TaxProf Blog

Editor: Paul L. Caron
Pepperdine University School of Law

Monday, July 19, 2010

Ideological Diversity and Law School Hiring

Douglas M. Spencer & James Cleith Phillips (both of UC-Berkeley) have posted Ideological Diversity and Law School Hiring on SSRN. Here is the abstract:

America’s legal professoriate does not appear to be particularly diverse. Recent studies have revealed that a majority of law professors are male, Caucasian, and attended law school at one of only a handful of law schools. To date, however, no scholars have examined one of the potentially most influential characteristics of new law professors who play an important role in shaping the law – their political ideology. Utilizing a unique dataset, this explorative study begins to remedy that deficiency by empirically measuring the political ideology of entry-level candidates to law schools between 2005-2009.

Using regression analysis, we find no statistically significant relationship between political ideology and prestige of hiring, although we identify a very large discrepancy between the proportion of new professors that can clearly be identified as liberal or conservative and those whose ideology is less clear. We argue that this discrepancy indicates either unequal hiring patterns or an academic environment at law schools that is less than conducive to openness and free debate. Further research is needed to test these hypotheses and to identify a remedy that will increase the intellectual diversity of American law schools, with ripple effects on the development of law in the United States and in the world.

Chart1
Chart2

(Hat Tip: Danny Sokol.)

Update:

http://taxprof.typepad.com/taxprof_blog/2010/07/ideological-diversity.html

Legal Education | Permalink

TrackBack URL for this entry:

http://www.typepad.com/services/trackback/6a00d8341c4eab53ef0134858be80f970c

Listed below are links to weblogs that reference Ideological Diversity and Law School Hiring :

Comments

Paul,

I think that the study has some problems. First, it does not look at multiple years of data. Second, the authors are unable to identify the ideology of the vast majority of those hired (those for whom they are unsure or only somewhat confident). Third, they only looked at those hired, not on those rejected. The main discrimination is likely to have its impact on the latter. They might even have acknowledged that some people are chilled from applying because of they believe that they will not get a job even though they are qualified. I am not sure if we will find mor liberal leaning minorities or conservatives/liberatrians in those that never apply because they have been chilled from applying. Finally, there are only 8 conservatives in the sample and that is too small a number to draw any definitive conclusions from the study.
I also think that once people are hired, it would be helpful to look at outputs and outcomes. What is the level of productivity in terms of number of articles? What is the placement of such articles, and if in law reviews to normalize the results based on the relative ranking of the law school?

Posted by: D. Daniel Sokol | Jul 20, 2010 1:25:26 AM

Sure liberals are overrepresented in law schools and block conservatives. Of course, it's not very Christian of them, but whoever accused liberals of being Christian?

Bias and bigotry in academia

For admissions officers at our top private and public schools, diversity is "a code word" for particular prejudices.

For these schools are not interested in a diversity that would include "born-again Christians from the Bible belt, students from Appalachia and other rural and small-town areas, people who have served in the U.S. military, those who have grown up on farms or ranches, Mormons, Pentecostals, Jehovah's Witnesses, lower- and middle-class Catholics, working class 'white ethnics,' social and political conservatives, wheelchair users, married students, married students with children or older students just starting into college and raising children."

Posted by: Woody | Jul 20, 2010 8:33:05 AM

The problem is that a lack of exposure to a significant ideological current in the US limits and handicaps these law students and does not provide the well rounded education that these institutions advertise they are providing. At some point somebody's going to file a class action...

Posted by: TMLutas | Jul 20, 2010 12:42:49 PM

Daniel, you make some very good points and I agree with most of your criticisms of our paper. While our data is actually drawn from multiple years (2005, 2007, and 2009), the sample size is still quite small (n=149) and I understand your skepticism of our findings. We are continually coding and hope to have ~1,000 subjects by the end of the year. By then we should be able to report our findings with more confidence.

You astutely point out that we ignore faculty candidates who are rejected, focusing only on those who get hired. This was not a research design oversight. Unfortunately, the annual FAR data is not available to independent researchers so we are unable to examine whether the hiring process is discriminatory, as you suggest. What we are left with is the descriptive information in the AALS Directory of Law Teachers, which we have coded.

Finally, I agree with you that it would be useful to look at outputs and outcomes. The anecdotal evidence of which I am familiar suggests that conservative law professors publish more articles because they fear that tenure committees will view their applications negatively. I am aware of no studies that support or reject this claim.

Posted by: Doug Spencer | Jul 21, 2010 12:33:50 AM

This paper has more than the problems pointed out by Mr. Sokol. This so-called 'research paper' implies the authors began with a presumption and have crafted their statistical conclusions to meet that presumption. The phrase popularized by Mark Twain, "Lies, damn lies, and statistics," might be too harsh but it illustrates the ways statistics can be presented in a way that bears little to no relation to the actual data. This paper has done just that.

149 entry-level professors were examined. Out of that 149, 52 liberal and 8 conservative professors were 'identified.' That leaves 89 non-'identified' professors.

Rounding up, 60% of the professors could not be identified as either liberal or conservative. This is a majority and statistically significant amount of the professors examined.

The authors, perhaps relying too much on their presumptions entering the study, ignored presenting this data and instead focused on the minority of professors who could be examined. From there, they offer breakdowns such as the above graphs which show a great disparity between liberal and conservative professors. 52 professors who could be confidently identified as either liberal or conservative were, in the author's determination, liberal. In contrast, 8 professors were conservative.

The problem is that this is, scholarly and academically, a disingenuous presentation that leads to disingenuous conclusion. If the authors are truly seeking a solid, scholarly research study then they should look beyond their presumptions that conservatives have a harder time getting hired. (One author above talks anecdotally about how conservative authors publish more out of fear of tenure committees. Interestingly, he further takes the polemical tactic of stating "I am aware of no studies that support or reject this claim.")

Looking at the evidence and the first graph above the real story is that so many professors cannot be readily identified using the author's suspect criteria. Further, a large number of professors fall "somewhat confident" categories for both liberal and conservative.

If the data is accurate, why is it that so many professors aren't identifiable? Is there self-censorship? How does this change over time? What happens after tenure is gained?

Nevertheless, the data suggests that, despite the presumptions of the authors, that the disparity between conservatives and liberals is not so great in law schools. It does appear there is a bias towards liberals in law school academia, but it is not so stark as the poorly presented 52 to 8 numbers suggest.

Finally, there may be significant problems with the methodology of the authors. I have not had the time while at work to pour over the paper, but one measure mentioned in another summary of the paper references is "the political party of the president who appointed judges for whom the professors clerked." I believe this measurement comes from a fundamental misunderstanding of how the clerkship process works.

A quick anecdote: An acquaintance of mine clerked for Justice Roberts. By this studies measurement this would weigh in favor of his being conservative. Except he's not; he's a liberal, if not unidentifiable as either liberal or conservative.

Similarly, I know multiple clerks who did not share the ideology of either the judge for whom they worked or the president who appointed that judge. Certainly there are those who get their clerkship because of their ideology, but the greater majority based on credentials or other reasons. In fact, who you know is far more important than what you believe when it comes to a clerkship.

I do believe the statistical data collected, even if it might be from flawed methodology, creates interested questions. My very sincere problem with this study is that, on top of the methodological problems it possesses, it also appears to consciously slant data to reach preconceived conclusions. That is poor research.

I hope the authors take a step back and look at their data again. It raises interesting questions as to why the majority of professors aren't identifiable as either liberal or conservative. It does not, however, provide a smoking gun to conservatives who claim, ipso facto, that law schools have a clear bias against conservative professors.

Posted by: John | Jul 21, 2010 9:17:01 AM

Hi John,

Thanks for the helpful comments, though the tone was remarkably caustic.

This was a paper that was not supposed to go public--a submission to a conference where we hoped to get mature and cool-headed feedback from fellow scholars so that we could fix obvious deficiencies in the study. The conference changed its policy this year, automatically posting papers to the public unless the authors prevented it. My co-author, who took care of submission, was out of town and did not get the e-mail warning him the paper would be public until it was too late--so this wasn't supposed to be available for people to view because we wanted to get it in better shape. Alas, it is too late now.

Of course the nominating party of the judge for whom one clerked is a poor proxy for ideology. That is why if that was the only indicator we did not rate ourselves as being confident in that person's ideology. I personally have spoken with appellate court judges who like to hire 2 conservatives and 2 liberals for clerks. But in the absence of anything else, it gives us something to start with.

As far as relying on presumptions, generally in social science a theory drives your research. That theory may be flawed, or may need to be tweaked, but atheoretical research is frowned upon. Our look at the previous literature indicated a heavy liberal presence, which then gave us something to explore. If you don't like that approach (you would like grounded theory it appears), then fault social science research methods.

You make a big deal about pre-conceived biases, but we didn't find a statistically significant effect for ideology. Neither, if you will notice we explicitly said in our paper, are we making any causal claims whatsoever. This is a descriptive study--nothing more.

You write: "they should look beyond their presumptions that conservatives have a harder time getting hired." We don't even address that. Our paper has nothing to do with the difficulty of getting hired, our paper is about the prestige of the school one is hired at, and we make no conclusions that liberals or conservatives have an advantage when it comes to prestige--our stats don't support it.

You wrote that we "ignored presenting this data," referring to the fact that we could only confidently identify the ideology of 40% of our sample. But we clearly stated that in the paper, including a chart that shows that.

You write: "If the data is accurate, why is it that so many professors aren't identifiable? Is there self-censorship? How does this change over time? What happens after tenure is gained?"

We actually addressed, if you'll recall, some thoughts as to why potentially so many professors are not identifiable, including self-censorship. But as for how this chagnges over time and what happens after tenure is gained, that is beyond the scope of the paper. For someone who lambasts our preliminary unpublished research for making too much out of too little, and then to throw questions at it that cannot be answered by the dataset seems schizophrenic.

Finally, you write: "It does not, however, provide a smoking gun to conservatives who claim, ipso facto, that law schools have a clear bias against conservative professors."

Nowhere in the paper did we claim this--in fact, we clearly stated that we can't answer this question, only describe what we found. It appears that your reading of the paper is not the same paper we wrote, and I encourage a more thorough perusal of the paper before critiquing it. We welcome comments and criticisms now that the paper is out in the open (prematurely we feel), but at least let them be of the paper they wrote and not of a mythical paper the reader created in his head.

And if you really stand behind your commments, you might try the moxey of putting your whole name.

Cheers.

Posted by: James Phillips | Jul 21, 2010 11:55:10 AM

This may be an incredibly obvious observation, but could it be that liberal law graduates are more likely to want want to be teachers? And conservative law graduates lean toward practicing at a large law firm? I don't find this surprsing at all.

Posted by: Beth | Jul 21, 2010 11:56:46 AM

Mr. Phillips,

The tone and stance of your paper is illustrated in its abstract. Here, let me quote from above:

"Using regression analysis, we find no statistically significant relationship between political ideology and prestige of hiring, although we identify a very large discrepancy between the proportion of new professors that can clearly be identified as liberal or conservative and those whose ideology is less clear. We argue that this discrepancy indicates either unequal hiring patterns or an academic environment at law schools that is less than conducive to openness and free debate."

This paragraph from your abstract highlights the discrepancy between those liberals you can clearly identify and the clearly identifiable conservatives. You then tie it together with your final sentence that this "indicates unequal hiring patterns" or an academic environment that does not foster free debate.

I appreciate that you didn't expect a paper posted publicly on SSRN to seep out onto the web. Nevertheless, trying to claim your paper does not raise questions about how easy it is for conservatives to get hired (see your above abstract) is not supported by your own abstract.

As for claims of "schizophrenia" in reference to pointing out how your data raises more relevant questions that might be explored -- seriously?

A bold charge there, Mr. Phillips. Perhaps a more politic and collegial term you can use at the conference where this will be presented is "logically inconsistent" -- even though my comments were not, as you seem to believe, logically inconsistent.

Finally, you talk about how social science research papers are 'designed.' Your attempt to defend framing statistics in a manner to meet a preconceived theory, rather than adjusting that theory to align with the later information your obtain through research, rings hollow.

Inflexibly sticking to a theory, or shaping statistics in a way to frame conclusions intended to support a preconceived theory, is not a research method. Rather, it is a method arising from the field of polemics rather than academics. Academic research molds theories and conclusions based on research data and information, it does not mold research and information based on theories.

Yes, you start with a theory. You might remember this starting point from your science classes. It's actually an hypothesis. This hypothesis is then tested with research and data until it is able to be supported and, therefore, becomes a theory.

Perhaps faulting you for failing to follow a traditional academic research path isn't fair. Perhaps past professors who rewarded a conclude first, support later approach to research papers are the ones to blame. That doesn't change the fact that such an approach is rife with problems and opens an author (or authors) up to questions of preconceived conclusions and presumptions.

The bottom line, Mr. Phillips, is that you and your co-author have framed this paper in a way that highlights statistically minor aspects of your theory and do not appear to have modified your hypothesis -- or original working theory -- in order to chase the interesting questions that have been raised by your data.

I hope to have more time to read the paper in full after I get home from work. You are right that it is not entirely fair of me to lambast your work as much as I have without having reviewed the paper thoroughly. Nevertheless, the way the paper is framed, and the data that has been highlighted, seek the swift conclusions that others, including myself, have arrived at as to the nature of your paper.

And, as an aside, I do wish you good luck at the conference and I hope it does provide good feedback. I may disagree with your approach, but the dataset you do provide is interesting and does seek to fill holes in the current research on this topic area.

Posted by: John Nelson | Jul 21, 2010 1:47:52 PM

Thanks for the response. I appreciate the change in tone, and will strive to do likewise. One lesson I have learned from this (besides don’t let your research get out until you are done with it), is that in the future I will not comment on other’s research until I have read it fully and carefully. I feel like you’ve looked at the findings section, maybe just the graphs and charts, and read the abstract, and then fired away.

I don’t mind devastatingly accurate criticism (in a civil tone), and in fact invite, but I do feel shortchanged when it appears the critic hasn’t really read the paper.

Please particularly note in the abstract we indicate this is an “explorative study”, and that after laying out an argument for the possible explanation behind our findings, we clearly state “further research is needed to test these hypotheses.”

While your judgment of our article based on the “tone” of the abstract is a good lesson to be more meticulous about abstract writing (with the understanding that unfortunately in scholarship one has to “sell” one’s article to get it accepted, highlighting the most interesting or potentially controversial findings), an abstract does not an article make.

Therefore, please further note: “…this study cannot purport to be anything beyond an initial foray into the area of law school hiring and ideology, it is hoped that these findings will lead to further research questions and answers” (p. 2).

“We acknowledge two important caveats at the outset. First, our data is limited to entry-level candidates who were ultimately hired. We do not have information about candidates who failed to receive, or who rejected an offer of employment as the AALS is reluctant to release that information to researchers. Thus, we are unable to measure whether ideology has an effect on the likelihood of getting hired and our study is limited only to the effect of ideology on the prestige of the hiring school. Second, as an extension, our study does not purport to explain any causal mechanism in the law school hiring process. Our goal is merely to engage in descriptive and somewhat predictive investigation and we provide very little analysis about why any trends may occur” (p. 8).

“Neither the categorical nor the dummy ideology variable was statistically significant,
indicating a lack of support for the hypothesis that liberal hires are more likely to land a
prestigious appointment…There are a few possible explanations for this finding. The simplest is that ideology is unrelated to the prestige of school one is hired at. A second reason for this finding might be the inadequacy of our coding.” (p. 16).

“…we are not very confident or completely unsure of the ideological
leanings of nearly 60% of the new professors in our sample. This might be due to the fact that these professors are more moderate in their political ideology, or it might be because they have done a better job of keeping signals about their ideology protected from public eyes” (p. 16).

“Of the 60 professors whose ideological signals were easily identifiable, 52 (87%) were liberal and 8 (13%) were conservative. See Figure 2. These findings suggest that the population of new hires who are transparent in their political leanings are disproportionately liberal in makeup. We are uncertain how representative these 60 professors, who are either more politically active or more politically transparent, are of the larger sample and of the larger population. If they are representative then the 13:2 liberal-to-conservative ratio does not speak well of intellectual diversity in American law school hiring. An alternative explanation might be that the large discrepancy is an indicator that conservative law professors, at least early in their careers, are more discreet in evidencing their political leanings and that, perhaps in reality, the ideological balance is more equal. Even if the latter is true, though, the implied academic freedom and environment of openness in law schools may be little more than a myth if conservative or libertarian professors perceive some level of hostility towards non-liberal
positions and politics. Finally, it might just be the case that conservatives are overly cautious when there is no need and that our data have very little normative implications. Our study cannot determinatively say” (p. 16-17).

All studies have imperfections, and ours is no exception…Second, we were unable to code
important information relevant to the hiring process such as the quality of one’s interview or
one’s letters of recommendation. Third, our data is limited to individuals who were hired as
entry-level, tenure track law professors. Our analysis ignores the lateral hiring apparatus and
does not account for candidates who go on the job market but are not offered academic
appointments. Fourth, our data do not account for exogenous influences that may cause a faculty
candidate to accept a position at a less prestigious law school for personal or other reasons
unrelated to any of the factors in our model” (p. 19-20).

Mr. Nelson, what more could we have done? Multiple times we warn the reader. We suggest multiple potential explanations for our findings, cutting in different directions. We say that further research is necessary to determine what is going on. We repeatedly inform that our data can only describe, not explain causation.

It is Stephen Teles and William E. Nelson, Harvey Rishikof, I. Scott Messinger and Michael Jo who write that liberals dominate legal academia. Then we find that we can’t tell on 60% (which includes the professors where the judge they clerked for was the only indication of their ideology) , but on the 40% that we could clearly tell, 87% were liberal. Please tell me, Mr. Nelson, how should we have treated such a finding? I know you in previous posts emphasized the 60%, but a lack of an indicator of ideology does not mean they are non-partisan—just that the way we are using to measure it is not picking anything up. Otherwise, previous literature all finds one thing, you find something similar in the data clear enough to analyze, and then you warn the reader over and over that this is description, not causation, and that there are multiple possible explanations that could account for this and you can’t prove or disprove any of them—what more should we have done?

No, I don’t like the way it is being framed by the media, and this is another lesson learned. If you will check out the updated version of the paper we have tried to make thing more clear that this is all preliminary. But how in the world could we be more clear than that which I quoted you from the paper?

As for my use of the term schizophrenic ,which you mistakenly or purposely re-quoted as “schizophrenia,” schizophrenic means “Of, relating to, or characterized by the coexistence of disparate or antagonistic elements.” Maybe incongruent would have been a less charged but equally accurate word and was a poor choice on my part. I did not intend mental disorders on your part. I still disagree with your interpretation of whatever you’ve read of our paper, and I’m guessing you reaction to our paper was so strong, so visceral (based on your initial tone), that reading the whole thing will not change your initial gut reaction, which I think is unforunate and speaks more of the polemic rather than the academic which charge you lay at our feet. That being said, I should have exercised better judgment in my response to your word choice.

I welcome your further input (I won’t be attending the conference) as I would genuinely like to improve the paper. However, please only comment after you’ve actually read the paper in its entirety. Then fire away as you'd like someone to fire at you. Thanks.

Regards.

Posted by: James Phillips | Jul 21, 2010 3:17:43 PM

Mr. Phillips,

I did not say I had not read the article. I said I had not read the article thoroughly. I have done that now and, as per your request, I will provide specific comments on problems I see in your paper.

First, some general points. To be frank, your paper read like the papers of undergraduate students. You have some good concepts in your paper but they should be executed better.

Having said that, I also understand that this is a draft. It is always much easier to find criticism in a draft than it is in a final product. Heck, I'd shudder to think what folks thought of some of my drafts.

Here's the general rub I have with your paper: it appears to presuppose something (lack of political diversity in law schools and a liberal slant) and then, by using a data model, appears to shape the data to meet the presupposed conclusion.

Clearly you don't feel that way. Clearly others have interpreted your paper in a similar way -- perhaps not as harshly as me, but nevertheless as a study indicating a higher percentage of liberals in the academy.

Couple this with the fact that I do not believe your data remotely suggests anything more than the majority of incoming law professors cannot neatly fit into a political ideology box.

Worse, the introduction and background on the issue of political diversity in law schools feels conclusory, which only reinforces my impression (and possibly others) that you have approached the data's analysis with a lack of objectivity.

Having thoroughly read your paper has not changed my stance on this. However, I'd like to offer some specific suggestions on ways my deep reservations in your data analysis and inferences might be addressed.

First, the background of your paper deserves more attention. Here's what I, as the reader, wants to know:

1. What is the issue?
2. Has this been addressed before?
3. Why should I care?

I believe you can approach this by breaking your background section (what you call your literature review) into the following sections:

1. Critics argue that the legal academy lacks a diversity of political ideas because it is dominated by liberals and lacks enough conservative voices.

-- This is where you should address the controversy. Discuss some of the less academic, more popular rants and discussions of this to add color and context. Add in more serious studies such as Nelson, Rishikoff et al; Klein & Stern; and others.

2. The legal academy's diversity, or lack there of, has been studied in the past -- but these studies have not addressed political ideology.

-- This is where you should discuss previous studies, I.E. Fossum, Bothwick & Schau, Redding, et cetera. Also highlight how these studies and others have not addressed political ideology.

3. Intellectual diversity in the legal academy is important and political ideology should be considered an important component of that diversity.

-- This section may be your most important. The first two can really be more glosses of the controversy/issue coupled with a quick rundown of what's been done before. This is where you need to convince the reader that this matters.

Clearly you think this matters. That's good. Me? I don't agree with you. Many on the ABA Journal's website didn't agree with you either. This is where you tell us why its important and, hopefully, convert some of us to the idea that political diversity in the legal academy matters.

Discussing works like Sears & Funk and Ditty & Lopez would be important here.

I know you say you are following the Social Science Research Method approach of Introduction, Background Literature, Methodology, Data, Analysis, Discussion, Limitations and next steps. That method, however, is the rote method followed by undergrads.

You can do better. This paper deserves better. People will like and appreciate your paper better if you break out of this schema.

Next, your methodology section deserves more work as well. You should explain why the methodological approach you chose was the best for this paper.

One of the problems with your approach is that it weakens the integrity of the data. You've designed a study where you've created the criteria and then determined whether someone else has met it. This creates an increased risk of your data being sub-consciously influenced by your desire to support your original hypothesis.

Nevertheless, you've mentioned on the ABA website that a polling method would not be viable for this study. Explain why such a method, and others, won't work and why this one does.

Then you need to beef up your independent variable section. Those variables, and their information sources, are some of the most controversial aspects of your study as you've already seen from some of the comments you've received.

Breaking this part into two sub-parts would probably help. Introduce it with a brief paragraph outlining the existence of two variables based on information sources. Then have a sub-part explaining each information source. Provide more detail than you have.

For example, using Facebook profiles may prove controversial. Explain why this is a possible indicator of political ideology. Provide some examples for both liberal and conservative indicators. (You only mention as an example being 'friends with Keith Olbermann. This will only serve to further alienate those who already view your study with suspicion and believe it has an agenda.)

Another example is the publication information source. Provide more details about the DIR variable of the Supreme Court Database. Summarize this variable -- don't presume people know it. Also, don't make people have to go read other sources on the variable in order to understand it. In addition, provide some context of how this variable might be used to rate a paper as having a conservative or liberal slant.

You should provide more detail and context for each of your information sources. This is where you should defend your use of the sources with details of why it is used and how it used.

The next sub-part should describe your application of these information sources to your two variables. One variable appears to be a sliding scale variable; another a binary variable. Provide examples of how these variables will interact with different values derived from the information sources.

The next section of your paper should not be a combined analysis and discussion. Instead, only analyze the data in this section. The reason this is important is two-fold. One, it serves to separate in the mind of the reader your data from your conclusions. Two, it serves to firewall the section off from other sections predicated upon argument, opinion, and conclusion. Therefore, even if someone vociferously disagrees with your argument, opinions, and conclusions that same person will still find your paper valuable for the data and its analysis.

The limitations section should be a sub-part of your data analysis. After all, the analysis should examine problems and surprises that presented in the data collection process. This includes imperfections or weaknesses that are apparent in the study.

The next section after your data analysis should be an application of the data to the issues presented in your background. This application may result in a conclusion that your data can't support an argument. Nonetheless, even insufficient or weak data can indicate that interesting possibilities exist.

You acknowledge this through your disclaimers that this is an exploratory study and the implicit conclusion that more data and more study is warranted based upon what you have. There's nothing wrong with that -- there is value in what you have, even if that value is indicating the need for more data and analysis.

Finally, you should have a section on conclusions. This is where you talk about possible reasons why your data ended up how it did. Further, suggestions of additional studies that may help clarify questions can be proposed. Finally, some pithy, single sentence conclusion of 'what this means' should exist. (The one-sentence summary of your findings, if you will.)

I highly suggest reading Eugene Volokh's Academic Legal Writing. It is an excellent book that examines ways in which to strengthen an academic paper, be it legal or otherwise. I also highly suggest you work to soften your highly-technical language explaining your process. Try and write in plain English.

Think of how you might explain something to a friend, or kid-sister or brother, or your parents. Sure, you'll necessarily have to keep some technical language in there. Even so, by striving for a more plain-english approach your broaden the potential audience, impact, and value of the paper.

Your paper is currently in a rough state. I know how that is. I went through three complete drafts before I finished a final draft of my most recent paper. Each draft was drastically different in style, approach, and tone.

Even so, I would also caution you on a non-stylistic matter. I do not agree with the following statement from your paper on page 20: " . . . yet the extreme discrepancy between the proportion of new professors who can clearly be identified as liberal or conservative indicates either unequal hiring patterns or environments less than conducive to openness and free debate in the law school setting." This is echoed in your abstract.

Firstly, you create a false either/or dichotomy. Your data might indicate either of those two options, or it might indicate something else entirely. Second, I do not really believe it indicates either of those possibilities. This conclusion ignores the large number of people you could not identify. Not being able to identify the political ideology of so many might indicate that the environment is very conducive to openness and free debate because the majority don't adhere to strict political dogma but, rather, are open to many ideas.

You may disagree, and that's fine, but this is a statistically significant aspect of your data. Drawing such a firm conclusion is not supported by the data your present. The data may suggest that such a conclusion is a possibility, it is far from certain and based a statistically small amount.

One way to avoid this kind of problem is to speak in terms of indications relative to arguments. For example, perhaps the discrepancy indicates there may be some credence to complaints that conservatives are underrepresented in law schools or disfavored in hiring, but more study and more data is required to fully examine this possibility.

So, Mr. Phillip, I do have sincere reservations about the approach you have taken with your study. Nonetheless, I believe you and your co-authors further work on this paper could at least defend those deficiencies I've pointed out.

Oh, and the whole disclaimer / warning bit: it is unhelpful. Sure, you have to disclose weaknesses in your approach or arguments. Here, though, you seem to be telling me that you've warned me to disregard any conclusions you've drawn, but you still draw those conclusions. Either don't try to draw them if they're not to be trusted, or make valid and sound arguments as to why those conclusions can be drawn.

Good luck working on it. I'll try and keep an eye out for future revisions.

Posted by: John Nelson | Jul 21, 2010 8:03:50 PM

John, thank you for the thorough and specific critique. This is the sort of thing that is helpful.
Sincerely, James

Posted by: James Phillips | Jul 22, 2010 11:17:48 AM