber]

Ok, no specifics from Sebastian. I hope you'll forgive me, then, Seb, for not stipulating to your assertion that these things happened.

That may just be me, of course.

[russell]

I'm willing to accept the relative certainty/difficulty of gaming a physical count to avoid introducing yet another politically gameable element into the system.

How is a physical count not gameable?

I'm with Brett on this. If the resources we currently apply to the census aren't getting it done, then bring the necessary resources to bear. Get it done, and move on.

But I don't see either a statistical or a physical count method as being more or less prone to gaming.

[ral]

Slightly OT, but since we're discussing crime rate trends (and statistics), I feel I should point to the Nevin study showing the correlation with lead levels. It's striking.

Excerpts here (the full study is behind a pay wall).

[Sebastian]

"I hope you'll forgive me, then, Seb, for not stipulating to your assertion that these things happened."

You aren't going to do the research on this well understood social history but instead want me to? When I've done so at length in the past?

You're forgiven. I wasn't asking anything whatsoever of you in any case.

Thanks!

[Sebastian]

"But I don't see either a statistical or a physical count method as being more or less prone to gaming."

The physical count is perhaps subject to gaming, but in ways that the average layman can understand when it is pointed out, so it is harder to get away with. Kind of like how gerrymandering was obvious in the old days, but doesn't appear to be so bad nowadays if you look at the maps because the computers can make it more compact with the same effect.

[Gary Farber]

"You aren't going to do the research on this well understood social history but instead want me to?"

No, I'm not going to try to mindread you and find I have no idea what evidence for I don't know what assertions. If you don't care to support your own assertions, hey, fine.

"When I've done so at length in the past?"

I guess you want me to research that, too, rather than offer links.

Maybe lots of people find this sort of argument persuasive: good luck with that.

[Bernard Yomtov]

But I don't see either a statistical or a physical count method as being more or less prone to gaming.

I don't think the physical count is so much subject to gaming, except through inadequate funding, as it is inherently biased low because some people are easy to count and others are not. There is not only error, but the error is in a particular direction.

Statistical methods try to correct for this bias. Sebastian and others complain that this has large errors. Phil's answer - that the issue is whether the results are more accurate with sampling than without - makes sense.

A second point is that reducing bias may justify increasing error. Suppose a count says that a certain area has a population of 90,000, but because of bias the actual population may be anywhere from 90,000 to 100,000. Now suppose we apply sampling and get an estimate of 98,000, with a confidence interval of 91,000-105,000.

For purposes of setting political representation, I think the sampling result is better, because while it may not be as accurate as we would like, it does not sytematically discriminate against hard-to-count populations, and is not known to be too low.

In other words, we should often prefer an unbiased count with large error over a biased one with smaller error.

[Sebastian]

"In other words, we should often prefer an unbiased count with large error over a biased one with smaller error."

But what does an 'unbiased count' have to do with it?

The statistical methodology is virtually certain to biased in one way or another. Once you get the poltical process involved it most definitely isn't going to biased in purely accidental ways. You can't assume away political influence when the stakes are going to be so high. You basically want to assume political influence on the physical count while denying the likelyhood of political influence on the statistical methodology (which will almost certainly be tweaked with on a county by county level or smaller). That just doesn't strike me as a realistic assumption. And it introduces bias in ways that hard to explain to voters. They can easily understand "not trying hard enough to count workers that move a lot" and can imagine possible ways to fix it. They cannot easily understand a 35 variable-weight, differs by each county, sampling 'error' correction method even if perfectly executed. It is like begging for conspiracy theorists to spread plausible rumors. And that is IF it is done without bias. If it is done more like gerrymandering (and don't even try to pretend that Democrats don't engage in vicious gerrymandering), it adds a huge potential for abuse.

Why do that, instead of using much more straight forward actual counting techniques?

[Catsy]

Seb: Catsy your system is interesting. It fits well with the backlash against the liberal law enforcement ideas of the 1960s and 1970s. Not only did you lose the conservative-types off the bat, but many of the more Democratic-party oriented communities saw that changes in the "punishing the guilty" part of the equation were hurting them with dramatically higher crime rates.

Like Gary I'm not sure I'd like to stipulate to some of your premises, but I don't think Gary got past that in order to understand that whether or not the facts you assert are accurate is beside the point--the point is that yours is an accurate description of how many people perceived the situation and how those perceptions drove their party alignments based on their priorities.

But for the most part, people in the debate choose to only recognize exactly one of those two facts. It is as if we want to cling so much to our ideological priors that when it turns out we are both right (such fraud happens at such a low rate and such disenfranchisement is nearly non existant that taking either path turns out to have virtually no consequence) we still can't accept it.

I see two things at work here.

The first is that you're absolutely correct about the propensity of people to cling to their ideological shibboleths and premises, sometimes in the face of evidence. It's not true of everyone, of course, but it's pretty common in politics.

The second is that if you think about it, what you said comports pretty well with the conservative-liberal dichotomy I've been trying to articulate. Consider: accept for the sake of argument that conservatives are right about voter ID laws having very little disenfranchising effect, and that liberals are correct that actual incidences of voting fraud are extremely low.

Despite these being the stipulated facts, for conservatives to continue to rail against voting fraud would suggest that even though the incidence of both harms is low, they still consider the prevention of small amounts of voting fraud to outweigh the small amount of harm to franchise that might result. Correspondingly, the liberal in this equation examines the low incidence of voting fraud and the low incidence of disenfranchisement, and judges that preventing that little voting fraud is not worth any potential loss of franchise. Both parties remain in their respective corners.

I don't know that most voters necessarily think this way, and I'm sure that most of them don't actually reason through it or identify their priorities as such. I think it's more likely that the balance between these competing priorities develops differently in each person and helps drive their identification with a given party, that party's alignment of priorities being well-defined by both tradition and by those who actually do reason through issues based on principles they can articulate.

I think Bernard's summary here nicely encapsulates the liberal end of this spectrum of priorities with regard to the Census:

Bernard: In other words, we should often prefer an unbiased count with large error over a biased one with smaller error.

[Jesurgislac]

Phil: Not only did I not say that, I don't think it. It's frustrating for me when I know you've seen well-reasoned arguments that flatly contradict shibboleths that you keep repeating, and act as if you haven't seen them. It's like, now that the Democrats actually are in power, you've reverted to GOP groupthink (on some topics) that you know doesn't cut the mustard, but can't seem to help it.

I agree with Phil.

[Gary Farber]

"Like Gary I'm not sure I'd like to stipulate to some of your premises, but I don't think Gary got past that in order to understand that whether or not the facts you assert are accurate is beside the point--the point is that yours is an accurate description of how many people perceived the situation and how those perceptions drove their party alignments based on their priorities."

Quite right; of course lots of people -- people who were therefore persuaded to vote for Richard Nixon, and later Ronald Reagan -- perceived "the crime problem" as Sebastian describes.

I'm far more interested in establishing how right or wrong they were. On that front, I'd love to see Sebastian read and review Perlstein's Nixonland.

[Bernard Yomtov]

You basically want to assume political influence on the physical count while denying the likelyhood of political influence on the statistical methodology (which will almost certainly be tweaked with on a county by county level or smaller). That just doesn't strike me as a realistic assumption.

I'm not assuming that the bias in the counting is due to political influence, but rather to the inherent difficulty of counting some people. I think the bias is an inherent part of the process. It varies according to levels of funding and effort, of course, but it's a fact of life. Just to be absolutely clear, by "bias" I mean not just error, but error in a known direction - undercounting in this case.

And it introduces bias in ways that hard to explain to voters. They can easily understand "not trying hard enough to count workers that move a lot" and can imagine possible ways to fix it. They cannot easily understand a 35 variable-weight, differs by each county, sampling 'error' correction method even if perfectly executed. It is like begging for conspiracy theorists to spread plausible rumors.

Statistical methods won't be perfect, but I'm inclined to think that a sensible objective approach can be devised by serious professional statistical mavens. No doubt the Limbaughs of the world and other know-nothings will scream, but I refuse to accept that we have to cater to ignorance.

And that is IF it is done without bias. If it is done more like gerrymandering (and don't even try to pretend that Democrats don't engage in vicious gerrymandering), it adds a huge potential for abuse.

Democrats? Gerrymander? Never happen.

Why do that, instead of using much more straight forward actual counting techniques?

Because it is more efficient, and given various battles over census funding it is hard to believe that the money will be there to do a count that is more accurate than a statistical estimate.