Law School Discussion

Nine Years of Discussion
;

Author Topic: Has ANyone Read NYT Best seller "Freakonomics" ? by Levitt and Dubner  (Read 2033 times)

TrojanChispas

  • Sr. Citizen
  • ****
  • Posts: 4702
  • , a worthy adversary
    • View Profile
haha aawsome.  Im gald this thread took off while i was away.

one thing that makes levitt's take on RvW so interesting is that he compared the crime rate drop with the relative legalization of abortion in different states.  Some states leagalized abortion 2 years earlier and saw a precipitous drop in crime two years earlier than the states that waited until the ruling in RvW... that adds to the weight and credibility of the argument

levitt also adresses all the causes that lie in conventional wisdom ie a strong economy.  Under his analysis, a strong economy had very lttle to do with the drop but increasing the police had a substantial effect.  RvW was not he only factor but a substantial one. 

here are my observations so far of the discussion:

hilljack is a rambling buffoon (good luck on the LSAT, HTH).

Everyone else should read this book because the abortion issue is a small part of the book.  he answers the question of why drug dealers live with their mothers (the top 2% make 50% of the profits.  the "dealers" or foot soldiers make between 3-7 dollars an hour), what the KK and real estate agents have in common (information asymetry over the public), and what teachers and sumo wrestlers have in common (they both cheat).  the whole book is fascinating
Arab Majority May Not Stay Forever Silent
http://www.nysun.com/article/36110?page_no=1

hilljack

  • Sr. Citizen
  • ****
  • Posts: 1669
    • View Profile
I don't know what your problem is 3peat.  You are obviously a very happy person who is very well adjusted; I don't want to screw with that.  I noticed you did nothing to counter any point I made, which is fine, you have no obligation to do so.  But when you just insult people in lieu of arguing their points, what does that prove?  Are you just lazy?  Surely you would like to counter the argument of someone for whom you have demonstrated a dislike?  You obviously don't even respect my intelligence or my reasoning or whatever, because you continually infer that I will do poorly on the LSAT; so shouldn’t it be quite easy to counter the points of someone as stupid as I am.  I don't know why you care about my LSAT; the only thing I hope is that you don't normally treat people with such disrespect.

haha aawsome. Im gald this thread took off while i was away.

one thing that makes levitt's take on RvW so interesting is that he compared the crime rate drop with the relative legalization of abortion in different states. Some states leagalized abortion 2 years earlier and saw a precipitous drop in crime two years earlier than the states that waited until the ruling in RvW... that adds to the weight and credibility of the argument

levitt also adresses all the causes that lie in conventional wisdom ie a strong economy. Under his analysis, a strong economy had very lttle to do with the drop but increasing the police had a substantial effect. RvW was not he only factor but a substantial one.

here are my observations so far of the discussion:

hilljack is a rambling buffoon (good luck on the LSAT, HTH).

Everyone else should read this book because the abortion issue is a small part of the book. he answers the question of why drug dealers live with their mothers (the top 2% make 50% of the profits. the "dealers" or foot soldiers make between 3-7 dollars an hour), what the KK and real estate agents have in common (information asymetry over the public), and what teachers and sumo wrestlers have in common (they both cheat). the whole book is fascinating

_BP_

  • Sr. Citizen
  • ****
  • Posts: 2594
  • Think. Wait. Fast.
    • View Profile
I never even said he was wrong, but just that people should consider the fact that making correlations is difficult, and it becomes more difficult when there is no control group.  Further, I thought that I made it clear that I would see great value in an evaluation immediately preceding and immediately following Roe v. Wade.  Usually, that is how this type of analysis works.  That is how we have studied government programs.  But even this does not account for unforeseen consequences, the other point of my contention.  We can all argue until we are blue in the face about the consequences of legalized abortion are.  I think it is easy to say that if nothing else, reducing the number of at-risk, as I have described them, births, has a negative relationship with crime.  But legalizing abortion and the abstract idea of lowering these birth rates is not the same thing.  If you make them equal, it is easier to say that legalized abortion leads to lower crime, no questions asked, but to make them equal is to ignore other effects.

I just ask people to be cautious.  I am a great believer in economics and weighing costs and benefits, but I am also a great believer in some scrutiny.  After all, how many times have two economists come to totally contradictory conclusions, one of them or both of them have to be wrong, the idea that this guy is so necessarily right is ridiculous.  He has provided us with a study, a theory or whatever.  But it is not price theory, it is not definite, so treat it as so at your own detriment.


In attempting to identify a link between legalized abortion and crime, we do not mean to suggest that such a link is  good  or  just,  but rather, merely to show that such a relationship exists


Okay, so in high abortion states, crime decreased.  Based on the effects of a reduction in sheer numbers, and on the notion that abortion is more likely in "high risk groups", this makes sense.  Even hilljack conceeded that it is reasonable to assume a plausible link between abortion and crime.

I see hilljack's larger point though.  I think Hilljack's main premise is that we should maintain some degree of skepticism in the face of "research".  If you torture data enough, you can make it say anything you like.  There is a well researched study that points to the correlation between the levels of rainfall in Michigan and stock market returns.  There is reason why nobody makes investment decisions based on rainfall levels though, and it's because of this: correlation=/=causation.  I see the writer went to great lengths to "prove" that the correlation he is finding is indeed causation.  The required, final step.  I know you guys appreciate the fact, however,  that it is in this final step from correlation to causation that models are subjected to more judgement calls (assumptions and methodologies drive the results vs actual data).  This is where we usually find the largest probability for error and the larger the opportunity for bringing model results into question i.e debate. 

The author certainly outlined the statistically significant link between high abortion rates and a reduction in crime. But isn't there a link between high abortion rates and almost every single societal ill that could be influenced by there being fewer poor people around? E.g:  SAT scores, health insurance coverage, mortality rates, Lsat scores...hahaa.  There you go, high abortion rates lead to higher mean Lsat scores!  I should write a book  ;)
UNITED STATES OF AMERICA
1776 TO 2006

Stroopwafel

  • Sr. Citizen
  • ****
  • Posts: 248
    • View Profile
Hooray, yet another thread hijacked by partisan BS.

Neo-cons aggressively disrespecting neo-libs? You don't say?

Increasingly profane, angry, inflammatory language, and name calling? No way!

I don't know which will come first, race wars, religious wars, or partisan wars.
Either way, a box of ammo and a .45 will get you a lot of soup.

A pox on both your houses.

to the OP: I'm reading the book now. Just bought it today. Good read. I'd bring up other interesting chapters, but with this group it would be like shoveling more $hit on the fan.
Rest not. Life is sweeping by; go and dare before you die. Something mighty and sublime, leave behind to conquer time.

-Goethe

InVinoVeritas

  • LSD Obsessed
  • *****
  • Posts: 5550
  • Fine! I shall also fix zee hobo suit!
    • AOL Instant Messenger - NVinoVeritasChi
    • View Profile
BP -- agreed that econometric models and their results deserve a healthy dose of skeptcism. 
however, keep in mind that this study is not new.  levitt published the study in the late 1990s, and it was, indeed, challenged on it's scientific merits -- i.e. were any other factors overlooked, was the data sound, etc.  levitt responded to each of these challenges, and his responses indicate that the analysis was robust.

The author certainly outlined the statistically significant link between high abortion rates and a reduction in crime. But isn't there a link between high abortion rates and almost every single societal ill that could be influenced by there being fewer poor people around? E.g: SAT scores, health insurance coverage, mortality rates, Lsat scores...hahaa. There you go, high abortion rates lead to higher mean Lsat scores! I should write a book ;)

sure, there may be a statistically significant link between high abortion rates and the things you listed.  however, what makes econometric analys meaningful, rather than merely interesting, is the way we can decern the relative impact of one variable over another. my guess is that abortion is not the most impactful factor acting on, for instance, SAT scores.  during the early 90s, however, it appears that the legalization of abortion and increased access to the procedure account for about half, of not more, of the decrease in the crime rate.

_BP_

  • Sr. Citizen
  • ****
  • Posts: 2594
  • Think. Wait. Fast.
    • View Profile
I know you guys appreciate the fact, however,  that it is in this final step from correlation to causation that models are subjected to more judgement calls (assumptions and methodologies drive the results vs actual data).  This is where we usually find the largest probability for error and the larger the opportunity for bringing model results into question i.e debate. [/b]

Good post In Vino Veritas, you made some really good points. 

I highlighted the above quote (I know, how weird to quote myself right?) because it seemed like certain folks were attacking hilljack just because he was questioning the results of the study.  To echo Stroopwafel's comments, it would have been good if we were able to discuss opposing views without the discussion degenerating......oh well.
UNITED STATES OF AMERICA
1776 TO 2006

hilljack

  • Sr. Citizen
  • ****
  • Posts: 1669
    • View Profile
I appreciate the support BP

This reminds me of a joke:

What is the difference between an economist and an econometrician?

An economist uses economic laws and models to be wrong and an econometrician uses computers to be wrong.


Just a joke, no malice intended.

InVinoVeritas

  • LSD Obsessed
  • *****
  • Posts: 5550
  • Fine! I shall also fix zee hobo suit!
    • AOL Instant Messenger - NVinoVeritasChi
    • View Profile
I appreciate the support BP

This reminds me of a joke:

What is the difference between an economist and an econometrician?

An economist uses economic laws and models to be wrong and an econometrician uses computers to be wrong.


Just a joke, no malice intended.

touche, although i heard this one too:

what's the difference between an economist who uses econometrics as a tool and an economist who doesn't?

the one who uses econometrics uses data to confirm or deny his hypotheses.  the one who doesn't use econometrics knows he's right just because.   ;)

ormachea

  • Guest
Wow, those are bad.

hilljack

  • Sr. Citizen
  • ****
  • Posts: 1669
    • View Profile
And they are probably two of the best econ jokes.