The Southern District of New York recently became the nation’s first federal court to explicitly approve the use of predictive coding, a computer-assisted document review that turns much of the legal grunt work currently done by underemployed attorneys over to the machines. Last month, U.S. Magistrate Judge Andrew J. Peck endorsed a plan by the parties in Da Silva Moore v. Publicis Groupe — a sex discrimination case filed against the global communications agency by five former employees — to use predictive coding to review more than 3 million electronic documents in order to determine whether they should be produced in discovery, the process through which parties exchange relevant information before trial.
The task of combing through mountains of emails, spreadsheets, memos and other records in the discovery process currently falls on a legion of “contract attorneys” who jump from one project to another, employed by companies like Epiq Systems. Many are recent grads who are unable to find full-time employment, or lawyers laid off during the recent recession.
Here is more, and for the pointer I thank Fred Smalkin.
And it’s about time.
Perhaps the predictive coding method could be applied to the process of deciding cases itself — after all, judges are deciding cases by basing their decisions on previous cases and by applying statutes and relevant regs, something a computer program like Watson should be able to do
Yes, “predictive coding” or some other form of artificial intelligence will make further inroads into the legal profession, and eventially AI will be able to perform the judges job better than humans. This is just another example of the relentless displacement of humans by robotics, AI, and other smart machines. Eventually humans will be made economically obsolete. Ray Kurzweil is forcasting that the “Singularity” will occur about 2045, but most of us will be economically obsolete before the singularity occurs.
I’m surprised that economists aren’t more interested in trying to analyze the brave new world that awaits us. In particular we should be trying to develop the legal and social structure of a nation where 99% of us are unemployed dependants of the amazingly intelligent and efficient machines. Perhaps a “robot tax” will replace the income tax?
We seem to have transitioned pretty smoothly to a world where the state supports you after 65, it’s just a matter of extending that downward. I think we’re headed for the John Barnes’ vision where AI does nearly all the productive work, while humans enjoy high standards of living earned through a period of (ZMP) work of 2 hours a day for a few years, which they will complain about bitterly. Most people spend all their time in immersive entertainment.
Oh, and eventually another civilization comes along and puts everyone’s heads in boxes, not because they’re evil but because they’re trying to help.
Are we approaching the point where a significant part of the cost of a email or data-backup platform is that of catering to potential legal discovery requirements?
Approaching? We’ve been there for some time now.
What Bill said.
This will hasten the reallocation of manpower from jobs that can easily be automated to farm labor which requires skilled human labor because technology has failed to find a way to produce tasty fruits and vegetables that can be massed produced and mechanically harvested.
By the way, I haven’t seen Tyler the foodie call out what he considers to be break through products that benefit from economies of scale, investment in capital, and investment in invention and innovation to reduce labor and ideally eliminate the human touch. Breakthroughs like Hostess Twinkies and Cupcakes, Pringles, Wonder Bread, Margarine, – all untouched by human hands.
Its my sense the market wants food produced with human hands, and these lawyers are ideally suited to farm labor because they chose a life of “a legion of “contract attorneys” who jump from one project to another” – an easy jump to “a legion of “contract farm laborers” who jump from one harvest to another”.
I don’t get it
Tyler, as a foodie, tends to advocate the least efficient foods, that require the most human involvement (fresh vegetables).
Although Mulp is speaking with his tongue firmly planted in his cheek, he is basically correct. Low skilled white collar work is shedding jobs very quickly, and low skilled, uncomfortable farm labor is trying desperately to increase the labour available to it.
No wonder the Mexicans have stopped illegally immigrating here — the contract lawyers stole all their jobs.
lol
I thought the whole point of his book was how big agribusiness was great and cheap foods could be excellent.
I got it. I loved it.
Thanks for explaining it?
Finally, the structural labor problems begin to resolve.
Sorry guys. As a farmer I am seeing first hand that many jobs we formerly thought could not be automated are now being automated. There are two new tree fruit pickers commercially available and the larger growers of “greens” are already automated. However, I get your point. It would be fun to see a field full of lawyers working away! We probably couldn’t afford the liability insurance to hire them though.
Perhaps those savings could be put toward luxury food items like heirloom turkeys and chickens which are labor intensive, but receive a premium.
BTW, this kind of thing for document review is not new, just new in this situation. Large contractors like Huron Consulting will use the filters to weed out a lot of unnecessary documents before contract attorneys ever see them. It can also happen that mid-way through a review the filter is revised, cutting short needed time for the contract attorneys. In Chicago, Huron is currently paying contract attorneys $29/hr. (overtime, too, if the case allows it)
A productive farm laborer can very often make $29/hr doing piecework — the problem of course is that no one who went to law school is capable of picking vegetables at that pace, and no one who grew up in the fields would make it through law school. So the labor is not substitutable.
I am being won over to the stagnationist view. In the early 1980′s, as a high-school dropout, I could make $35/hr programming computers.
I’m curious, what kind of programming was that? COBOL? Fortran? Pascal? FORTH?
I think the big labor problem in IT is basically WORM — code is infinitely re-usable at near-zero marginal cost. That means you get billionaires at ORCL, MSFT, AAPL, Facebook, etc, while the ratio of coding-hours (i.e., programmer employment) to value provided becomes very high.
Of course, you can always try to write your own WORM software — that’s how Mark Zuckerberg dropped out of college and became a billionaire programming computers not that long ago.
I have some trouble seeing the “problem”. Are you saying that the average programmer is under-compensated?
Their vast wealth seems to me to be the same as all others – they found or created a way to employ something at very high value. There is a lot of variance and the payoff is not necessarily proportional to programming skill.
Once software is developed, it can continue to produce profit at little marginal cost — unlike a steel mill, where you still have to buy iron ore, coal and limestone. At least that’s the theory — tell it to MySpace, Word Perfect and Lotus 1-2-3.
Well, the labor problem is that in programming, labor inputs can approach zero.
That doesn’t mean, of course, that there can’t still be large demand for programmer labor (there is). But it’s a limiting factor.
Basically you’re saying that the variable cost of producing a piece of software falls to near-zero after it has been released, right? Isn’t that a good thing from the point of view of the economy? If the value of programming labor is basically nil after the software comes out, that labor will be immediately deployed to new projects. In effect, every time a programmer gets a job, he’ll be working on a new product rather than simply “reproducing” a previous one. To return to TallDave’s original argument, while the coding hours-to-value ratio may be high in accounting terms, I’d expect that there’s a positive externality (in terms of innovation, the pace that the tech industry as a whole develops, etc.) to this model of labor utilization.
Contrast this to zbicyclist’s example of a steel mill where you have to purchase a unit of labor (plus iron ore, coal, limestone, etc.) every time you want to reproduce an existing product (i.e. steel). Would we really be better off if developers had to individually program every unit of Microsoft sold? What benefits does society see from this other than job security for programmers?
Err, every unit of Windows. Note to self – stop commenting after midnight.
The change has been going on for a long time, but this changes the quality of legal work from artistic interpretation to one of vector resolution. The machine pumps out 73.2% in favor of the plaintiff, and wham case closed.
Unfortunately, it ossifies the law into something akin to civil law, something that has been happening in the US criminal law for some time.
In the end it is just an arms race and digging for the nuggets will need to be done by humans. Much like having a chess program like Houdini, the lines are deep and amazing, but there are so many that at the board (in the courtroom) you still need to preform according to human standards.
Anyway, who needs lawyers.
We’re talking about document review, not real legal work.
This article seems to point to some analysis in the software as well.
http://www.nytimes.com/2011/03/05/science/05legal.html
And who is to say in the future this won’t be applied to analysis of legal cases?
At a time when the entire economics “profession” seems hellbent on discrediting itself as a collection of hacks too cynical or stupid to overcome their own confirmation biases, it’s hard to know if any of you can be believed or trusted. But when push comes to shove, I’m inclined to disregard the sociopaths that seem to take some kind of glee in that fact that thousands of people who have invested a heartbreaking amount of work and money into a promise if a better life are being made redundant.
If you wanted flowers and unicorns, you should not have ventured into anything that attempts an unbiased view of reality. If you prefer, we can outlaw excel and instantly create demand for thousands of low skill white collar workers.
There is no divine right of lawyers to unending six figure salaries and country club memberships just as there is no divine right of autoworkers to Cadillac insurance plans.
Obsolescence and creative destruction are a part of life, and legal education is a human capital investment that can have negative returns just like any other investment.
The only guarantees in life are death and taxes. Grow up.
Very few lawyers made six-figure salaries and dined on foie gras at the country club. Even fewer of them do so today. The distribution of salaries is basically a power law distribution: a small number make almost everything, a few do very well, and everyone else is surprisingly normal, or even well below normal. I know plenty of lawyers who would weep tears of joy for a $50k salary and health insurance. Not exactly dick-swinging power brokering.
“Grow up!” is all well and good, but most of us didn’t ask to be Fleet Street partners pulling in $2m a year. It’s just too much misery to get there. But all of us wanted to get jobs and make enough to have, at the very least, decent middle-class lives and the ability to service our loans. That hardly seems like irrational optimism or childishness, though it could to your standard half-aspie econ-worshiping right-wing twit trolling Asian bars for 15 year-old hookers.
I’m not being sarcastic, but if the average lawyer is making a not-so-good living why do so many Americans aspire to be lawyers? What’s the mystery?
Information assymetry or irrational optimism?
Yes and yes. Also, “lawyer” has a prestige that “IT consultant” does not, even if they make equivalent salaries.
I saw a self-employed plumber stiffen his back when a buddy said “Lawyer” even though the plumber made six figures and spent his weekends on his boat, while the buddy was doing DUI appearances and struggling to make rent and $1200/month student loan payments and keep his 15 year-old POS car running. The plumber is better off in every way, but the lawyer gets the prestige.
Aye, it’s a queer world.
Whatever information assymetry was there is disappearing. Apparently college students are realizing that it’s not that great a career.
http://blogs.wsj.com/law/2012/04/09/number-of-law-school-applicants-continues-to-slide/
Everybody wants to go to law school because law school is easy: no math, no science, no computers.
The part about lawyer prestige is confusing too: with the number of lawyer jokes and lawyer hate thrown around its hard to think of a lawyer as prestigious.
The other interesting part is lots of youth aspire to be models, filmstars or star athletes and obviously that’s a target with high failure rates too. Many end up as secrataries, baristas or some such. Somehow that cohort doesn’t garner they sympathy and attention that out-of-work lawyers do.
A generation ago, you could major in English and spend ten years trying to write the Great American Novel. Then, when that didn’t work out, you could go to law school and get a decent job — maybe in the District Attorney’s office.
Sure, you weren’t doing what you wanted to but you could make a good living, there was no heavy lifting, and you could still spend a few hours working on that novel occasionally. [two friends did exactly this]
That hasn’t been the case for many years, but the news got out to law school students slowly – maybe that’s where “Law Schools Lie” is coming from. Plus the tuition increases leave people far deeper in debt than in the days where you could work yourself through.
Irrational Optimism.
Starting lawyer salaries are bimodal: one mean is around $50k-60k and the other mean is around $160k. And the mean of means is around $110k. If you go to a top tier law school and graduate near the top of your class (and work on the law review), then you are likely to get hired by one of the top tier law firms at a large salary. If you go to any school below top tier, or you go to a top tier school and graduate below the top 10%, you’ll end up making about $100k less your starting year – if you can find work. Almost no one starts between $70k and $130k, but that is where the mean of means is located.
http://en.wikipedia.org/wiki/Bimodal_distribution
Irrational optimism leads the folks who can’t do math into believing that they will end up in the top bracket when even a modest amount of study will show that they won’t. One could call it the Lake Wobegon illusion: everyone thinks they are above average. Casinos are very profitable because people can’t do the math.
The mystery is that folks look at the “average” salary, and combine it with TV shows about what lawyers do, and the past prestige of the industry and they then make faulty decisions that result in them graduating with massive student debt. My local law school estimates that you’ll need about $40-50k/year for the 3 year JD program. They aren’t “top tier” but then they aren’t in the bottom half of law schools either.
And as a personal anecdote of how irrational this behavior is, at the end of one of my classes in business law (I’m working on a 3rd bachelors, this time in accounting to change careers *from* programming), while I’m chatting with the professor, another of the students comes up and says that he was going to law school and he wants to be a patent attorney. Me: “what is your undergraduate degree in?” Him: “business” Me: “do you have a minor in engineering, math or physics?” Him: (strange look) “no”. Me: “then you cannot possibly pass the patent examiner test because they won’t let you sit for it without a degree or minor in engineering, math or science”. As someone who has studied for the patent examiners exam, but didn’t take it, I explained what the qualifications were. This guy was too preoccupied with getting “good” grades and doing “good” extracurricular activities in order to game his law school application to actually do what he thought he wanted to do with his life. He had made it to the last semester of his bachelors degree without actually checking the required qualifications for what he thought he wanted to do as a living.
“Very few lawyers made six-figure salaries and dined on foie gras at the country club. Even fewer of them do so today.”
That’s a ridiculous statement and far from the truth. Your average American lawyer makes a six-figure salary.
From the BLS: Occupation Lawyer: 2010 Median pay was $112,760 per year
http://www.bls.gov/ooh/Legal/Lawyers.htm
From the folks that brought you the Birth/Death Adjustment and falling unemployment rates during a collapse of the labor pool, we get a number pulled from thin air.
Who was surveyed? How many? Is the reporting self-selected? Are we including lawyers who can’t find a job and are sitting at home unemployed? Are we including lawyers who have to work at Starbucks because there are no lawyer jobs?
Lies, damned lies, statistics.
I was at a CLE just last week hosted by the Louisiana State Bar and the President-elect stated that their surveys show 60% of the Louisiana Bar is earning between $50-$75,000 with the remainders above and below that range.
Remember, that’s Louisiana. Half the earnings of the Bar are in bribes, graft, drug peddling and murder-for-hire.
To second LawSchoolIsALie,
those figures show “terminal success”, ie the people who have “made it”. As a thought experiment, supopose there was a 3 year degree called “Play for the Yankees”. Of all who started, very very few would play for the yankees.
Some wouldn’t even pass the 3 year school, some would graduate but flush out right out of the gate to other, less “tournament contest” work, some would graduate but later be injured and leave employemnt at the yankees, some would be 5th string ballboy type positions on the yankees…and some would properly make it in all senses of the word.
so when “Play for yankees” school says, “the average salary of grads 5 years later is a jillion dollars”, they’re NOt lying, but they are also omitting that ~1% of the matriculated class is captured in that figure
$60k in Baton Rouge is equivalent to, what, $200k in NYC?
The mean is even higher:
http://www.bls.gov/oes/current/oes_nat.htm#23-0000
The methodology is described here:
http://www.bls.gov/opub/hom/pdf/homch3.pdf
They’re lawyers, not people. Just look at the post title, it’s a joke about murdering thousands of lawyer.
No wonder the Mexicans have stopped illegally immigrating here — the contract lawyers stole all their jobs.
If the cost of litigation decreases, you should expect to see the number of suits increase and the percentage of suits that settle early decrease. I’m not sure if that’s what TC meant by a good start. It’s certainly not what Dick the Butcher meant.
Tyler was referring to the old lawyer joke: What do you call 5000 dead lawyers at the bottom of the sea?
Yes, I was aware of that.
this will speed up student loan reform lickedy-split. Scores and scores of people who, almost definitionally, are a hundred thousand in debt AND are great as suing? It’s only a matter of time.
Creative destruction is all well and good. But the “destruction” part includes serious damage to people’s lives. The contract attorneys to be displaced by this technology are not making six figures, they are, almost by definition, the least successful attorneys in the economy. They cannot find permanent employment and move from temporary job to temporary job. They have no benefits and can rarely count on being employed more than 2 or 3 weeks out. Life is not good or easy for these people, and the loss of these jobs will only make it worse. And, they are not bad people, whose tough situation should be enjoyed by others. They went to law school yes, that doesn’t make them bad people.
I think victims of creative destruction would just like to see a little more sympathy from the more fortunate. Or at least, don’t laugh at our pain. That’s where the “sociopath” accusation comes from. Some of us understand the larger benefits of creative destruction, and we accept our lot, but its no small thing to deal with. It isn’t as simple as just “growing up”. That won’t help us pay our bills.
It depends on what kind of sympathy you are looking for. I’m all for unemployment benefits and job retraining. People make mistakes in life, including law school.
If your looking for the “I went to law school for three years and took out 150,000 in student loans, therefore I deserve a middle to upper middle class job” kind of sympathy, you’re not getting it from me.
You get sympathy from me, at least in the sense of a shiver of “There but for the grace of God go some of my family”.
As I pointed out above, lowering the cost of litigation should lead to an increase in litigation, i.e., demand for lawyers’ services. I’m not ready to weep yet.
Could they all become adjunct professors?
In the US, attorneys are the priests of the secular religion that’s called democracy. Naturally, many individuals wish to be ordained because this priesthood has the potential of high income, prestige and a small amount of physical effort that doesn’t require one to get dirty. As a whole, the legal community exacts an immense tax on the rest of the population, far in excess of that levied by the Roman Church in the Middle Ages, and, even though legal billings are part of the GNP, they produce nothing that we can eat, drive, wear, shoot or fondle. Centuries from now a saner society will marvel that we allowed these parasites to suck our blood.
What immense tax? The “immense tax” where people voluntarily pay lawyers to draft good agreements for them in case deals go sour? The “immense tax” where people choose to hire lawyers for complex litigation?
How do you propose we resolve disputes then? Swords?
Lawyers or swords? Yeah, those are the only available options for dispute resolution. Before Harvard Law probate was done at the point of a rapier.
I thought I made it clear, though perhaps I did not. I am not, nor are most un/underemployed lawyers demanding society give them a job. I am just saying, it would be nice if people did not voice pleasure at our suffering. You see/read that a lot on comment threads and not for any good reason that I can think of. You probably don’t personally know anybody in this situation, but if you did, you would take not pleasure in their situation. And….you very easily could. This isn’t something that just happens to us stupid people who don’t know better than to predict the future (as 18 year olds) and see what career path will lead to strong employment 10 years out. In this economy, you can’t do that. This could happen to anybody.
The idea that un/underemployed law school graduates (or college grads in general) are “demanding a job” is a straw man. If people are upset, it is because they were told a story about the value of education, a story about borrowing huge sums of money at very high interest to “invest” in their future, a story that did not turn out to be true. And the people who told them that story aren’t taking responsibility, but rather are blaming them for being stupid enough to believe it.
The story you laid out in your first paragraph is perfectly sympathetic — why did you have to add the second? Saying “don’t mock me for making a big bet that didn’t work out” is a lot different from saying “some amorphous group of people (or ‘society’) are responsible for things not working out – ‘they’ need to take responsibility!” The former is a totally respectable response to anti-lawyer gloating and prejudice, the second not so much.
I think the glee that you are seeing (and I see it too, and do not approve of it) is simply the flip side of prestige. If you are willing to accept the prestige that accompanies being a successful lawyer, then you must be willing to accept the disdain that accompanies being an unsuccessful lawyer–even if the reasons for the lack of success are beyond your control. You might not have pursued the law in order to reap prestige, but it is inevitable that some will assume you did. Thus it is inevitable that they will delight at your misfortune.
I think some of the delightees would disagree that “[t]his could happen to anybody.” It doesn’t happen to failed carpenters or plumbers or janitors. It happens to people who might be supposed to be pursuing an advantage in relative status–including, unfortunately, those who just happen to want to study law, with no designs on prestige.
I’m an equity partner in a large law firm and offer the following as food for thought.
Even back in 2006 or ’07 when we raised starting base salaries to $165,000 and were passing out $100,000 bonuses to the senior (typically 31 – 35 yrs old) associates who were billing 2400+ hours (taking some of them above $300,000 for the year) i think deep down we all knew that it couldn’t last. If truth be told, associates doing privilege and due diligence reviews produced almost nothing of value beyond billable hours. More senior associates did similar make-work projects – drafting discovery responses for partners who would never read them and churning out forms that could be assembled by a $15/hr clerk. In fact, times had been so good for so long that we not only had lots of associates who functioned at the level of a paralegal yet were paid as much as equity partners had made just a few years before we also had equity partners who functioned at the level of senior associates who were making $1,000,000+. And the associates were telling us that the market for first year associates was going to $190,000/yr within the next 12 months.
Instead we shortly had lots of ZMP associates and a bunch of grossly overpaid “service” partners. The former got outsourced to doc review companies while the latter got deequitized. As time went by we scaled back the summer clerkship program and some partners argue for its elimination. Meanwhile, clients say they’re only willing to pay for advocates (think ethos, pathos and logos), counselors (who negotiate and navigate deals through a sea of laws and agencies) and lobbyists who sell their contacts list. Everyone else is subjected to strict scrutiny.
Last year was our most profitable ever and mostly it was the result of getting rid of the associates who never would have been hired but for the flush time in which they were hired and of deequitizing those partners who never would have been brought into the partnership but for the endlessly escalating profits per partner that had lifted, for a time, the burden of making tough choices (“We’re sorry Fred, you’re loyal and hardworking but you just don’t have the business to support yourself and contribute to the enterprise as an owner.”)
The unrelenting pressure to be ever more efficient that had transformed our clients is now transforming us. Our clients make more and more widgets with fewer and fewer pipefitters and operators. It’s our turn in barrel.
“It’s our turn in barrel”: will “our” include “my”?
Sorry. Should have been “in the barrel”. That said, I’m happy to report that you won’t be able to mine any schadenfreude from my circumstances. Throughout the good times I paid myself first and cash for everything else. On the other hand, you can take some satisfaction from the plight of a few of my partners. How, for example, does a partner in a law firm wind up getting his lakehouse foreclosed on when he’s made $9 million in the last 10 years? Having a combined $38,000/month in mortgage payments helps. Unbelievable.
Good analysis. The mystery to me is have hourly rates for lawyers fallen much? With such a glut of out-of-work lawyers the rates don’t seem to have fallen much.
Per hour consultation rates of lawyers are atrociously high, the last time I checked.
the hourly for new lawyers has effectively fallen. the rise in price has been outstripped by the rise in tuition cost
Rates at big firms haven’t fallen, no more than the fees at the big Wall Street banks have fallen and for the same reason: you aren’t necessarily paying for the service, you are paying for the internal butt-covering. The exception is big litigation suits, where you have to pay a big firm because only a big firm can service the enormous amounts of paper being shuffled. Which brings us back to the original post and the question whether smaller firms can nudge their way into the defense side of the bar by outsourcing the paper-shuffling.
My rates have gone up yet there’s enormous downward pressure on the rates of the legal widget assembly line workers. As a result our partners are making more from our own production and less from associate leverage.
If predictive coding largely replaces the contract attorney and causes the cost of discovery to go down significantly, then the number of lawsuits will go up. Plaintiffs will balk less at the upfront costs. Attorneys operating on contingency agreements will be able to determine whether it is worth continuing to pursue the case for much less out of pocket costs, which will both allow them to reduce contingency fees and take cases where the amount in controversy is less. Settlements may be less common or will be closer to the plaintiffs original asking price. Ultimately, it will become much cheaper to bring suit, and people will go to the legal mattresses much earlier in disputes.
It is unclear to me whether this would increase or decrease the number of attorneys (though it would change the composition of the bar) or whether or not it would increase or decrease the amount of money paid to attorneys collectively (though I think money paid as damages would go up significantly).
Here’s how you analyze this. The demand for legal services stays constant. The supply curve for legal services shifts down (computers can now do work more cheaply than people could). The total market for legal services (equilibrium price) shrinks. As a result, lawyer salaries go down. There is no way this help lawyers.
If that is too abstract, think of an equivalent machine beats man analogy. If we invent a robot that can build houses, does that benefit construction workers because more people will be able to afford homes (lawsuits?)?
Nevermind; retract all this. Thanks.
Nice ad hominem. Actually, on many issues I lean progressive, ktv girls are way better than bar girls, and most Asian women tend to look 5-10 years younger than their actual age which should obviate the need for underage girls except for pedophiles.
Above comment intended for law schools lie.
Comments on this entry are closed.