Wednesday, May 24, 2006
Digestible Numbers from Law Profs
A couple of days ago (that's a different eon in blog-time), Yair asked whether statistics should be a required course for law students. My quick response is that required statistics courses aren't feasible at law schools until the elective curriculum in numbers-based learning proves itself. Show me a law school that has a half-dozen courses with some statistics component, and I'll show you a law school that's ready to start talking about a required course.
But what about the rest of the legal world? How can law schools encourage more numeracy (and not just literacy) among the lawyers now in practice? Do you think that lawyers who've tasted the real life needs of legal practice have more interest in statistical learning? Possibly law schools and providers of CLE courses should be offering brief "Statistics for Practicing Lawyers" seminars.
Some law professors might also take on the job of providing useful, digestible numbers for practicing lawyers. We could become their guides through the massive data bases that lie around, unused by lawyers who don't have the time to appreciate what they're seeing. Perhaps most law professors, along with producing a law review article every so often, could also generate a periodic report on interesting and useful statistics from their specialty field. The aim would be to give lawyers some insight into the landscape, producing numbers that show a pattern that otherwise remains hidden.
For an example of this sort of thing, here's an SSRN link to my recent effort to create a set of useful numbers for practicing lawyers. In this case, the numbers involve the charge reductions that prosecutors make, reductions that lead to predictable drops in sentences. A criminal lawyer should know the statistical lay of the land before starting plea negotiations, no?
TrackBack URL for this entry:
Listed below are links to weblogs that reference Digestible Numbers from Law Profs:
» "A guy named Professor Heise" from Empirical Legal Studies
Over at Prawfsblawg, Ron Wright has an absolutely terrific post on constructive steps toward a legal profession that can take full advantage of empirical research and statistics. The comments are also worth reading. This comment on the value of a [Read More]
Tracked on May 24, 2006 9:27:06 PM
» The Value of Stats for Law Students and the Legal Profession from Empirical Legal Studies
Over at Prawfsblawg, Ron Wright has an absolutely terrific post on constructive steps toward a legal profession that can take full advantage of empirical research and statistics. The comments are also worth reading including a comment on the value of [Read More]
Tracked on May 25, 2006 3:33:07 AM
Ron, great suggestions. Law professors are well positioned to create a market within the profession for empirical data on legal issues. bh.
Posted by: William Henderson | May 24, 2006 10:41:08 AM
I took a "stats for lawyers" class while in paactice. It was taught by an economics consulting firm that was essentially peddling its services as expert economists. I thought the class was very good; I wish I remembered the name of the folks who offered it, but the company offers this course in three cities, I believe.
Posted by: Scott Moss | May 24, 2006 11:19:23 AM
A beneficial place for a full quarter's course in statistics and probability is high school, though, likely, not public high school, save in affluent communities in which, statistically speaking, most of the children are above average. There is a commentator who launches a monologue with a similar depiction.
The economics vantage is important, but becomes intermixed with a studious view of the unfurling of history. A Big-5 accounting firm (then there were 3 post Sarbanes Oxley) survey is an interesting concept. The best statistics savvy attorneys I have known first obtained a degree in economics, then entered law school. Though, your idea sounds like a wonderful map for a new website waiting to see the light of cyberspace.
Posted by: John Lopresti | May 24, 2006 12:55:05 PM
My firm offered a course in "Financial Accounting for Lawyers" when I was in practice. It was put on by an outside group and was very useful for learning to read balance sheets and other financial documents.
Posted by: Carlton Larson | May 24, 2006 12:55:40 PM
The notion of legal academics providing useful, digestible numbers sounds good ... but my concern would be the "consumers." Practicing attorneys may not have the statistical, methodological or other appropriate background to use such data sets ... which comes back to the question whether they should learn about it in law school. Of course, legal academics as well may not have the appropriate background ... which, I suppose, comes back to the question whether they should learn about it in law school.
I think an introductory stats and methods survey course would be very useful in law school, perhaps more so than more focused ones given by accounting or economics firms. A survey course wouldn't be just focused on finance, or multiple regression, etc., but could give a good sense of probability, experimental methodology, causality, and applications to legal problems such as those identified in the original post. I know there are some such animals in various law schools - did anyone take one and find it worthwhile (or otherwise)?
Posted by: Jeremy A. Blumenthal | May 24, 2006 1:50:11 PM
I took a course in law school from a guy named Professor Heise. I believe he has since moved on to teach at Cornell. Although the class was more rigorous and clearly different than standard law school fare, it was easily the best course I took in law school. The material was interesting and useful in future litigation practice. In fact, I use it all the time. Also, Prof. Heise was a very enthusiastic teacher. Always willing to help out outside of class. A real class act, I am sure my alma mater is missing him now. Incredibly worthwhile, and a class I'd recommend every future lawyer take (especially if you can take it from Heise).
Posted by: Jason M. Gerber | May 24, 2006 3:57:55 PM
It's "Yair," not Greg, or "Yiar". YAIR's post. Please keep your fellow bloggers' names straight. You're thinking of Greg Lastowka, the IP scholar at Rutgers, not Yair Listokin, the Yale Law & Econ hotshot. Greg Lastowka isn't a blogger here. Please get with the program. The names aren't really that similar.
Posted by: Anonymous Annoyed | May 24, 2006 11:56:05 PM
As "Anonymous Annoyed" pointed out, I mistakenly renamed Yair as "Greg." My bad; I've edited the post.
Posted by: Ron Wright | May 25, 2006 9:14:28 AM
Ron, I think numeracy is basic to competence as a business lawyer, and there is an even more deconstructive reason for understanding, at least in concept, most commonly used aspects of numeracy, like financial accounting, business valuation techniques, stock option valuation methods, etc. The reason is captured in the old saw: "lies, damned lies and statistics." Sorting out methodology from the assumptions underlying the methodology is a critical analytical skill.
Two examples. (1) there was a paper out recently using D&O insurance premium dollars as an indicia of risk: that is firms paying more for D&O insurance over time indicates a greater risk of liability on corporate fiduciary claims. I mentioned to the study authors that using gross premium dollars of D&O premium as indicia of risk, without regression out of other factors influencing premium levels, may be problematic. Other factors include supply side issue - are more or fewer firms in the market, types of coverage (premium may go up or down because the D&O product itself has changed), or specific claims histories of individual firms. (The problem is they can't get the data they would need to do the regression. (2) Business valuations, say on a discounted cash flow, notoriously manipulable based on things as simple as the assumed multiplier in the terminal value, or the discount rate, both of which at the root of it, are fixed by judgment, not scientific precision. And the cash flow itself is a projection, and subject to the infamous "hockey stick" graph (flat now, sharply upward later).
There are philosophical issue as well imbedded in the use of data bases to find meaningful patterns, and that is whether we bring the pattern to the data, or the data brings the pattern to us. (I'm reminded of Annie Hall and Alvy Singer each complaining to their respective therapists. Annie: "He always wants sex: three times a week." Alvy: "We hardly ever have sex. Only three times a week.") Moreover, the leap from the historical data to the decision what to do is laden with issues of rule following and rule skepticism, as well as Hume's observation on the conflation of the "is" with the "ought."
All of that is to say statistics and algorithms and quantitative analysis in the realm of human interaction are tools to be used, debated, critiqued, and argued, but there's been no fundamental shift in the world by which statistical analysis provides a shorter route to the TRUTH (notwithstanding the hope of many that it would be so!)
Posted by: Jeff Lipshaw | May 25, 2006 11:04:14 AM