I've been involved with the LSAC since
1989. And so I've learned a lot about the test. And hopefully we'll be able to
share some of that with you today. Yeah, I see a lot of friendly faces, good! I
was also asked to talk about this article by Conrad Johnson that appeared
in The New York Times that's entitled "Law School Admissions Lag Among
Minorities" and what was surprising reading this article when it initially
came out was that the New York Times published it. It didn't really say
anything new. It's somewhat misleading in that it compares 1993 to 2008. In
1993 -- I'm sure not many of you were in the market then -- but 1993 was a huge year
for law school admissions. It was one of the biggest if not the biggest year. And
so if you compare '93 to 2008 and don't compare the intervening years you get
some distortions. And if you care about minority admissions between 1993 and
2008 what you'll see is that there've been some ups and downs but essentially
African-American matriculants peaked in 2008. And they have gone down slightly
-- well, they went down slightly last year. And there was a
little peak also in 2006. They went down again in 2007, so a little peak-and-valley action. The numbers roughly -- and I'm just throwing it -- roughly about 3400
African-Americans are being admitted to ABA-approved law schools. And the number
doesn't vary that much and so we're talking about small differences even if
you compare 2008 with 1993 What's problematic -- and the article does state
this and deans and others have been aware of this -- is that the number hasn't
grown while the seats for law students has grown by about 10,000 since the
1990s because of all the new law schools that have been given ABA approval. And
that is somewhat troubling that the numbers have not kept pace with the
number of additional seats. And so for that this article is correct. It's also a
little bit misleading because it talks about the decline in
Mexican-Americans and what he doesn't talk about is that
basically between 1993 and now with the 2000 census we of course -- like everyone
else -- went in and developed new categories. And so we now like the census
have 15 boxes and so you can check boxes and whereas 1993 you only had five
boxes now you have a lot more boxes so actually the data at least with
respect to race -- especially with respect to Hispanics -- is non-comparable and so
that's another thing I would caution you about. The last thing,
and I'll definitely get into this. The thing that troubled me is that in this
article he says -- and I quote -- "Even though their" -- he's referring to African- Americans
and Mexican-Americans. "Even though their scores and grades are improving and are
very close to those of white applicants African Americans and Mexican Americans
are increasingly being shut out of law schools." That's flat-out wrong. And
I'll talk more about that so I don't know where that statement comes from
there's no cite for it and I don't know what metric he's using to say that the scores
are the same giving -- basically the effect that African-Americans, Mexican-Americans
are scoring the same on the LSAT but not being admitted. As I'll talk about
there's still a significant score gap and a documentable score gap so I was like
where did that come from? Anyway, okay, so, where to start? I wanted to tell you a
little bit -- couple minutes -- about the history of the LSAT to give you some
flavor of the organization, how it started. It's not this big, you know, corporate
entity. If you think about testing organizations there's ETS and ETS is a big
nonprofit. They do a lot of the tests -- GRE, GMAT -- and they're located in Princeton,
New Jersey. The LSAC was incorporated in 1947. And it was incorporated by law
school deans and indeed one of the founding law schools to incorporate the
LSAC was the University of Virginia and one of the individuals signing the
initial articles of incorporation was an old dean -- Emerson Spies, the late
Emerson Spies -- and the problem with this -- In 1947 the GIs were coming back from
World War II and they were applying to law school. And before World War II if
you went to Princeton, if you went to Harvard, if you
went to Yale, those sort of schools, and you applied to law schools and you had decent
grades, meaning C or better, you got into law school. And then after World War
two you had this flood of applicants and these applicants came from all over. They
went to schools like Poughkeepsie State, UCLA, schools that admissions officers
had never heard of and they had no idea how to evaluate these students based on
their grades because they didn't know the quality of the schools. So what
happened was if you went to one of these quote "elite Eastern schools" you got in.
If you didn't, you did not get into a law school in 1945, 1946 with the GI Bill. And
that created some issues because the people who went to the elite Eastern
schools were all WASPs and so there was no ethnicity in law school, there was no
way to sort of include others -- and now I'm not talking about African-Americans
I'm talking about Italians, Jews, Greeks whatever, and so they these people got
together -- law school deans -- and they came up with the LSAT. And they came up with
the LSAT as a tool of inclusion to expand, if you will, the entering class's
ethnicity, to reach out to groups that had been excluded and it worked very
well from 1950 to about 1970 when the issue became well how do we get more
people of color into our law schools and I'll talk about that. But essentially -- I think this is so true -- the LCC views this organization as one that is
beneficial to underrepresented students and underrepresented students today are
African-Americans, Hispanics or Mexican-Americans, and Native Americans. Asian-Americans are no longer underrepresented and that's why they have a standing
committee called the Minority Affairs Committee and that's why they do things
like the fee waiver and some other things and we can talk a little bit more
about that in the Q&A. But essentially it's a tool of inclusion and they view
the LSAT as providing opportunity for those who can't go -- or choose not to go --
to Harvard or Yale or Stanford. They choose to go to Georgia State, they can
prove that they belong in law school by taking this test. And so that's sort of
the framework and that's how I sort of view the test. And I took the test -- I
won't tell you when -- but it was way before 1984. Anyway,
so the test -- oh and I should add, the LSAC is still owned by the law schools so
if you're an ABA-approved law school -- and they're now 200 -- you get to vote on who
goes on the board of directors. The board of directors, of course, is the governing
body. The board of directors hires the president, the CFO, the CEO and all that
stuff and there's a volunteer organization and as Melinda mentioned
I eventually became chair of the board from 2001 to 2003. So still a law school
entity and so it still cares very much about law schools and what law schools
think about the test. And we've discovered over the years some issues
that create some of the problems that lead to, well, less than adequate
number of African Americans and Hispanics in law schools. The first
thing is there is a score scale differential and it is greatest between
whites and blacks and it's typically on, in most administrations, one standard
deviation. So if you look at the LSAT, 153 is roughly the average score. The average
score for African-Americans is still about 143 -- it went up to 144 but it's
roughly 143 -- you can play around with the numbers a little bit. Now what's interesting about
that -- and people point to that score-scale differential, which is the way the
psychometricians -- you all know what the psychometricians are? Psychometricians are
people who will go to very interesting colleges, get PhDs in test taking. And
they study statistics, sociology, psychology, it's a very specialized area of
expertise but those are the people that put together the LSAT and the GRE and
the GMAT and all those other tests, and the SAT, that come up with the item types
and do the score scales and -- I know a lot of psychometricians but anyway, the
psychometricians would say it's one standard deviation and the persistence
of that gap is real. At one point it was about one and a half. And so the LSAC is actually quite proud of the fact that they, I think, have the narrowest gap and
that they are narrowing the gap from one-and-a-half to down to one. So when you
look at standardized tests you see this
persistent score gap in every standardized test in the United States
today and indeed you can document it as early as second graders taking
standardized testing. Now here's the problem: no one knows why.
If you assume that race is not a biological attribute which I definitely believe it is not,
I've written a lot about that, then you cannot point to race as the rationale for that gap. That gap persists along what we call the SES, the socio-economic
spectrum. So the wealthier you are, the better you tend to do on standardized
tests and obviously, if you think about it for a minute, that makes sense. But
even among the wealthiest blacks when you compare them to the wealthiest
whites you still see the gap. That's why I say the gap is persistent across the
SES spectrum and that leads to a claim that either the LSAT's
discriminatory or all standardized tests are discriminatory. And I don't know
about all standardized tests, but I can tell you that I would not be involved
with an organization for over 20 years that is discriminatory to African-Americans. But that's not why it's not discriminatory. There are other reasons. First is -- and you would know this -- for
every item on -- and they call questions on the test items -- so for every item, they do
what's called DIF analysis. You remember the experimental section?
That was in your tests? All of you took the LSAT right? I'm starting with that knowledge. And you
didn't know -- you weren't supposed to know which one was the experimental section?
Although we can -- some people actually can figure it out. We get some tests where
they answer every question correctly and then on the experimental they just go
"E." We're trying to figure out how they do that that's another story. So on the
experimental section that's where we try out new items and what they do is they
test the items and there's this thing called a Hensel-Metal -- it's a mathematical
formula -- where you take basically someone of this group, say, gender, woman versus
man of the same ability and see how they score on an item and if there's a
difference like men get it better, men of that ability get the answer 80% of the time, women of that ability get the answer 60% of the time, they throw it out. There's a statistical difference in the
way it's answered based on a comparison between groups you throw out the item.
And so they do that for all subgroups. They do it for gender, they do it for
race and they do it for ethnicity because we call Hispanics ethnicity and
not race. And while I was president -- and when you find one of these it's a big
deal in the experimental section because they're very rare. When you find one of these they actually, you know, have to bring it to the board and during my -- I was, well, I
was both president-elect and president for two years in the past -- so four years we
had two items that showed up with the DIF analysis that we kicked out.
One benefited African-Americans, one didn't. And it's kind of interesting, the one
that did, I can't tell you because they might use it in -- but definitely benefited --
anyway this question so, they do DIFF analysis so on each individual
item they can show that there is no differential outcome based on subgroups.
Second thing: if you look at, once again, all the other standardized tests, you see
the same sort of gap and therefore, in order to prove discrimination,
you'd have to assume that all the test takers get together for every test --
architecture, nursing, GRE, whatever -- and they all get together and collude --
and believe me, these people are not smart enough to do that. And so it's a standardized testing issue not LSAT issue. The third thing, and this you're not going to like, is that if the tests were discriminatory, it would have a different
predictive validity. And so if you could show that African-Americans would
actually do better but the test somehow masks their true ability because it
discriminates against them, then it would have a poor predictive validity for
African-Americans. Unfortunately it doesn't. It actually does better for African-Americans than it does for whites. And so when you look at the test and we look at
those three factors: the fact that DIFF analysis, the fact that all
standardized tests have this issue and some a lot worse, and the test has pretty
good predictive validity, even better for African-Americans, it's not
discriminatory. In fact, I would argue the LSAT is a very good test. One of the best if not the best sort of -- what they call
high stakes or power test -- because it has a correlation of 0.43 which is very
high and when you couple that with the UGPA it exceeds .50 which is
pretty good for a standardized test. I can give you, you know, rattle off the GRE or GMAT but they're less. So it's a good test.
The real question -- I don't have an answer to it, no one has an answer to it -- is why does -- well when you go through
DIFF analysis -- why then do you have the score-scale differential
between whites and blacks who have the same UGPA, who have the same
socioeconomic status, who go to the same school, have the same major.
You don't know. No one knows. And if I knew, I'd win the Nobel Prize. I'm serious.
And so if I knew I'd tell you and I would fix it! And it's not as easy, the -- most people don't know
this but the SAT gender normed its test back in the '50s because they were seeing
a differential between males and females and so they consciously went in and said
let's make sure they come out the same. You can't do that with the SAT it's not --
that's not an issue. I mean the LSAT. And no one knows why these, these score-scale
still persist. But it's there. There's some hypothesis that it has something to do with being subjected to discrimination, there's some anecdotal evidence that
French Canadians in Canada who live in Toronto don't do as well as the British-speaking and go to Montreal it flips, and so some stuff like that but nobody
knows and so I can take questions about that but I don't have an answer.
It's there. The gap is there. So what does that mean: .43 correlation?
.43 correlation -- you have to square it, this is what the psychometricians tell me -- and
that means that if you square 0.4 16% of your first-year LGPA --
law school grade point average -- is attributable to the tests. 83 or 84 %
is attributable to other factors: how much you like law schools, how much
you study, whether you get along with your colleagues, whether you're having
fights with your loved ones, and when you add basically undergraduate GPA, it comes
up to about 20%. So basically 80% of grades that you achieve or students
achieve are attributable to other factors but that 16 to 20 % is still very high
for standardized tests. And what does that mean if you have and -- let me see if I can do this right- -
if you had a class made up hypothetically of 100 students and 25 had
170, 25 - 160, 25 - 150, 25 - 140, at the end of the first year -- I'm going to get this a
little bit wrong, I'm going to get some criticisms from my psychometrician friends but 12
of these people would appear in the top quartile -- let me see, I think I got this
right -- six would appear in the second quartile of grades, four and three. And so you
can see being in this group your odds of being in the top quartile in terms of
your grades are very good. The numbers flip in the bottom quartile: three, four, six, 12.
So even if you're in the bottom quartile, three of these people will probably -- excuse me --
even if you're near the bottom rank, three of these people will be in probably the top quartile. Although
we'll see that doesn't really happen because we don't admit those people.
So basically the test will tell you that a certain percentage of these people will
be spread along the range and will do so well but it doesn't tell you which 12. It
doesn't tell you which three are not going to be in, they're going to be in the
bottom quartile. And that's the other factors that build into, well, how well
you do in law school. And what the LSAT people do -- let's see, I used to know this.
I no longer am involved in admissions here at Virginia so I can't speak to it directly.
When I was dean at Minnesota, obviously I was involved in admissions there. Every school uses an
index. I'm trying to remember what the index is here.
It used to be 50 to 58. And the index is basically a mathematical formula which
combines and weights your UGPA and your LSAT score and so -- ah, the old
one -- I don't think we use this one anymore. The old one here at Virginia was
weighted 60% I believe for UGPA and 40% for LSAT and they went through a
mathematical formula and it gave you a number between -- don't hold me to this --
but 48 and 60. And maybe it was a little bit lower than that, I think might be 40
to 60. And what happens is, when I was on admissions if
you were a 53, 54, 55, and you were in-state you got in. If you were 55, 56, 57, or above
you got in if you're out of state. And what the LSAT people do is they take
that number, your index number, and then they take your first year grades and --
grades you achieve -- and what you should see on a very good test -- oh by the end of the
first year the Dean gets a report, it's called a correlation study and you guys are
just numbers and it shows for everybody that's in the school that had an index
of whatever, this was 50, the grade should be down here. this might be 2.3 and this might be 2.6, 2.9, you get the picture, and if
you had a very good correlation back -- if you had a perfect correlation your graph
would look like this: all of your students would be graphed along this
axis and so it would show you that the student who had the highest index had the best
grades. And what actually happens when you get those, and I got those when I was
Dean at Minnesota, you see very strong correlations. You see outliers sometimes.
You see things over here and things up here. But there's a lot of bunching
around here so that tells you it's a very good test.
And it's one that people rely on. So it demonstrates predictive validity. The one
problem which I'll come back to when we talk about admissions is that the test
has even stronger validity -- back up -- you all remember the test had scores between 120 and 180? You don't want to get a 120, right?
You sign your name you get a 120. Now here's what's interesting. I looked this up.
You remember this too, right? US News and World Report? Virginia
admits between 166 and 171. That's your 25th, 75th is 171.
Now I haven't gone in and looked, but I would hazard a guess, and I'm going to make the contention that
there's -- this is a guess, so don't sue me -- but I would -- I can't guarantee it -- I
strongly believe that there's no one in the law school who got below 155. Strongly
believe no one in law school got it. In fact I could probably make the case
if there's no one in the law school who got below 160. We'll talk about why in a minute.
Now if that's the case, if there's no one in the law school got below 160, and if in
fact 75% of the people above 25 -- in fact I'm going to argue probably 90% of
the people got between 166 and 171 or better -- then you have a problem with this.
And the problem is something called range restriction -- it creates
a skew. The range restriction is the test is based on 120 to 180 but all of you
are between 166 and 171. So even though overall because you're at the top in
terms of the score, you're all fighting each other for grades and we don't have
a bottom. We don't have the 120s, the 130s, the 140s. So that chart that I had on
the board before is invalid. You have all these 166s and 168s fighting it
out and somebody's got to be on the bottom and somebody's got to be on the top.
So when you do see variations it's usually -- not usually -- but a lot of it is
attributable to what they call range restriction because you don't admit 120 to
180.
It's kind of fascinating but we've -- as an aside -- we're giving
the test now in Australia and some other places and they do have a whole range of
people who take the test and in Australia law is an undergraduate
discipline and there you see much higher validity because you have an entire
range of applicants taking the test and matriculating at the various schools and
getting grades, so the 125s, 130s are on the bottom and the 170s and 180s run on
the top. So the LSAT's a good test one of the best. It's not the end-all be-all, it's
only a number. Maybe it counts 16% of what it takes to get a grade and so that
leads to the question: what's the problem? Why have African-Americans not been
admitted to law schools at a higher rate given the increased number of seats?
Two problems. One: Gruder -- and Gruder was a victory. But before Gruder was decided
the deans all thought -- I shouldn't say all -- most of the deans thought that
Gruder would come out the other way and so many of the deans changed their
admissions processes to reflect basically a race-neutral admissions
process.
Some were required to, under their state law like Michigan, like
California. But many of them, being fearful of potential lawsuits, just
changed their policies so that race was no longer a variable in their process.
And even after Gruder those deans -- many of these Dean's have chosen not, for
various reasons, threat of lawsuits, some have a belief in fairness not to take
race into account leading to a race-neutral process and then when you
couple that with the differential and scores African-Americans overall are
disadvantaged compared to white applicants. The second problem, which I
think is the real problem, and not that problem is this one, US News, okay? And US
News is the bane of most deans' existence. Deans have been fired,
literally, over not improving or falling in US News and World Report. And if you
care about US News and World Report it contains a number of different variables
upon which schools are evaluated. The metric that US News uses and the guy who
puts this together -- he's nice guy, he comes to the deans meeting, he explains what
they're going to do -- you look at the metric. When I was dean -- you try to
manipulate the metric so that you can take advantage -- obviously
not game the system but achieve the best rating you can. The best -- er not the
best -- the heavily most heavily weighted attribute of US News rankings is
reputational. 40% of it's reputational.
I think it's 25% by -- 25% by judges and lawyers,
15% by academics. The next most heavily weighted component
is admissions which comes in at 12.5%. Now here's what
happens. Around this time, in fact it should be starting about now, all of the
law professors in the world, especially in this country, and all the deans in
this country get flooded with what we call law porn, okay? And law porn is
glossy brochures and cute things from other law schools saying how great they
are, you know, like this is the best program of Trial Ad in the world and
here, glossy pictures of X Y and Z. This is you know, here are our grads, they're all employed
they're all doing great things. Every school -- I can't say every school, I
can't prove that -- most schools in the United States spend thousands of dollars
on this stuff. Guess where it goes. Circular, okay. Everybody throws it away.
The people spend money on it and people try to improve their reputation but
because of some things called echo effects and some other things, doesn't
work. The one thing that deans think they can control is that admissions
variable. And there's three reasons why they think they can control that admissions
variable. The first is cost. To improve your reputation -- another big variable also
worth 12.5% percent is the resources that you spend per student. So
there's actually a calculation: you take a budget, you divide it by the number of
students, and it gives you roughly how much you spend per student and
that's a big one. Here we spend about -- I'm guessing, I used to know the number. Here we spend about $65, $70,000 per student. So see, you're --
even though you're paying $40,000 it's still a good deal, right? And at Yale they
spent about a $150,000 per student and Harvard they spend 100 so
that goes into -- $70,000's not that bad. Minnesota we only spent about 40. It could be worse you could be on the other end of the ledger -- we're spending 20, you're paying 40. So you're getting -- you're getting a good deal. If we -- if you want to increase that then you're talking millions of dollars. If you want to
increase your reputation you're talking millions of dollars. But when you're
talking about increasing your median LSAT score or your 75th and 25th
percentile, all you need to do is tell your dean of admissions,
"I want that score to go up. I don't want you to admit so that we wind up with the
167 next year and a 172." It's very easy to do. It's very mathematical. You get
these applications in. You admit people they accept and then you keep people on
the waiting list and eventually you wind up with the numbers you want at a very
low cost. The other thing is certainty. These numbers are certain. You can, as a
dean, say "I want to have a 167 next year and a 172 and if you don't" -- I've never
had this conversation with my admissions officer -- "you will be fired."
I know deans who've had that conversation with their admissions officer, their
admission staff, and those numbers had better go up or -- but it's certain. You
look at the numbers. You admit a class if you admit my hundred -- what this means -- if
you admit my hundred students, the 75th student has to have a 166 and all the 75
students that you've admitted or 74 students you've admitted before her have
to be 166 or better and you can do that, it's very easy to do and I can do it. I
did when I was dean I did it and so you can get that number and achieve certainty. You cannot achieve the same certainty with things like reputation. There's no way
you can you know gauge whether you're really affecting your reputation, whether
your message is coming across or whether you've made a tremendous mistake.
So deans like that number. The third thing: it's correlation. We call it
three C's. At one point although -- I noticed this year was a little bit off --
there was an exact correlation between the school's rankings and the LSAT score.
And deans see that. And so -- where is the page here -- schools of law, Yale: number 1. '08 LSAT score: 25th and 75th percentile: 169 and 177.
Harvard: 176. And you go down the list and so there's almost a perfect correlation.
There wasn't this year but there has been in the past. There's one school -- I won't
say which one, I better not get in trouble -- that you
don't have the correlation but they did something different. So anyway, deans see
that and they they say, "Well, we can effect that number." Here's what happens. If you
want to shoot for this number and you say this is our 25th percentile
number, what that means if you're admitting 24 students below that in my
hypothetical hundred-person class, you can admit whomever you want with
whatever LSAT score you want and they will not affect this number because by
definition your 25th percentile is 166. So you would think that would
give you a range to admit people from say 140 to 166 and you can do whatever
you want on that 25 percentile but people don't do it. Deans don't do it.
Admissions officers don't do it. They don't do it for three reasons. One:
they're afraid of being sued. If they admit only students of color in this
range they might get sued. Two: they like the
numbers. They don't understand math. That's why they went to law school. You
can tell deans until you're blue in the face, "You can experiment below this
number." They don't see that, they see, "Oh, my 25th percentile is 166. Well I
better not admit a 164, that'll affect it. Well, I better not admit a 150, that'll
affect it." And you'll say, "No, no, I'm sorry. I know you don't have a math degree but
you can do that" and they say, "No, no, I have to admit 165, 160.
You will see schools -- and I've seen schools -- where if they had these numbers
there will hardly be anybody below 166. the 25th percentile will be
basically the first percentile and these are -- and they just get caught up in that.
They also think that if they do it this way, if they admitted 172 it will change
it. The other thing -- and maybe this is perhaps a good reason -- and we can talk
about this in the Q&A -- you clearly don't want to admit a 140 to compete with all
the 166s, so at some point, if you're a principled dean, even if you know that you
can admit below the 166 and that you should admit -- especially if you want to
diversify -- if you want to give people a chance, you're still going to try and stay
close to these numbers because you don't want to put anyone at a competitive
disadvantage, and say, "Gee, you're my 150. You're the only 150 in the class.
The odds are almost overwhelming that you are going to wind up in the bottom
quartile of this class," might you be better off going to another law school
where your median -- where your 150 is close to the median? Or maybe even at the
top? And so a principled dean could come up with a rationale for not diversifying
along the spectrum. There are some principled deans. Most of them, however,
just don't know the math and they just try and stick close to this number. And so
that's what happens in the admissions process. And so the impact on admissions,
when you couple the score-scale differential together with US News &
World Report and this pressure for the rankings -- when I went to law school there
was no US News, it was very different how you chose a law school. I've had students
come to my office during admitted students week and say -- but one year -- I
guess we're number 10 now, but one year we were number eight and NYU was number
seven and they said, "Why should I come here, you're number eight, NYU is seven" and
I'm like, "hello this is Charlottesville, that's New York. Isn't that a reason to
try and compare these schools?" No, they're number seven, you're number
eight. And that pressure is just -- you know, I won't to ask you how many of you came
here because this was the highest ranked school you got in. But that has
changed the game. The metric has changed we, unlike other schools, have one
evaluator. We're unlike business schools who have four, six different value
measurements, they measure on different metrics. We're also -- law schools are
unlike business schools or colleges -- we're all the same, and we're all the
same because of something else you don't know: the standards and rules for a procedure for
approval of law schools, the ABA section of legal education and admissions to the
bar. This is the group that the Department of Education has tapped to be
the accreditor under federal law. The Department of Education has to have an
accrediting body. The Department of Education decides for each type of
school who will be the accrediting body. For us, it's the
ABA section on legal education and the ABA section on legal education basically
mandates -- if you want to know why all the 288 approved schools look pretty much alike,
it's because of this. And so, since we all look alike, since we all teach the same
thing, since we all use the same admissions process, since we all have the
same curriculum, we're very easy to measure.
We're very easy to compare. We're very easy to say, "Oh, Yale is number one,
Virginia's number 10" and differences turn on that. And so that's really part
of the problem. The other problem is: once you know what these numbers look like --
and I don't want to depress you but I used to have the exact numbers and I emailed my
head psychometrician at Newtown this morning to get the exact numbers but
guess what I discovered. Unlike here in Virginia, the rest of the world is on
holiday so I was assuming I'd get these numbers, they never came, because it hit
me, oh, my god, they're closed today. Because they even called after I didn't
get the email reply because as being former chair there's some respect -- anyway,
so I called him and I said -- I emailed him and I said -- Peter, his name's Peter
actually, "I knew these numbers five years ago, I don't know now.
Tell me the number of African-Americans who scored above 170 in the last
administration cycle." So the last administration cycle would be the one -- I
can't remember now. It's February to the four cycles and then -- and the number used to be -- the last time
I checked -- it was 125 but I'll be charitable, say it's 150 so 150 above 170. And then you have -- last
time I looked -- then you have about another 250 above 160 and then you have
about 500 above 153, which is the median. and this is out of 7,000 test takers
roughly. And so what you see here if this is our pool here at Virginia.
This is what we're looking to. These are guesses -- they're not that far off. I mean
it's not hugely different, it may be some up, some down. We're looking at basically
400, maybe 500, students in any year take the test who do above 160 and you're
looking at about 800, 900 who do above the median. If you think about the top 20 law
schools -- Minnesota, number 20 -- their median LSAT was 160. So
everybody's chasing these students. Everyone is chasing all of these
students and all of these students are receiving, no doubt, hopefully, great
financial aid packages and -- that's another story. We've also changed the way we finance legal education. The students who are not sought after basically pay
for the other students. That's why tuition's $40,000. It's the have and
have-not model of financing. So you have all -- half your class usually on free
rides, the ones that you want: the high scores, those who diversify. And then you have
the other half: the low scores, the ones who don't diversify who pay for the rest.
It's a fun world. All due to US News and World Report. The
last piece of this -- and I'm running out of time -- I want to leave ample time for
questions. And all this -- in fact I should have mentioned this -- I wrote about all
this in an article and the article is -- let me give you the cite in case you care to read
about this -- "The Destruction of the Holistic Approach to Admissions: The
Pernicious Effect of Rankings." And it's in 81 Indiana Law Journal 309 in 2006. So
all that's laid out with some empirics and some -- obviously more -- detail in that
article. I've just finished a new article that's going to appear in the Journal of
Legal Education and in that article I run through some other numbers and what
I discovered is that roughly -- I don't know how to approach this as quickly as
I'd like -- if you look at the individuals who take the LSAT, after they graduate
from college, it is clear that law schools get their fair share of
interest. And let's see, I wrote it down somewhere -- in any given year, 22.5% of bachelor's degrees are awarded to members of
underrepresented groups with 9% to African-Americans. The good news is
32.5% of LSAT test takers during 1999 to 2000, which is
the last year we have statistics, 11.5% were African-Americans. So we're
attracting more African-Americans to take the test than are graduating from
law school -- excuse me -- from undergraduate schools. Of that
number -- and that was nine thousand four hundred and something -- of the number of
test takers, roughly 75% choose to make an application to a law school. And
when you look at the numbers, that means 7300 basically
choose to apply to law schools and that number is comparable to both whites,
Hispanics, and Asian-Americans. So if you look at the pipeline for students of
color, at that point in the pipeline the numbers aren't skewed, they're about the
same proportionately. So about 25% of people who take the test, whether they're
white, black, whatever choose not to apply and they choose not to apply for a lot
of reasons: they get a low score, they got a better score on this test, they
got a better opportunity, they're going to business school and maybe applying in two
years. But they don't apply that year. That makes sense so far. Okay, then what's
interesting is if you look at the 74,500
people who applied in the year 2000 -- once again, that's the year that I have data on -- when
you look at the people who are admitted, 65% of the white
applicants were admitted but only 54% of the Hispanic applicants and
only 43% of the African-American applicants were admitted. That's where
you see the drop. That's where you see the gap. And so we're getting applicants --
we're gettomg people who take the test, which obviously you have to do -- you're getting
applicants but you're not getting matriculants and I don't have enough time to go
into this but in this article that's coming out in the Journal of Legal Education this year -- I don't have a cite obviously -- one of the things that I try
to document is that African-Americans disproportionately, when compared to
whites and even Hispanics, misapply to law schools. They apply to the wrong law
schools. We get, for example here in Virginia -- at least when I was on the
committee -- we would get over 200 applicants every year with below a 150
LSAT. And although we don't have a cut-off score --
nobody has cutoff score because that's not good -- not one of those people got in.
And when I was dean in Minnesota I saw the same phenomenon. Our median
was 163 and we would get applicants 143, 148, 150. If those same applicants had
applied to -- there are two other schools and, well three now in Minneapolis:
William Mitchell, Hamlin and St. Thomas, the three of which basically had medians of
155 -- those students would have gotten in to any one of those three schools but by
applying to Minnesota, you're not going to get in. And what I
can show, what I try and demonstrate in this article is that people misapply.
They apply to the wrong schools: schools that they have no shot at getting in. And
then they're given bad advice. They're told: if you don't go to Harvard, if you
don't go to Virginia, if you don't go to Minnesota, you shouldn't go to law school.
You shouldn't go to a second tier, third tier or whatever tier law school. You
can't get into the best. So today actually there's -- we joke about this -- well
I used to joke about when I was dean -- there's a law seat for everybody.
In fact, some schools have some vacant seats. And so we say, there's a law school
for everybody. But you have to pick the right law school. And so one of the
things that I attempt to document is that African-Americans are getting bad
advice. I don't know if it's a factor of their belief that affirmative action means
that irrespective of the LSAT you'll get into Harvard or Virginia or Yale or
Michigan or whatever. Well, no. But it's clear that's also part of the problem.
And if people were given correct and actually good advice they'd be better
off. There is some hope -- and this I'll conclude with and leave 10 minutes. All of this is
created by a standard in the ABA section -- in fact, I could have started with this --
there's a standard in here called 503 and 503 -- let me read it to you --
ah, here it is: standard 503 says a law school shall require each applicant for
admission as a first year JD student to take a valid and reliable admission test
to assist the school and the applicant in assessing the applicant's capability of
satisfactorily completing the school's educational program.
It goes on. But then there's an interpretation: a law-school that uses an admission test other than the law school admission test sponsored by
the LSAC shall establish that such other test is a valid and reliable test.
So basically, if we put those together, that means every school has to use the LSAT
and that score gets reported to the ABA and is therefore verifiable. And that's why that score is in here. What the ABA section is debating right
now is to eliminate the LSAT from the admissions process. There is a huge
debate and apparently I've been told that it's going to pass: that they're
going to eliminate standard 503 and no longer arer schools going to be
required to use an admissions test. We're no longer -- they're going to be directed,
if you will, to use the LSAT. And I haven't decided whether that's a good or a
bad thing. Because we may go back to a world where
it's who you know and where you went to school and, you know, who your friend is and who your daddy knows as opposed to, you know, can you perform. So it could be a
whole new ballgame in a couple of years, and maybe as early as this year and all
this may be moot. We'll see.
Yes, definitely. This is from some years back, and in the older days they didn’t mix up experimentals in the manner they do today. So there were ways to figure it out during the exam if you knew what to look for. These days it’s much, much harder to tell during the test (and wouldn’t really abide by their rules).
Crazy motherfuckers. What if you were wrong (even if just a small chance)?
I figured it out on the Nov test half way through