Timothy P. O'Neill

Chicago Daily Law Bulletin
Volume: 154 Issue: 037

February 22, 2008

Back in the 1960s, the old Baltimore Colts had a running back named Alex Hawkins. Hawkins was more memorable for his activities off the field than on.

The story goes that he was caught in a raid on a poker game in Baltimore during the wee hours. After the Colts bailed him out of jail, he was told to report to Coach Don Shula's office.

''Hawkins,'' Shula said. 'What's wrong with you? Did you know that one of the guys you were with has been arrested 33 times?''

''Coach,'' said Hawkins, ''it's hard enough to even find a poker game at 5 a.m. You expect me to check their rap sheets, too?''

Shula laughed so hard he didn't even notice that Hawkins had left the office. Tom Callahan, ''Johnny U.: The Life and Times of John Unitas'' (2007).

Guys like Hawkins don't need a lawyer; they can go pro se. But for those who do need counsel, several questions arise. What makes a criminal-defense lawyer ''effective''? And even if the lawyer is ''effective,'' are there meaningful differences between ''effective'' criminal-defense lawyers?

First, look at the language of the Sixth Amendment. The criminal defendant is merely promised the ''assistance'' of counsel. There is no adjective. The first time the U.S. Supreme Court inserted an adjective was in the notorious ''Scottsboro Boys'' case in 1932. The court held that to pass constitutional muster the counsel's assistance must actually be ''effective.'' Powell v. Alabama, 287 U.S. 45 (1932). Yet for years the court did not elaborate on what exactly this entailed.

The Burger Court issued two important holdings. First, in 1980 the court made it clear that the standard of ''effectiveness'' was the same for both retained and appointed counsel. Cuyler v. Sullivan, 446 U.S. 335 (1980). Second, four years later the court tried to articulate a substantive definition of what ''effective assistance'' of counsel entailed. Strickland v. Washington, 466 U.S. 668 (1984).

Prior to Strickland, states and circuits were hopelessly split on how to evaluate effectiveness of counsel. Some opted for a ''checklist'' approach that set out the duties of a defense attorney. (See the discussion of this in U.S. v. DeCoster, 199 U.S.App.D.C. 359 (D.C. Cir. 1976).) At the other extreme, some jurisdictions such as Illinois refused to find ineffectiveness unless the performance reduced the trial to a ''farce or sham.'' People v. Torres , 54 Ill.2d 384 (1973).

Strickland took a middle ground. It found that the issue of ''effectiveness'' did not exist in a vacuum. Rather, it saw ''effectiveness'' in an instrumentalist context, i.e., as a means to an end. And that end is ''the ability of the adversarial system to produce just results.'' The key question is always whether counsel's performance was so deficient that the trial ''lost its character as a confrontation between adversaries'' thus resulting in an ''actual breakdown of the adversary process.''

Strickland created a two-part test for a defendant to prove that his counsel had been unconstitutionally ineffective. First, defendant must establish that his counsel's performance was ''deficient'' in that counsel made errors so serious that counsel was not functioning properly. Second, defendant must establish that counsel's errors were so serious ''as to deprive the defendant of a fair trial,'' defined as ''a trial whose result is reliable.'' The test for this is whether there is a ''reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.''

Yet Strickland made two additional points. First, it warned appellate courts against regularly second-guessing the performance of trial counsel. It specifically held that ''scrutiny of counsel's performance must be highly deferential.'' Thus, the thumb is already on the judicial scales in favor of the status quo.

But the second point has had an even more deleterious effect on the development of a body of law defining what is ''effective assistance.'' Strickland explicitly told reviewing courts that, when faced with a claim of ineffective assistance of counsel, they need not invariably decide whether counsel in fact was ineffective. For if a court concludes that any possible ineffectiveness could not have created a reasonable probability that the result could have been different, then they could simply skip the first step and affirm the conviction without deciding whether counsel's performance was deficient.

In a narrow legal sense this may be proper. After all, a finding of ineffectiveness that doesn't result in prejudice may only be dictum. And, as Strickland says, ''The object of an ineffectiveness claim is not to grade counsel's performance.''

But on a broader level, courts that consistently find ''no prejudice'' while refusing to evaluate effectiveness are missing an opportunity to provide guidance to defense attorneys in this crucial area of the law.

For this reason, it was good to see the Illinois Supreme Court at least intimate that that a defense attorney in a recent case could have done a better job. People v. Manning, No. 104300 (decided Feb. 7). Manning, a person with a history of mental health and addiction issues, was charged with residential burglary. After consulting with his defense attorney, he rejected the state's recommended sentence (24 years) and, hoping to get a lower sentence, entered an open plea of guilty. The judge sentenced him to 22 years. He appealed, alleging that his attorney was ineffective for not explaining the option of ''guilty but mentally ill'' (GBMI).

The Supreme Court found no prejudice because ''there is virtually no difference in the treatment that defendant will receive under GBMI as opposed to a guilty plea.'' (Whether GBMI is the reincarnation of ''farce and sham'' will be left to another column.) What is important is that the Supreme Court at least critiqued counsel's performance and, referring to the ABA Standards for Criminal Justice, concluded that it would have been ''better practice'' for defense counsel to have fully discussed all plea options with Manning. The Supreme Court should be commended for offering some useful guidance to defense lawyers.

But constitutional ineffectiveness aside, are there appreciable differences in attorney ability within, say, a single public defender office? A recent empirical study answers ''Yes.'' David S. Abrams and Albert H. Yoon, The Luck of the Draw: Using Random Case Assignment to Investigate Attorney Ability, 74 Univ. of Chicago Law Review 1145 (2007).

Abrams and Yoon focused on the Clark County Public Defender's Office in Nevada because it possesses two uncommon institutional features. First, the office makes a random assignment of all non-capital felony cases to its 76 public defenders. The office allows absolutely no ''cherry picking.'' Whether a non-capital felony goes to the most or least experienced attorney is based on pure ''luck of the draw.'' Second, the office engages in vertical representation, meaning that an attorney assigned at the start will be the attorney of record all the way through the final resolution.

Abrams and Yoon then arranged all 76 lawyers on an ''ability'' scale that took into account only two variables: the incarceration rate and the sentence length rate of the attorney's non-capital clients.

They discovered several statistically significant results. For example, they found that a defendant had a 1.7 percent lower probability of incarceration for every extra 10 years of his attorney's experience. The overall mean incarceration rate is 39 percent; thus, having a 25-year veteran public defender could reduce the likelihood of incarceration by more than 10 percent, relative to a new public defender.

How about length of sentence? The average sentence length in the study was 7.2 months. The study found that an attorney with 11 years of experience will, on average, obtain sentences that are 1.2 months shorter than an attorney with only one year of experience. This is a reduction of 17 percent.

Another significant variable was whether the attorney was Hispanic. In terms of length of sentence, the study found Hispanic attorneys consistently outperforming white, Asian, and African-American attorneys. Defendants represented by Hispanic attorneys were less likely to receive prison sentences, and those that did received sentences about 27 percent less than the average. True, the differential performance between Hispanic and other attorneys is greatest for Hispanic clients (perhaps reflecting language concerns). But the study showed that Hispanic lawyers achieved better outcomes than their peers regardless of their clients' race.

It is difficult to make broad generalizations based on statistics for a single county. But the Abrams and Yoon study is definitely worth a careful look. Both judges and lawyers should welcome empirical studies that aid in improving the delivery of effective criminal defense.