The Best Test of a New Lawyer's Writing
Let's say that you need to hire a new lawyer. No small decision, and you don't want to go wrong, so you take the usual steps: sort applications, review transcripts, read writing samples, interview candidates, check references—and then pick someone from the short list. You might think that you have covered all the bases, but you would be wrong. You haven't done enough to assess the candidates' most important skill—their writing.
No one, I'm sure, will dispute that lawyers speak and write for a living. In a telling study by the American Bar Foundation, about 1,200 practicing lawyers were asked to rate lawyering skills from a list of 17 different skills. At the top of the list, in a class by themselves, were oral and written communication.1 The American Bar Association has said the same thing, and in one report after another has encouraged, urged, pleaded with law schools to improve their legal-writing programs. One report, for instance, says, "Legal writing is at the heart of law practice, so it is especially vital that legal-writing skills be developed and nurtured through carefully supervised instruction."2
To confirm how central writing is, look over the advertisements that appear in legal publications. In the latest issue of the Michigan Lawyers Weekly (at the time I write), there are 26 ads under "Employment Available—Lawyer." As varied as the ads are, with many seeking expertise in a specific practice area, 10 of them include statements like this:
• "Excellent writing skills a must."
• "Excellent writing skills required."
• "Only candidates with demonstrated research and writing abilities need apply."
By comparison, "strong academic credentials" is mentioned just three times. And it's the same pattern—the search for the same defining and distinguishing skill—every week in those ads.
The Best Test:
What could be more obvious? To see what the candidates can do, have them do it. Once you get down to a short list of finalists, have each of them take a performance test—a writing exercise. The time and effort required of you and the candidates is piddling when compared with the investment that a decision to hire will entail. You can put the test together in a matter of hours, then reuse it to your heart's content.
For the candidate, the performance test will take between two and six hours, depending on which test or tests you choose. At most, the candidate will have to spend a day at your office. If anyone seems to feel degraded or put upon by the test, that in itself might reveal something about the disposition of the person you would be working with.
Now, before turning to specifics, I want to explain that I'm talking mainly about testing a lawyer's ability to analyze and apply law in a clear and coherent way—to think straight on paper (or on a computer screen). At the same time, of course, you can assess the work in a general way for style and grammar. Is the writing tight, readable, and mostly error-free? You can decide how to weigh the different qualities and how much to forgive because of the time constraints. At any rate, you can bet on one thing: give a good performance test, and you will not hire a bad writer.
Here are the possibilities.
A closed-world performance test. By "closed world," I mean that the writer does no research; you furnish the legal problem and the legal sources needed to address it.
A good example is the Multistate Performance Test, now administered in about half the states. Candidates get a file—the factual background of a case, including relevant documents—and a library with the authorities they can use to analyze the legal issues. They have 90 minutes to write, for instance, a memorandum, a letter to the client, or a settlement proposal. You can allow more time if you want more polish.
Tests given in years past, along with the "point sheets," or answer guides, can be purchased quite reasonably from the National Conference of Bar Examiners at http://www.ncbex.org/Tests/mpt.htm. Is it within bounds to use these ready-made tests for your own private performance test? Yes, it is, although the organization is obviously not trying to market its product that way.
As an alternative, you can easily create a closed-world test yourself.3 You must have an office file that you can adapt. Put together a packet modeled on the Multistate Performance Test: an outline of facts, including the question you want answered and the instructions for what to write; perhaps a disputed document, a pleading, or excerpts from depositions; the relevant statutes or rules; and not more than three or four cases. You could, as the multistate test sometimes does, include an irrelevant statute or case. Again, you have to assemble a packet just once. And presumably you have already done the analysis yourself, so you have a good idea of how the answer should go.
A research-added performance test. The only difference here is that you would not provide the selected library of legal authorities. You would provide only the file—whatever facts and documents you want the writer to use—and the writer would do the research, either in the office library or at a public law library. I'd keep the research fairly basic. Adapt a file that you would give to a new lawyer—probably a one-issue state-law problem that does not involve more than one or two statutes and a few cases.
The research-added test should take
A test for grammar and style. Conveniently enough, you will find just such a test (complete with answers) in Volume 5 of The Scribes Journal of Legal Writing. The test, called "The Legal-Writing Skills Test," was devised by Bryan Garner, a top expert on legal writing and legal language. The test has two parts: an editing section and an essay section. If you give one of the two performance tests described earlier, then skip the essays. The editing section has 35 items, most of them single sentences, that cover a range of skills: grammar and punctuation; correct usage (the difference between affect and effect, for example); converting the passive voice to the active voice; tightening wordy passages; and eliminating legalese. The editing test would add about 90 minutes.
In the end, you have to decide how important writing is in your practice, how confident you want to be about your decision, and what combination of tests to use.
There are traditional and obvious ways to gauge writing: look at what the candidates have already written or at the grades they earned in their required law-school writing classes. These credentials are worth considering, as long as you understand their not-so-obvious limitations.
A writing sample. A 10-page writing sample will probably involve a more complex analytical exercise than a performance exam does, so you can assess the writer's ability to handle a tougher intellectual challenge. Also, because the writer had the luxury of time, you won't wonder whether you should excuse deficiencies, and to what extent. The sample ought to be polished, and you can feel reasonably confident that it presents the writer at his or her best.
The trouble is, the sample may not be the writer's solitary best; it may be, at least to some extent, a collaborative effort. If it came from a legal-writing class, then it was probably critiqued (students would say "ripped") two or even three times: as a first draft, as a final draft, and possibly as a rewritten final draft. Moreover, it's not unheard of for students to ask a different writing professor to "look over" a writing sample before it departs into the real world. I've looked over my share in 16 years.
There's nothing wrong with any of this. Good writing instruction assumes good feedback, and the final product is still the writer's work, primarily. Just so you know.
A published article. Certainly, an article would be a plus. It tends to show intellectual ability, academic accomplishment, an interest in writing, advanced course work in writing, and the approval of other readers.
Again, though, you can't be sure how much editing the article needed or received from a scholarly-writing professor or from the journal's own editors. A portfolio item is not the same as a live performance.
One other suggestion: if the article inclines toward the plodding and overwrought style of most law journals, ask whether the writer can convert to the plain language that most readers prefer in practice documents.4 You might even pull a few pages from someone else's article and ask for a rewrite.
A grade in a law-school writing class. An A is good news and a C is bad news, but grades in between are harder to weigh unless you happen to know the program or the professor. Although a B+ looks good, maybe the professor gave no grade, or just a couple of grades, below a B. A C+ looks pretty bland, but maybe the candidate earned a better grade in a second required writing class. Then again, maybe the first professor was an experienced teacher and the second was not. There are many other variables. I would treat writing grades as one more indicator.
Speaking of indicators, let me ask a few questions that I hope will not give you pause. But talented new lawyers do tell discouraging stories about the attitudes and practices of some supervisors.5
Do you resist, and maybe resent, the idea that lawyers ought to write in plain language? Do you regularly strain against the page limits that courts impose? Do you try to raise every issue imaginable, rather than settling for just your best ones? Do you wait a few pages before stating the issues and then state them superficially, rather than putting the deep issues up front?6 Do you give a lengthy analysis of most cases, use lots of block quotations, and take few pains to make clear how each new case connects and moves the analysis forward?
Do your sentences average more than 25 words? Do you favor the passive voice and commonly turn verbs (like consider) into abstract nouns (give consideration to)? Do you end affidavits with "Further affiant sayeth naught"? Do you end contracts with "In witness whereof the parties hereto have affixed their signatures"? Are you fond of prior to and in the event that and hereinafter?
If you answered yes to any of these questions, you might look into a good book or seminar on legal writing—to help you judge writing smartly and mentor well.
This article is reprinted from the June 2000 issue of Trial. ©
1. Bryant C. Garth & Joanne Martin, Law Schools and the Construction of Competence, 43 J. Legal Educ. 469, 473, 477 (1993).
2. Council of the Section of Legal Education and Admissions to the Bar, Long-Range Planning for Legal Education in the United States 29 (1987).
3. See Ritchenya Shepard, Firm Exam Tests Writing Skills, Nat'l L.J., Feb. 15,1999, at A16 (noting that the Chicago firm of Connelly Sheehan Moran regards its test as the most important element of its hiring process).
4. See Joseph Kimble, Writing for Dollars, Writing to Please, 6 Scribes J. Legal Writing 1, 19–31 (1996– 1997) (summarizing a number of studies showing that plain language is clearer and more effective than legalistic style).
5. See Symposium, The Politics of Legal Writing, 7 Scribes J. Legal Writing 29 (1998–2000).
6. See Bryan A. Garner, Issue-Framing: The Upshot of It All, Trial, Apr. 1997, at 74.
"Plain Language" is a regular feature of the Michigan Bar Journal, edited by Joseph Kimble for the State Bar's Plain English Committee. The assistant editor is George Hathaway, chair of the committee. The committee seeks to improve the clarity of legal writing and the public opinion of lawyers by eliminating legalese. Want to contribute a plain English article? Contact Prof. Kimble at Thomas Cooley Law School, P.O. Box 13038, Lansing, MI 48901. For information about the Plain English Committee, see our website—http://www.michbar.org/generalinfo/plainenglish/home.cfm.