Coaxing information out of the political universe with statistics

In summary: I assume you've at least glanced at it) really speaks volumes. Originally posted by Zero Sorry, but I think you just really don't know what your talking about. In fact that you use the term 'study' in scare quotes in reference to the very same ideological scores that are the basis of Berkleys very impressive voteworld program they've undertaken to develop with data sets internationally and nationally (available here as well as it's prescence and availability for researchers use at the Princeton site among other well known and respected institutions...and it's overwhelmingly positeve reference throughout the political science,...I assume you've at least glanced at it) really speaks volumes.
  • #1
kat
42
0
Origionally posted by Zero
Wow, how hard did you have to dig to link me to somebody's college thesis? And, really, all that paper says is pretty much teh opposite of what you suggest it does. Nice try though, I suppose.

Uhhhh...Wow..that's Keith T. Poole of Poole and Rosenthal. Here's the google on that to head you in the right direction poole and rosenthal

So, what I would like to do is really examine the article and the data and see if this direct quote from the article (the quote that Zero is referring to) is supportable by the data. Please don't post if you are not directly referencing the data. I really am not interested in more..opinion without support from the data

The regional differences within the Democratic Party have almost completely disappeared and Democrats as a whole have become more liberal. In contrast, Republicans have become more conservative since the 1980s. The combination of the Democrats becoming more liberal and the Republicans more conservative has resulted in a trend towards ideological polarization within both chambers.
http://voteview.uh.edu/chminds.pdf [Broken]
 
Last edited by a moderator:
Space news on Phys.org
  • #2
I read it differently from you, kat. I saw the point of the entire article to be that the Dems and Repugs are more consistantly voting along party lines, not that there is any real shift in ideology. After all, the data is only based on yes and no votes of what comes through Congress.
 
  • #3
Originally posted by Zero
I read it differently from you, kat.
Honestly, I don't know how you can read it differently. It says..what it says..and I used a direct quote from Keith Poole.
Democrats as a whole have become more liberal. In contrast, Republicans have become more conservative since the 1980s. The combination of the Democrats becoming more liberal and the Republicans more conservative has resulted in a trend towards ideological polarization within both chambers.
I read the point of the article as being that Ideologies do not shift, despite the will of the constituent. However, I read this particular statement..exactly as it was stated.
 
Last edited:
  • #4
Originally posted by kat
Honestly, I don't know how you can read it differently. It says..what it says..and I used a direct quote from Keith Poole. I read the point of the article as being that Ideologies do not shift, despite the will of the constituent. However, I read this particular statement..exactly as it was stated.
Yes, but you read it outside the context of the overall presentation, to support your preconception.
 
  • #5
Originally posted by Zero
Yes, but you read it outside the context of the overall presentation, to support your preconception.

Zero, you're mistaken. I really don't know how you can disagree with the statement. I'm sure you agree that the Republicans have become more conservative since the 80's? Secondly, Democrats have become more liberal, (I would assume, particulary in regards to civil rights)..."when taken as a whole"...the statement includes the Southern democrats shift as noted above in my origional quote... I don't know how you can possible disagree with this!
Also, if I'm not mistaken roll call issues were given point values?
 
  • #6
Originally posted by kat
Zero, you're mistaken. I really don't know how you can disagree with the statement. I'm sure you agree that the Republicans have become more conservative since the 80's? Secondly, Democrats have become more liberal, (I would assume, particulary in regards to civil rights)..."when taken as a whole"...the statement includes the Southern democrats shift as noted above in my origional quote... I don't know how you can possible disagree with this!
Also, if I'm not mistaken roll call issues were given point values?
I would say that there is nothing conservative about the Republican party either. The world doesn't work the way this 'study' claims it does. Actually, the more I think about it, the more bogus I think this whole thing is. Thanks for bringing it up, though: I'm sure some folks would find it more interesting than I do.
 
  • #7
Originally posted by Zero
I would say that there is nothing conservative about the Republican party either. The world doesn't work the way this 'study' claims it does. Actually, the more I think about it, the more bogus I think this whole thing is. Thanks for bringing it up, though: I'm sure some folks would find it more interesting than I do.

Sorry, but I think you just really don't know what your talking about. In fact that you use the term 'study' in scare quotes in reference to the very same ideological scores that are the basis of Berkleys very impressive voteworld program they've undertaken to develop with data sets internationally and nationally (available here as well as it's prescence and availability for researchers use at the Princeton site among other well known and respected institutions...and it's overwhelmingly positeve reference throughout the political science, economical and political historian "world".
In actuality by posting here I was looking for input from those who might have a better grasp of scientific/mathematical statistics and the use of these data sets then I, as opposed to getting into an unnecessary debate about the slight nuances of your understanding of such broadly abused terms like "right", "left", "conservative, or "liberal"
 
  • #8
Well, like I said, thanks for posting it, maybe someone will be interested. I personally think statistical analysis is too easily manipulated to be trusted for socialogical purposes.
 
  • #9
I think the study is misleading in a number of ways. I'm not accusing the authors of intentional deception, I just think that they lead the reader to infer some incorrect information.

They discuss democrats becoming more liberal and republicans becoming more conservative since 1980. Much of this can be attributed to dixiecrats leaving the democratic party. In the early '80's they voted more conservatively than most republicans. When they switched to the republican party, it gave the impression that democrats became more liberal. Nobody's ideology changed, but statistically, it looked like the democrats moved far to the left.

They analyze roll call votes. This is a very poor measure of liberal/conservative bent. The real measure is what gets to the floor to be voted upon. The voting just determines how partisan congress has become, not how liberal or conservative. Consistent, polarised voting could still occur with both parties moving to the right. While I have no data to back it up, I maintain that the republicans have moved far to the right, and the democrats have moved right to a lesser degree. The gulf between the parties has widened, and polarisation in voting has increased, but it does not mean the democrats have become more liberal.

Njorl
 
  • #10
Originally posted by Njorl
While I have no data to back it up, I maintain that the republicans have moved far to the right, and the democrats have moved right to a lesser degree. The gulf between the parties has widened, and polarisation in voting has increased, but it does not mean the democrats have become more liberal.

Njorl
Statistically, you could 'prove' that the Democrats are more liberal, because they have moved to the right to a lesser degree than the Republicans, don't you think?
 
  • #11
Njorl, I really appreaciate how you not only reply but spell out your reasons in such a clear, and often educating manner. Thanks a lot.
I think though, that in the quote above "The regional differences within the Democratic Party have almost completely disappeared" (if you use the find property you can easily see the statement in context} it is referring to a combination of the dissapearance of the issues and shift of members party affiliation in regards to the Southern Democrats (Dixiecrats). In fact they are even tracked separetly to show you that. This is why (and I quoted earlier)"Democrats as a whole have become more liberal.". I don't see the article as misleading at all, it clearly states that the purpose of the study is to show that once individual members achieve office...despite what the voters may..need, desire..percieve..or what the members have declared they would achieve...their ideologies do not shift...and that the parties Ideology as a whole only shifts...when indiviual party members with new/differing ideologies enter the picture.
I also think your both incorrect in stating that the data used in the mathematical equations that the study is based upon is simply the nay or yay votes in roll call..this ignores the entire first several paragraphs of the article where he clearly spells out the method used...and it's aplication.
 
Last edited:
  • #12
Originally posted by Zero
I personally think statistical analysis is too easily manipulated to be trusted for socialogical purposes.
I didn't read the study, but I had to comment on this. For sociological purposes, statistics are often the ONLY data we have to go on. No they are not perfect and yes they can be manipulated, but they are all we have and its important to know how to interpret them. I'm not saying you, Zero, but the lack of understanding of statistics is a pretty big problem in the US, mostly in/from the media.
 
  • #13
I'm sorry, but I can't accept the model where the Dixiecrats left, taking their conservatism with them, and leaving the remainder of the democrats statistically more leftish.

The dixiecrats were only conservative on the one issue of race. On economic issues, they voted the majority ticket. And that was a more leftish ticket than you see today. Highly progressice income taxes (over marginal 90% in top brackets), govenrment programs frankly designed to afftect social change, Keynesian pump priming, the whole program of the New Deal, which was never questioned by Democrats of any wing in those days.


And what do we have now? Clinton's biggest success was balancing the budget, with a lot of GOP help. His other one was welfare reform, in a style invented by Wisconsin's GOP governor Tommy Thomson. Unions are sinking in power and influence in the party - watch Gephardt's candidacy. As a life long Democrat, it seems clear to me that the national party has drifted way to the right. Is there now a "silent majority" of quasi socialists?
 
  • #14
Originally posted by russ_watters
I didn't read the study, but I had to comment on this. For sociological purposes, statistics are often the ONLY data we have to go on. No they are not perfect and yes they can be manipulated, but they are all we have and its important to know how to interpret them. I'm not saying you, Zero, but the lack of understanding of statistics is a pretty big problem in the US, mostly in/from the media.
Yeah, even statiticians can screw up on a semi-regular basis, after all., and the media certainly don't have any idea how to cope.
 
  • #15
From page 12:

"Both chambers became substanially more liberal from the 1940s into the 1960s and both chambers have become more conservative since the Reagan era."

I missed that when I skimmed the article. At first glance it seems to run counter to:

"Democrats as a whole have become more liberal. In contrast, Republicans have become more conservative since the 1980s"

One could argue that this merely reflects the change of control of congress from democrats to republicans. While the fact that the majority has shifted from one party to another has had monumental effects, the percentage shift in congressmen has been small. A small democratic advantage changing to a small republican advantage may have a big political effect, but the statistical effect is small.

I think this conflict is actually caused by the way the authors take data. Again, I come back to the use of yea and nay voting to try to reflect conservative/liberal trends. It is useful, but inadequate. The authors reasoning is that a liberal (or conservative) on one issue is a liberal (or conservative) on all issues. They analyze voting records to demonstrate that it is true. While I do concede that it is true on those issues that get voted upon, I maintain that it is not true in general. Some issues are too liberal for some liberals, others are too conservative for some conservatives. The author's premise implies that if a waiting period for the purchase of handguns can pass, so too can a bill calling for the confiscation of all guns. That is ridiculous. These issues never make it to the house floors for votes, and so are not reflected in their statistical analysis. Where that line is drawn is even more reflective of the extent of liberalism or conservatism in a party or the legislature as a whole.

Njorl
 
  • #16
I'll have to read your comments again later when I don't have the headache I have atm, Njorl. I do want to say that I STILL think you're overlooking the complexity of the methodology they use in the NOMINATE program. It touches on it here http://voteview.uh.edu/nominate/nominate.htm [Broken]

much more info on their data and methods here: http://voteview.uh.edu/default_recpap.htm [Broken]

and to just give an idea I quote:
Our solution was to go "high tech" in that we transferred a standard decision model from economics to a legislative setting. In this model legislators have utility functions and they vote for the alternative on a roll call for which they have the highest utility. This utility function consists of (1) a deterministic component which is a function of the distance between the legislator and a roll call outcome in the basic space; and (2) a stochastic component which represents the idiosyncratic component of utility.

We assumed that the stochastic component was a random draw from the logit distribution. Given these random draws, we could then calculate the probabilities of each legislator voting Yea or Nay. Therefore, given a matrix of roll calls the problem is to estimate legislator ideal points and roll call outcomes that maximize the joint probability of the observed votes. That is what NOMINATE was designed to do.
In reality, Howard and I had to push the complexity of NOMINATE to the point that we were working with a dynamic multidimensional model (D-NOMINATE) which required a supercomputer to perform the estimation. We could only study our results by turning them into animations that we could then view on videotape (these can now be viewed on our website – http://k7moa.gsia.cmu.edu [Broken] or http://voteview.gsia.cmu.edu [Broken] -- in the form of animated gifs). These videotapes were the culmination of our efforts. The animations are relatively easy to understand and interpret despite the tremendous complexity of the theory and the computer programming that produced them. This simplicity of understanding of political events over time would have been impossible without our "going high-tech".
 
Last edited by a moderator:
  • #17
They don't go into enough detail of their methodology on any of those sites for me to make a judgement on it. They talk around NOMINATE a lot, but don't get into the guts of it. I'll keep checking though.

I may have been too quick to judge. There is a lot there, and I have only skimmed the surface. I am a sucker for rigorous statistical analysis, and clearly, these guys have done much more work than anyone else I've come across. Of course, doing more work doesn't necessarily make you right.

I do modelling of electron transport in devices. My models are relatively simple. A good buddy of mine does modelling of natural gas prices for a large energy trader. His models are very complex, and are usually proprietary information. He warns that making models too complex can lead to perfect modelling of the past and worthless modelling of the future. Another pitfall of modellers is misinterpretation of the models. A sufficiently complicated model will show you what you want to see, so it is very important that you want to see the truth more than any other outcome.

Poole himself stated in his "History of NOMINATE" that he believed in his method before he applied it. That makes me a bit suspicious. I'd like to see some disinterested critiques of his method if there are any. I did some half-assed googling of Poole, nominate etc, but all I found was his stuff, no appraisals of him.

Njorl
 
  • #18
Originally posted by Njorl

I'd like to see some disinterested critiques of his method if there are any. I did some half-assed googling of Poole, nominate etc, but all I found was his stuff, no appraisals of him.

Njorl
I'd like to as well, but it appears that it has perhaps developed to the point that it's beleived to be reliable enough for a multitude of respectable institutions to base numerous studies and reports upon. That is however, one reason I posted it here, I would like to see a critique by someone who is knowledgeable in this area. If I don't find something online soon I may email him about it.
 

1. What is the role of statistics in extracting information from the political universe?

The use of statistics is crucial in extracting information from the political universe. Statistics allow scientists to collect, organize, and analyze large amounts of data to identify patterns and trends. This data can then be used to make informed conclusions and predictions about the political landscape.

2. How can statistics be used to uncover hidden information in politics?

Statistics can be used to uncover hidden information in politics by examining data from various sources and applying statistical methods such as regression analysis and data mining. These techniques can reveal relationships and patterns that may not be immediately apparent, providing valuable insights into the political universe.

3. What are some potential challenges in using statistics to extract information from politics?

Some potential challenges in using statistics to extract information from politics include bias in data collection, incomplete or inaccurate data, and the complexity of political systems. Additionally, the interpretation of statistical results can be subjective and influenced by personal biases.

4. Can statistical analysis be used to predict political outcomes?

Yes, statistical analysis can be used to predict political outcomes by analyzing historical data and identifying patterns and trends. However, it is important to note that predictions are not always accurate and can be influenced by various factors such as unforeseen events or changes in public opinion.

5. How can scientists ensure the reliability of their statistical findings in the political realm?

To ensure the reliability of their statistical findings in the political realm, scientists must use rigorous data collection methods, carefully analyze their data, and consider potential biases. They should also use multiple statistical techniques and replicate their results to increase the validity of their findings. Additionally, transparency and peer review are important in the scientific process to ensure the accuracy and integrity of statistical findings.

Similar threads

Replies
96
Views
9K
Replies
19
Views
2K
  • General Discussion
3
Replies
74
Views
9K
  • Cosmology
Replies
4
Views
1K
Replies
5
Views
7K
Replies
1
Views
3K
  • General Discussion
Replies
30
Views
5K
Replies
39
Views
6K
Replies
2
Views
2K
Back
Top