A Right to Appeal?

by Michael Dorf

In my latest Verdict column, I discuss the Supreme Court's recent cert grant in Davila v. Davis. The case presents the admittedly technical question of whether a criminal defendant's state habeas lawyer's ineffectiveness counts as "cause" sufficient to excuse the default of a claim that his state direct appeal lawyer was also ineffective, such that a federal habeas court can hear his petition for relief. Got that? No? Okay, go read the column, where I explain the details as well as I can for laypeople. Also, if you're still a law student, take Federal Courts and/or an advanced class in federal habeas corpus. The law still won't make sense, but at least you'll know exactly how it doesn't make sense.

In the Davila case the issue is how narrowly or broadly to read two prior precedents (Martinez v. Ryan and Trevino v. Thaler), which held that ineffectiveness of state habeas counsel is "cause" that can excuse defaulting a claim of ineffectiveness of state trial counsel. In a sense, Davila poses the question of whether a claim of ineffectiveness of state direct appeal counsel is more like a claim of ineffectiveness of trial counsel--if so, the habeas petitioner wins--or more like other claims, which are subject to the rule of Coleman v. Thompson that ineffectiveness of state habeas counsel is not cause--if so, the state wins.

Martinez and Trevino themselves no doubt will strike many readers as somewhat peculiar. There is no federal constitutional right to state habeas. So how can there be a right to effective assistance of counsel on state habeas--even if only for the limited purpose of excusing a default? The Supreme Court in Martinez  tried to duck that question by claiming that it was only recognizing an equitable principle, not a legal right. If that sounds mysterious to you, you're not alone.

Still, the notion of a conditional right isn't all that hard to understand. We find conditional rights in other areas of the law as well. There is no right of the public to have the government dedicate public property as a municipal park, but if it does, then there are various rights of access guaranteed by the First Amendment under the public forum doctrine. Closer to the context of a case like Davila itself, there is no federal constitutional right to an appeal in a criminal (or civil) case, but if the state provides for an appeal, then there is a right to counsel (including state-appointed counsel for indigents) on appeal, per the rule of Douglas v. California (1963).

Thus, the idea that there is no right to state habeas, much less to free counsel for state habeas, but that equity will excuse a default of an otherwise valid ineffective assistance of trial counsel based on ineffective assistance of state habeas counsel falls within the general category of propositions that deny that the greater power to deny some right or benefit entirely includes the ostensibly lesser power to deny it partially on some problematic basis.

In itself, the idea that the greater power does not always include the lesser power is sensible enough, especially where the exercise of the ostensibly lesser power would result in illicit discrimination. The greater power of the government not to provide health insurance at all does not include the lesser power to provide health insurance only to white people. The greater power not to create a municipal park does not include the lesser power to create a park open only to pro-Democratic rallies and marches. Etc.

Likewise, some of the injustice of allowing a death sentence to stand based on lawyer incompetence surely sounds in principles of equality. Who lives and who dies should not turn on the fortuity of who has the better lawyer.

Yet that principle doesn't seem to carry us far enough in the current context. For one thing, the fact that there is no right to a lawyer at all in state habeas means that even after Martinez some defendants could end up being sentenced to death based on the quality of (non-)lawyering at the state habeas phase--namely, those defendants who lacked any counsel on state habeas. That's why the Court in Martinez suggested--but did not decide--that perhaps there is a right to a lawyer on state habeas where state habeas is the first opportunity to raise a particular claim, such as ineffectiveness of trial counsel. In such cases, state habeas functions as de facto first appeal and thus the rule of Douglas applies. That, at any rate, is the possibility left open by Martinez. I am suggesting that the equal-protection logic of the opinion argues in favor of extending Douglas in this way.

But that brings us back to a problem with Douglas itself. The Court's opinion in that case sounded chiefly in equal protection, concluding that "[t]he indigent, where the record is unclear or the errors are hidden, has only the right to a meaningless ritual, while the rich man has a meaningful appeal." That is indeed problematic, but Douglas ought also to have sounded due process themes. Sure, if the state gives an appeal, it should have to provide a lawyer. But is it really permissible for the state to vitiate that right for everyone by denying any right of appeal? The Douglas Court thought so. It left undisturbed the proposition--articulated in McKane v. Durston (1894) and reaffirmed in Griffin v. Illinois (1956)--that the Constitution does not require any appeal in a criminal case. No subsequent case calls that proposition into question.

To be sure, the issue is hypothetical, because every state and the federal government affords at least one layer of appeal in criminal cases. But what kind of law professor would I be if I permitted the hypothetical nature of the problem to preclude my examining it?

So, suppose some state were to do away with a right of appeal in criminal cases, preserving (let's say) discretionary review by the state supreme court to be followed by the discretionary grant of cert by the U.S. Supreme Court on federal questions that could prove dispositive. Under current doctrine, that would be permissible. Should it be?

I would like to think not. For one thing, the very fact that all of the states in fact permit a right of appeal in serious criminal cases is evidence that such a right has become "fundamental" to what the Court in Duncan v. Louisiana (1968) called the "American scheme of justice." Just as in the Eighth Amendment context the widespread rejection of a practice by the states serves as evidence that the practice is "cruel and unusual," so in this context the widespread--indeed universal except in my one hypothetical state--practice of granting appeals in criminal cases argues for its fundamentality.

There is also the problem of error. Some years ago my then-colleagues Jeff Fagan and Jim Liebman examined capital cases from 1973-1995. They reported that reviewing courts found prejudicial error in 68% of those cases. Many of those errors did not directly bear on guilt or innocence, but some did, and even the ones that didn't involved serious error. Perhaps the error rate is somewhat lower in non-capital cases, where the rules are somewhat less punctilious. At the same time, however, one would think that the high stakes of a capital case would give prosecutors and trial judges extra reason to get it right. In any event, even if the error rate in capital cases is several times higher than in non-capital cases, the Fagan/Liebman findings imply that trial judges err enough to warrant appeals in all cases. A 1993 NY Times story about an especially error-prone judge cited a 20% reversal rate, even as it noted an 8.5% nationwide reversal rate for the most serious felonies. A system in which trial court judges did not have to worry about reversal would likely produce an even higher reversal rate.

To be sure, the fact that an appeals court reversed a trial court is no guarantee that the appeals court was right and the trial court was wrong in some absolute sense. But our one-way ratchet system (in which a defendant can appeal a conviction but the government cannot appeal an acquittal) will catch the errors that most threaten our core principles--that it is better to let the guilty go free than to punish the innocent.

That brings me, finally, to a practical point, a way in which my plea for a right to appeal is relevant in a country in which criminal defendants already have rights to appeal. Both the Fagan/Liebman study and common sense suggest that each additional layer of review adds to the reliability of the criminal justice system. Of course, one can reach a point of diminishing returns. But I do not think we had reached that point in the early 1970s when the Burger-then-Rehnquist-then-Roberts Court and then Congress began to substantially cut back on the scope of federal habeas corpus as a collateral remedy for state prisoners.

The Davila case arises in the aftermath of that cutback. Under the Warren Court precedent of Fay v. Noia, state court procedural defaults were generally excused for federal habeas purposes so long as the defendant did not "deliberately bypass" the state courts. Counsel's incompetent decision not to present an issue to a state habeas court would not count as deliberate bypass under Fay. The baseline rule of Coleman and other post-Warren Court cases--requiring the habeas petitioner to show cause--is what creates the need for Davila to argue that his case is more like Martinez than like other defaults through the ineffectiveness of state habeas counsel. And those post-Warren Court cases, in my view, under-value federal habeas because they under-value the right of appeal. Indeed, as Douglas shows, even the Warren Court, which decided Douglas, under-valued the right of appeal.