A second shot at a plea bargain
Your client's sitting in his yard, minding his own business, when the next-door neighbor, with whom he's been feuding for years, comes over and starts arguing with him. The argument escalates, the other guy has some pruning shears which he starts waving around, but your client has a gun, and in any papers-rock-scissors contest, a gun trumps everything, with the result that your client now has a dead neighbor, and is facing a murder charge.
It's an iffy case; after all, the other guy did have the shears. Still, he was standing ten feet away when your client shot him, and you're going to have a tough time proving self-defense by a preponderance of the evidence in those circumstances. The State offers to let your guy plead to one count of voluntary manslaughter with the agreed minimum sentence of three years.
Your client's willing to take the deal, but you tell him not to. You've read something on some blog by a Cleveland attorney about the Castle Doctrine, and that it shifts the burden of proof on self-defense to the prosecution, so you convince your client that he has a good chance to get acquitted, and off you go to trial. Where you find out that in addition to reading that blog, you should have checked out the law, because if you had, you would've learned that the Castle Doctrine requires you to be in your house, or at worst an attached porch; it doesn't apply if you're just in your yard. Your client is convicted of murder, and probably will wind up spending about seven or eight times as long in prison as he would've if he'd accepted the deal.
Can he use the fact that you screwed up and gave him bad advice to get out from under that?
That's the issue in a couple of cases before the Supreme Court next term. In Cooper v. Lafler, the 6th Circuit affirmed habeas relief where the lawyer rejected a plea deal of 51 to 85 months on an attempted murder charge because he wrongly believed that, because all of the shots hit the victim below the waist, the jury couldn't find an intent to kill; the defendant was convicted and sentenced to 185 to 360 months. The situation in Missouri v. Frye was even worse: there, the State offered a plea of ten days in jail on a drunk driving charge, the offer expiring at the time of the preliminary hearing. The lawyer never told his client about the offer, and Frye eventually pled guilty and wound up getting three years imprisonment.
Both courts held that the lawyers had rendered ineffective assistance, with the 6th Circuit requiring the state to sentence Lafler to the 51 to 85 months it had offered, or cut him loose. The situation in Frye's case was more complicated: shortly after the plea deal expired, he'd picked up another drunk driving charge, which is how he got three years in prison. Had he pleaded to the first when he had the chance, he could've only gotten one year on the second. The Missouri court decided it couldn't force the state to make the same offer, so it simply remanded the case to allow Frye to plead guilty again or go to trial.
While there are a number of cases, like this one from the 6th Circuit, which have held that the failure of an attorney to advise his client of a favorable plea deal constitutes ineffective assistance, these two cases, expecially Lafler, tread on more unfamiliar territory: ineffective advice during the plea bargaining stage. The government's argument in both is that such errors are irrelevant. Under their view, the right to effective assistance of counsel is intended to ensure that the defendant receives a fair trial, and as long as he does, how his attorney performed during the plea bargaining process is irrelevant. Simply put, one does not have a constitutional right to a favorable plea bargain, and if your attorney screwed up in not getting you one, that's too bad.
Whether this analysis survives Padilla v. Kentucky is another matter. In Padilla, discussed here, the Court held that counsel's affirmative misadvice regarding the immigration consequences of a plea rendered the plea invalid. To be sure, that case involved only a plea, rather than the plea vis-a-vis trial questions presented by Lafler and Frye, but it still recognized that counsel's role extended to matters which had previously regarded as merely collateral. In fact, that opinion rejected the previous view, asserted in Scalia's dissent, that "we have held that 'defence' means defense at trial, not defense in relation to other objectives that may be important to the accused."
What's more, given that the vast majority of criminal cases end in pleas, it makes little sense to view a defense attorney's role purely in terms of how his performance affects a trial. As I once wrote, a criminal defense attorney's primary role is damage control: we are dealing with clients who not only are usually guilty, but whose guilt can rather easily be established. Our ability to properly manage the plea bargaining process, not only with the prosecutor, but with our own clients, is critical to obtaining the best results. That doesn't mean that a client should have an ineffective assistance claim merely because he didn't get the deal he thought he could, or that a more skilled or experienced lawyer could have gotten him, any more than he has such a claim simply because another lawyer might have won an acquittal at trial. There's still the requirement of deficient performance. But not communicating a plea offer is deficient performance, and so is not knowing the law sufficiently to accurately appraise a plea offer and properly advise a client about it. And when a client suffers an adverse result because of that, it's just as much a failure of effective assistance of counsel as if it had occurred in trial.