Back in 2003, Ayres and Donohue found some coding errors in Lott's "More Guns, Less Crime" data. They found that if you corrected his errors, Lott's results went away. Lott's reaction to this? Well, for four months he refused to admit to the existence of the errors. When he finally admitted to the errors, he changed his model to bring back his results, making a clumsy effort to try to hide the changes he made.
Fast forward to 2005. Now Foote and Goetz have found a coding error in a paper by Donohue and Levitt. The Economist reports:
But Messrs Foote and Goetz have inspected the authors' computer code and found the controls missing. In other words, Messrs Donohue and Levitt did not run the test they thought they had—an "inadvertent but serious computer programming error", according to Messrs Foote and Goetz.
Unlike Lott, they immediately admitted making the error. Nor did correcting the error make their results go away, though it did reduce the size of the effect. The Economist article has more on where that leaves their thesis, but what I find very interesting is Lott's reaction to all this. He writes:
Personally, I think calling this a "programming oversight" is being much too nice. More importantly everyone who works with panel data knows that you use fixed effects.
Yes, he's implying that they deliberately cooked their results. I think this tells us more about Lott's approach to econometrics than it does about that of Donohue and Levitt. He must think that everyone else operates like he does.
There's a much bigger lesson here. It's about the failure of the economics profession to do simple reality checks on complex econometric modeling.
Back in my debate with Levitt in Slate.com in 1999 over his abortion-cut-crime theory, I showed his theory failed various simple reality checks: e.g., nationally, homicide had shot up among the first cohort born after legalizaton.
I also showed that his assumptions about human behavior were simplistic and likely inaccurate:
http://www.slate.com/id/33569/entry/33571/
I think if you read it now in light of this week's news, you'll find it very interesting.
Levitt's response was to re-assert that his black box econometric work proved his theory. I'm not an econometrist, so I couldn't disprove that. Now, two econometrists have finally redone his work and disproved it, which strongly suggests that my reality checks should have been taken seriously all along.
Instead, much of the public policy world whooped up his abortion-cut-crime theory, making it Instant Conventional Wisdom. I doubt if this week's debunking of Levitt's theory will ever catch up with the popular impression that some smart guy PROVED abortion cuts crime. "A lie is halfway around the world before the truth gets its boots on."
So, I believe economists should study this entire fiasco to learn how to avoid being duped by black box analyses like Levitt's in the future.
On the contrary, Lott's record of doing excellent, unbiased work in the field of econometrics (in addition to his nobel-worthy work elsewhere) provides him with the sort of stature and insight that allow him to perceive when sad little biased researchers who wish they were John Lott falsify their data.
In fact, this is exactly the sort of karmic retribution that I imagine John has been expecting ever since these rabidly anti-gun mudslingers began attacking his carefully-researched work.
I am not an expert by any means, but I know one when I read one, and after reading all of John's excellent books I can say that I trust his judgememt implicitly.
That's an interesting way to characterise the National Academy of Sciences.
"Sok Pupit"
Humm, someone's having a joke here!
More importantly everyone who works with panel data knows that you use fixed effects
Except, of course, when you use random effects. The more Lott writes about econometrics (particularly that ghastly business with the tobit model) the more I think he doesn't really understand what he's doing at all; he's just got the STATA handbook and throws randomly selected chapters of it at whatever dataset comes around.
FWIW, as far as I can tell from the working paper, the programming errors don't actually change the results all that much; the conclusions are still there in the full specification. What does get rid of the L&D results is moving from absolute numbers of arrets to per capita numbers of arrests, assuming that F&G's population data is reliable. I've commented on my own blog that I think it's a bit poor of Levitt to pretend that the whole issue is about the programming errors, but that doesn't mean I can't say the same thing about Lott.
You want more evidence that Lott doesn't understand what he is doing, here's an old exchange we had on Usenet:
Me:
To be found, of course, halfway down this post (for those of us who have not yet had our "Duh!" moment today).
The most depressing thing about dedicated idiots like Luskin, Blair, Lott and Fumento is that you cannot hold a real discussion on line with clueful folks lest the dreck pick on any disagreement as evidence of malfeasance.
Its good to see that Tim Lambert (TL) is honest enough to now acknowledge that Levitt has some clay on his feet. There were others who were dubious of Levitt right from the get-go. It would be nice if TL acknowledged that those Levitt-skeptics were onto something real. In the past TL seems to have been more concerned with debunking the Levitt-skpetics rather than debunking Levitt.
The Levitt-skeptics were in the minority and in striking contrast to the elite members of the social science and pundit community. This lot were generally uncritical towards Levitt's headline claim and is still in a state of denial about the how easily they swallowed it.
There are some problems with TL's comments. Tim Lambert says: Dec 3rd, 2005 at 2:47 am
He is incorrect to say that Foote & Goetze's (F&G) critique merely weakened Levitt's conclusions. In fact, they said that Levitt's abortion-crime cutting effect, when properly tested, disappears. This quote is lifted from the abstract of their paper:
This is clear enough. Why bother to try and salvage something from the Levitt wreckage?
The Economist has to go to, of all places, Romania - the basket case of Europe in transition - to find some inverse relation b/w abortion and crime. Would it not have been proper to look towards representative countries in normal times?
Most moralists have associated increased abortion and increased crime with a more general decline in individual moral standards. For example, abortion was illegal in the US during its rather straight-laced Roosevelt-Eisenhower period. Yet the crime rate was much lower then.
By contrast, abortion was legal in the USSR during the same period. Yet over this time the Russian Mafia got its claws into the state and was strong enough to form a black governement in Russia the moment the USSR collapsed.
This tells me that the relationship b/w abortion and crime is much to weak to warrant Levitt attributing it so much causal power. And that is before we mention the Crack Wars, which Levitt swept under the carpet via an obscure footnote.
It also tells me that traditional moralists - insufferable old bores, old wives talers and the like - can sometimes leave fancy hi-falutin' social science modellers for dead. Perhaps the scientists are succumbing to the temptation of contrarianism for its own sake?
Its a pity that TL also uses this post to drag out Lott for another belting. Lott is a bit of a soft target, in no small part due to TL's constant pummelling. But in this case Lott deserves some credit as he was one of the few social scientists who checked Levitt's propensity to go off half-cocked on this issue.
Maybe its time that TL started a detailed attack on hard targets with a heavy duty scientific track record, such as Levitt, Murray and Darwinian social scientists. That would do more justice to his talents than riddling soft targets, like wannabe scientists such as Lott & Fumento or bloggers such as Insta-pundit & Luskin, full of holes.
Eli Rabett Says: December 6th, 2005 at 8:12 am
Its the height of chutzpah for Eli to carry on as if he is one of the "clueful folks" given the clueless way he got carried away with Levitt's theories. On this blog I had a "real discussion" with Eli Rabet about the validity and veracity of Levitt's abortion cuts crime theory. He was a Levitt-booster, I was a Levitt-skeptic.
Eli was so confident of Levitt's large claims were valid that he likened Levitt's critics (myself included) to fools or knaves. Here is a sample of his rhetoric Says: May 29th, 2005 at 11:47 am
The Levitt-skeptics are not looking so dumb now. It appears that a sound analysis of the evidence proves that Levitt's theory is seriously flawed, if not fatally wounded.
So, Eli "baby you wanna stick your fingers in you ears" about the evidence that you were sucked in by celebrity marketing spin? Go ahead, but if so "don't expect me to do anything but laugh at you".
It appears that a sound analysis of the evidence proves that Levitt's theory is seriously flawed, if not fatally wounded.
I am curious. I have admittedly read little enough on the subject, so could you explain just how the analysis showed a serious flaw, or a fatal wounding?
Thanks...
Actually Jack, Eli Rabett never defended Levitt, he just pointed out that you were consistently and purposefully misstating Levitt's argument.
http://scienceblogs.com/deltoid/2005/05/abortionandcrime.php#comment-798
and he is doing it again. So let us repeat what set Jack off.
Having gone a few rounds there with Jack, I finally concluded with a quote from Scott Berkun which pretty much summarizes dealing with the Jacks of the world
This is but a small part of a great essay. Go read it http://www.scottberkun.com/essays/essay40.htm
Eli Rabett Says: December 8th, 2005 at 1:22 am
ELi Rabett is consistently and purposefully propagates falsehood both about Jack's arguments and about his own Levitt-boostering or Levitt-apologetics. This is shameful to him, unjust to Jack and a disservice to the intellectual community.
First off, Rabet is a Levitt supporter alright. He never ventures a criticism of Levitt and always bends over backward to interpret Levitt in a favourable way.
Rabet even buys Levitt's interpretation of the crack epidemic as a "short term ... factor that stimulate crime". (Hey Eli, I've got a real nice bridge on the East side of Manhattan that you might be interested in buying...)
For the nth time: the crack epidemic is crime, fer crissake! To treat a refuational fact as if it were a complicating circumstance is the definition of apologetics. (Rabet summarising modern cosmology to the Church: "Um,...Er...your geocentric theory is correct except for those complicating occasions when the earth spins around the sun.") Therefore Rabet is a Levitt-apologist, if not booster. So he is blowing smoke when he says he "never defended Levitt".
He also brazenly misrepresents the facts when he accuses me of "misstating Levitt's argument". In fact, far from travestying Levitt I bent over backwards to fairly construe his take on the Crack Wars. I dug out the quote that Rabet later tries to get so much mileage out in his Levitt-apologetics.
Jack Strocchi Says: May 28th, 2005 at 12:37 pm
I did not misstate Levitt's case. I state it verbatim to show how ludicrous it was. One does not need to resort to reductio ad absurdum when the statement is absurd on the face of it.
I dug this quote out to highlight Levitt's slippery style and to lampoon his ridiculous tactic of calling the crack epidemic a "complicating short-term factor" in crime rather than what any honest person would call it: crime. I was in the US at the time and I know what of I speak.
It is clear from the above that I have accurately quoted and summarised Levitt's argument, which amounts to first, redefining the Crack War crime wave as non-crime and second, re-defining a variables causal power as being sufficiently attenuated as to be unobservable. This is intellectually worthless. My characterisation is not a "misstatement", it is a judgement.
Lets take Levitt's escape clauses to pieces bit by bit, so that the slow learners can easily follow.
Levitt ackowledges that the Crack Wars are an "ostensible inconsistency" to his abortion cuts crime theory. He then redefines this refutational fact into a complicating "short term factor that increases crime". And the massive black economy and open gang warfare that came with crime wave are reduced to extranneous noise through the use of the innocuous soundng formula "the rise of crack and the easy availability of guns". Oh if only it was so easy to write off all problems in the world with a some cheap phrase making.
Levitt knows that the crack epidemic is deadly to his theory. Instead of admitting this and going back to the drawing board (and kissing his dreams of celebrity hood goodbye) he buries the incriminating evidence in an obscure footnote.
And then Levitt has the cheek to suggest that "caution [is] required in drawing any conclusions regarding an abortion-crime link based on time-series evidence alone." So Levitt showed great caution in marketing his abortion-cuts-crime headline grabber the wide world over? If this is Levitt caution I would hate to see Levitt recklessness.
What a contemptible series of ruses! And how many people were suckered by this transparent bit of intellectual legerdemain? Well Rabet for one.
Finally Rabet introduces an amazing new innovation into empirical social science: the reversible sign.
Translated into the methodology of social science this implies that a theory's independent variables (x1, x2...xn) may predict an increase or decrease in the dependent variable (y). So any observation is consistent with the verification of the theory.
This is great news for dummies because with this marvellous new methodology one can produce any amount of analysis and never be wrong! Nobel Prizes for every one all the time.
This methodological tactic amounts to an insulating barrier constructed by Levitt and his foolish defenders in order to protect his headline theory from any unwarranted damage by collission with brute facts.
Now that Rabet has "come face to face with someone who is tenacious enough to dissect [his] logic, and resilient enough to endure the thinly veiled intellectual abuse...[will he] "question [his] ability to defend bad ideas"? I am not holding my breath.
Hmm. Strochi again.
Go read the thread folks
http://scienceblogs.com/deltoid/2005/05/abortionandcrime.php#comment-798
The bunny was not the only one pointing out that Strochi was spinning. Our humble host himself mentioned the same point
As a matter of fact, as he said 25:
as did several others in the thread.
Eli Rabett Says: December 8th, 2005 at 11:46 am
Hmmm...Rabet again, up to his old tricks of relying on ad hominum argument, discredited authorities and selective quotation. This is discouraging as Rabet's failure to disclose cheats the reader of the full story.
He compounds the error by failing to record the fact that I refuted Tim Lambert's criticisms in the next comment. Jack Strocchi Says: May 27th, 2005 at 9:53 am FTR, I will reguritate my refutation, although it is tiresome to be forced to bang on about a point which is obviously been decided.
Says who? This would be the Levitt who has just been shown to be incorrect in his testing procedures? Levitt is most definitely not "correct when he says that you need to separate out the effects of the crack cocaine epidemic".
I work on common or garden definitions of public facts. In case Eli Rabet or Tim Lambert are unclear on this issue, let me remind them that trafficking in crack, and robbing and murdering people because of it, is still a crime. It is irrational and manifestly disingenuous to drop the most horrific criminal episode in recent US history down the memory hole for the professional convenience of an ambitious intellectual.
Since the Crack Wars were a criminal episode there is no good reason to "seperate them out" from the general crime rate. This would be like saying that one should seperate out the crimes of the Mafia in the twenties and thirties from the US crime rate as a whole on account of the fact that these crime were associated with extraordinary Prohibition.
Sailer's graphs are derived from figures collected by the US Department of Justice. They might be presumed to have some knowledge of crime rates and its causes. Their figures show that US violent crime was still around peak levels until the mid-nineties, a good 25 years after abortion was legalized.
As late as 1994 the overall violent crime victimization rate was still 51.2 per 1,000, just a fraction under the all-time crime peak of 52.3 achieved in the "Son of Sam" days of 1981.
As late as 1993 the homicide offending rate per 100,000 for young black males, the key variable that Levitt's abortion theory depends on, skyrocketed to 253.0 (14-17 yrs), 361.6 (18-24 yrs). These figures are several multiples of historic averages.
If abortion was supposed to weed out crime-prone boys from the population, so that the residual were likely to be better cared for and better behaved, then what happened to such people during the peak crime years in the early nineties?
It is ridiculuous for Tim Lambert to Eli Rabet to accuse me of denying that "Levitt failed to notice the upswing in crime in that cohort". In the course of this thread I have twice linked and commented on Levitt's unsatisfactory "notices". Several comments earlier I pointed out that Levitt noticed the problem with his theory, surreptitiously in his 2001 paper and conspicuously in his 2005 blog.
Jack Strocchi Says: May 26th, 2005 at 9:53 pm
I have also been industrious in digging out Levitt's sneaky footnotes. This is more than enough to satisfy any duty of disclosure.
I can't do much more in this topic by way of web research, data analysis or fallacy/fabrication refuting. If Levitt and his apologists want to insist that the Crack Wars were not genuine crimes or that sign-reversing evidence does not spell doom for a theory then they are beyond help. Their evasive attitude towards facts reminds me of those ultra-Darwinians who Stove chafed over their blinkered attitude towards inconvenient facts:
Jack, no-one has said that the Crack Wars weren't crimes. It is disappointing that you have so completely misunderstood what I wrote.
Yer gotta remember here Tim that Jack isn't so much interested in making a point but rather in admiring himself being seen to make a point.
I must say I don't understand why Levitt decided to deal with measurement error in the Alan Gutmacher Institute abortion figures by constructing an IV estimate using the CDC figures. In what possible world would the measurement error in the AG numbers be correlated with the error term but the CDC not be?
dsquared,
Levitt is absolutely correct here. The point I think you're misunderstanding is that it isn't the measurement error that's correlated with the error term, it's the AG figures (and the CDC figures) that are correlated with the error term.
That is, Levitt is talking about the type of measurement error where some nicely behaved noise is added to the true X variable. Maybe the AG institute had a computer generate a random number that they added to the data before reporting it. That's an optimistic assumption, but it's still enough to cause the OLS estimates to be inconsistent, and probably biased towards zero.
Specifically,
Let Y = a + bX + e, where X is the true measure.
Let Xa = X + u1 and Xc = X + u2, where Xc and Xa are two noisy measures of X.
Levitt is assuming that X, e, u1, and u2 are all independent of one another.
Given these assumptions, the following results are true. Regressing Y on Xa is an inconsistent estimate of b (because Xa is correlated with the error term). Similarly, regressing Y on Xc is also inconsistent. However, Xc is a valid instrument for Xa (and vice versa), and IV can yield consistent estimates of b.
Tim Lambert Says: December 8th, 2005 at 8:54 pm
It is pretty hard to construe Tim Lambert's talk of the Crack Wars as an "epidemic", the effects of which should be "seperated out" from the general crime [?], in any other way. But I would be happy to withdraw my criticism if this could somehow be clarified.
Still clarification is not going to save Levitt or his apologists. If the Crack Wars are treated as part of general US criminal history then the strong version of Levitt's theory, the one that grabs headlines and sells books, is falsified. Its that simple.
Nabakov Says: December 9th, 2005 at 12:19 am
If blogospherists relied on Nabakov's substantiated points, rather than jokes, for food then we would all die of starvation long before we died laughing.
Jack, there are many factors that affect crime. That means you can't just look at abortion and crime and ignore everything else (like the crack wars, for example). That's why you need a multivariate analysis that (hopefully) allows you to isolate the effects of abortion.
Hi jck stchy. Y my nt hv ntcd that crack consumption was an epidemic and the crack wars to control the traffic were crimes. An important difference. OTOH, you may not have.
Tim Lambert Says: December 9th, 2005 at 10:50 am
In the just quoted sentence I read Tim Lambert as saying that the "crack wars" are an "example...[of] everything else...that affect crime". But this can't be right since in a preceding comment Tim Lambert went out of his way to emphasise that "no-one has said that the Crack Wars weren't [a] crime" in itself.
Obviously these two comments are irreconcilable in logic. Or I need to adjust my browser.
Well we agree on something. I have never decried the presence of multivariate analysis in teasing out the effects of abortion on crime. I have decried the absence of common sense and straight thinking in identifying these effects. These are not inconsistent demands.
Here is a multiple of variates that might have had a significant affect on the crime rate in the nineties. They are at least as plausible crime prophylactics as abortion. And Levitt did not give them sufficient analytic attention.
That should keep the nerds busy.
Eli Rabett Says: December 9th, 2005 at 11:59 am
No, there is no real difference - so far as the criminal law is concerned - between breaking a law against crack consumption and breaking a law against crack distribution. Both are examples of a particularly vicious form of law-breaking. (If a crack house established itself next to Rabett's family kindergarten would he call the cops or an epidemiologist?)
Crack consumption is not an "epidemic", at least not in the biological sense of an infectious disease. There is no "crack pathogen". The use of crack is a conscious choice which many in "infected areas" refused to make at the time.
Persons with a literary imagination are free to use the term "epidemic" as a metaphor for rapid replication of behaviour patterns. But if they take that metaphor too seriously then scientists are then free to laugh at those people.
Likewise the "Crack Wars" were directly associated with crack usage. These crimes included the use of crack which is known to turn the user into a criminal psychopath, gangsta wars between rival crack supplier/dealers and the state's "War on Drugs" against such supplier/dealers. These are some of minor matters that Levitt managed to overlook.
On the contrary, I was in the US at the time (early 1993) when the Crack Wars peaked. No non-comatose person could have failed to notice the catastrophic breakdown of morality and legality associated with crack-usage and trade. Although Levitt and his apologists seem to have had a pretty good shot at it.
Rabett's moral compass appears to have gone completely haywire. He has the luxury of layers of social insulation with which to protect his intellectual vanity and ideological fantasies.
I would dearly love to take him back to those dim, dark days and then become a fly on a projects wall the moment after I dumped him deep in the heart of Bedford Stuyvesant. It would be amusing to if he took the opportunity to fall for the notion of a non-criminal way to engage in the crack economy.
Jack, the Donohue and Levitt paper is linked in my post. You might want to read it and then retract your false claims about them not considering the effect of ather factors.
The Crack Wars were crimes and they affected the crime rate. Why you think there is some sort of contradiction here escapes me.
ragout; of course you're right about IV estimation. Though I still think it's pretty odd to make the implicit assumption that the errors in CDC and AG estimates will be uncorrelated (and thus that u1 and u2 in your formulation are independent). Wouldn't this tend to introduce the opposite bias in the IV estimates (christ I am rusty on this, though I never really learned all that much about measurement errors since we economists usually assume our data is perfect).
I grew up in Brownsville Jock which, for your information was down market of Bedford-Stuyvesant. I've been in the heart of Bed-Sty many times. I also know the difference between an addict, an addict-dealer and a dealer. I have some appreciation of good and bad effects of the criminalization of drugs from personal observation (not use, my Dad worked on the Bowery in the 1930s which made him alcohol adverse).
I probably live and have lived a lot closer to a bunch of shelters, half-way houses and drug treatment facilities and even the occasional crack house (although they are disappearing as the neighborhood improves) than you do.
Actually, the neighborhood I live in is interesting in that we accept these facilities, we just don't want the city opening additional ones and overloading the area.
Now, of course, you can say I just made that up, and short of offering you a guided tour, which I am not interested in doing, I am not going to prove any of this to you. OTOH readers of this blog have some idea of your reliability, and my reliability and will judge for themselves as to who is likely to be telling the truth. Which is why your tactic of sticking a rake up your nose and then insisting that others are wearing it is self-defeating.
I also notice you are trying to change the subject again.
What you don't appear to get is that neither Tim or I or anyone else here who is watching you do the dervish dance is really invested in Levitt, but we want to discuss it on an honest basis, not by exaggerating his arguments to turn them into strawmen. You on the other hand appear to enjoy straw. That is hard to digest.
Try again.
Tim Lambert Says: December 9th, 2005 at 1:52 pm
Far from harbouring a contradiction this statement is a tautology, and one I would happily sign my name to. Unfortunately it is flatly inconsistent with Tim Lambert Says: May 26th, 2005 at 11:02 pm
If the criminal effects of the Crack Wars are "seperated out" from the crime rate then they are not being measured as crimes. Calling the Crack Wars an "epidemic" underlies the false notion of treating the Crack Wars purely as a health crisis rather than a crime spree.
Why Tim Lambert does not "think there is some sort of contradiction here escapes me".
Tim Lambert might want to take his own advice and "read...and...retract [his] false claims about [me] not considering" these matters. Far from ignoring or misrepresenting Levitt on various aspects of his theory, I have bent over backwards to quote, chapter and verse, and summarise his detailed views.
FTR, I have read, linked and quoted from the Donohue and Levitt 2001 paper on three occasions in the course of this debate.
Jack Strocchi Says: May 28th, 2005 at 12:37 pm [quoted footnote # 21]
Jack Strocchi Says: May 30th, 2005 at 3:58 am [quoted footnote # 8]
Jack Strocchi Says: December 8th, 2005 at 7:53 am
I have Jack Strocchi December 8th, 2005 at 6:44 pm linked to Levitt's 2005 blog, to highlight the one time he finally gave proper attention to the defects in his theory.
In these instances I went out of my way to emphasise that Levitt did address the "empirical problem" that the Crack Wars posed to his theory. Of course I also got stuck into Levitt for burying reservations about the theory in obscure footnotes, and being tardy in making a detailed reply.
I have never made the "false claim" that Levitt failed to "[consider] the effect of ather factors". In fact Jack Strocchi on December 9th, 2005 at 1:04 pm maintained that:
My own view, repeatedly stated, is that abortion may have acted as a weak "demographic prophylactic" to crime, but that is effects were no where near as strong as Levitt makes out in order to "sex up" his theory. The Crack Wars proved that criminal tendencies in the post-legal abortion generation were still extremely strong.
As far back as May 28th, 2005 at 12:37 pm Jack Strocchi agreed with Steve Sailer's conclusion:
My biggest grouch is the way that, normally hard-headed and fair-minded, critics of intellectual shonkiness were completely star-struck by Levitt's celebrity hood that they gave his dodgy theories a free-pass. Yet they had plenty of energy to turn their fury onto the miserable, low-life Levitt-skeptics. Who have now been vindicated! Where's the outrage?
Lets not forget the Big Picture here. Sailer's "moot" verdict on Levitt has now been vindicated by Foote and Goetze. The Levitt-apologists were wrong and the Levitt-skeptics were right. It would be nice if Lambert and Rabett had the honesty to admit this.
Eli Rabett Says: December 10th, 2005 at 3:30 am
Unlikely. I have for long periods of time made my own residence a "shelter, half-way house and drug treatment facility" for substance abusers in various states of disarray. It was this experience that has given me a life-long detestation of the hard-drug culture. Rabett may have walked near the wild side but I doubt he knows where its at.
Rabett implies that I am a liar. This is depressing as I have many times been tempted to hurl the same accusation at him. But I have forgone this guilty pleasure to maintain the assumption of good faith amongst disputants and to preserve a measure of civility.
Readers of this blog have a good measure of my reliability. This is my strong tendency, almost to the point of mania, to back up every statement I make with a specific fact or quote, invariably linked to source. And to do likewise to my disputants.
This trait is not in evidence in my disputants, which is prima facie evidence that they have not made proper efforts to disclose, are misrepresenting or evading the point. This calls into question their reliability, if not "truth telling" propensity.
This is the Rabett chutzpah grotesquely in evidence again. I think that Foote and Goetze have squarely planted the rake right up Rabett's nose. My Levitt-skepticism has been vindicated by their studies of the matter. His Levitt apologetics has been thoroughly debunked. One would search in vain for his honest admission of this fact.
Rabett is being disingenuous about his own position. It is true that Tim Lambert expressed reservations("Levitt may be wrong") about Levitt's conclusions. Although he had no reservations about (falsely) slamming Levitt's critics.
I cannot find a single instance in this debate where Rabett did anything else but find excuses for not acknowledging the weaknesses in Levitt's theory. This includes his his amazing methodologlical innovation to give Levitt a free pass: "The balance between 1 and 2 can be positive or negative". That is a licence to publish unfalsifiable theories if ever there was one.
My advice to Rabett is to link to specific areas where I have uttered (alleged) falsehoods. Quote actual words and cite true facts. It is what I do. It might force him to make comments on an "honest basis" for a change.