The Good Faith Exception and Changing Law: The Costs of the Exclusionary Rule, and Balancing Costs and Benefits

This is my third post in a planned series on why the good faith exception to the exclusionary rule does not apply when a police officer conducts a search that is lawful when it occurs that is later ruled unlawful before the conviction is final. This issue is being litigated all over the country right now thanks to a clash of two recent Fourth Amendment cases, Herring v. United States, — U.S. — (2009) and Arizona v. Gant, 556 U.S. ___ (2009). Herring has language suggesting a broad approach to the good faith exception, while Gant ruled a common and widely-accepted law enforcement practice unconstitutional. The combination of the two cases raises a question: Should the exclusionary rule apply for the many violations that occurred before Gant was handed down, or does the good faith exception apply so the evidence is admitted? Here is post #3: The Costs of the Exclusionary Rule, and Balancing Costs and Benefits. Post #1 is available here; post #2 is available here.

* * * *

Now we turn to the costs of the exclusionary rule.   The key to determining the costs of the exclusionary rule for any good-faith case is understanding the subset of cases in play.  The question is, is the cost of suppression for that class of cases worth the benefits explained in Post #2?  Here I think the answer is clearly “yes.”  To see why, we need to start by identifying the precise class of cases in which suppression would be imposed.

1.  The Class of Cases. The trick to understanding the scope of suppression is to recall the different limitations on the exclusionary rule discussed in my first post.  As I explained there, whether the exclusionary remedy applies depends on three basic questions:

1. Is there sufficient connection between the constitutional violation and the discovery of the evidence to warrant suppression?
2. Is the defendant one of the individuals who is permitted to assert the challenge?
3. Is this the kind of constitutional violation that supports suppression?

The exclusionary rule applies only if the answer to all three questions is “yes.”  Further, if the evidence was wrongly admitted, a subsequent conviction will be overturned only if the error was not “harmless.”    These limitations are critical to understanding the costs of the exclusionary rule, as  it means that the exclusionary rule will only have a cost when no other exception to the exclusionary rule would apply.

In the case of the proposed good faith exception for changing law, costs would only occur in those cases in which (a) the illegal search directly caused the discovery of the evidence (fruit-0f-the-poisonous tree), (b) the government would not have discovered the evidence had the unconstitutional search never occurred (inevitable discovery), (c) the government did not know the evidence some other way (independent source), (d) the defendant’s own rights were violated (standing), (e) the conviction is not yet final (retroactivity), (f) the kind of constitutional violation is the kind of violation for which the Supreme Court has concluded that suppression is necessary (good faith), and (g) the error is not harmless.

Think about that for a second.   The only cases in which the exclusionary rule would apply are those cases in the direct review pipeline that would have never been brought in the first place had the Constitution been followed, and for which the violation is so serious that the Court believes the evidence must be suppressed.  That’s a limited group, consisting of the most core constitutional violations out there.  These are the cases involving the serious constitutional violations in which suppression has been deemed a necessary cost.  And it’s those cases, and only those cases, that would lead to suppression.

Powell v. Nevada, 511 U.S. 79 (1994), provides a helpful example of how narrow this class of cases will be.   In 1989, Powell was arrested for murder and held by the police without an initial hearing for 10 days.  On the 4th day of his detention, he made incriminating statements that were later used against him at trial.  While Powell’s case was on direct review,  the Supreme Court decided County of Riverside v. McLaughlin500 U.S. 44 (1991), which held that the initial hearing was required within 48 hours of the initial arrest. The Nevada Supreme Court initially held that McLaughlin was not retroactive, so the McLaughlin case did not apply.  The Supreme Court unanimously reversed,  holding that McLaughlin was retroactive to Powell’s case because Powell’s case was on direct review.  But the Court then stressed that while McLaughlin applied, the exclusionary rule may not:

It does not necessarily follow, however, that Powell must be set free, or gain other relief, for several questions remain open for decision on remand. In particular, the Nevada Supreme Court has not yet closely considered the appropriate remedy for a delay in determining probable cause (an issue not resolved by McLaughlin), or the consequences of Powell’s failure to raise the federal question, or the District Attorney’s argument that introduction at trial of what Powell said on November 7, 1989 was “harmless” in view of a similar, albeit shorter, statement Powell made on November 3, prior to his arrest. Expressing no opinion on these issues, we hold only that the Nevada Supreme Court erred in failing to recognize that Griffith v. Kentucky calls for retroactive application of McLaughlin’s 48 hour rule.

Indeed, the Nevada Supreme Court upheld Powell’s conviction despite the Constitutional violation.  Powell v. State, 113 Nev. 41 (1997).

2.  The Cost of the Exclusionary Rule And the Merits of Gant. At this point you may be thinking, well, okay, so maybe not all of the bad guys will go free.  But surely some of them will.  And that’s a cost, right?  People who violated the law will get out of jail.

I agree it’s a cost.  But this cost is the cost that the Supreme Court will have just announced is worth incurring by establishing the new rule that is subject to a possible good faith exception.  Consider Arizona v. Gant, which dramatically changed the law of searches incident to arrest for automobiles and led to the flood of cases on whether the good faith exception applies to changing law.  The majority decision in Gant concluded that it was not constitutionally reasonable based on a balance of legitimate interests for the police to have searched Gant’s car while Gant was arrested and in the back of the squad car.  The search went too far, and should not have been allowed.  That is, a weighing of the interests produced the result that the government should not have known about the drugs in Gant’s car in the first place.

In this setting, applying a “good faith” exception to Gant and other cases with identical searches before Gant is akin to revisiting Gant and concluding that the Gant court got it wrong. Applying the exclusionary rule for Gant violations puts everybody back where they would have been had the police done what was permitted in Gant. The question must be, did the Gant court accurately gauge the competing costs and benefits of allowing the search in that case? Whatever the answer is, presumably that should be settled by the Supreme Court in announcing the new rule, not by a later court under the guise of the good faith exception.

3. Good Faith for Changing Law and the Problem of Technical Violations.  There’s one more objection worth exploring.  What if the new rule announced by the Supreme Court is a somewhat technical Fourth Amendment rule, rather than a rule announced after weighing costs and benefits?   Specifically, what if the Court’s new rule is a technicality, imposed for reasons of history or something like that, and it isn’t worthwhile from a cost-benefit framework to let people out of jail for violating it?  Wouldn’t  imposing the exclusionary rule for such a technical violation be excessive?

The answer is “yes,” but that this problem is already dealt with by the preexisting scope of the good faith exception.   There is no need for a good faith exception for changing law to cover this, as it is already covered by the traditional function of the good faith exception focused on the nature of the violation.   In such a case, the proper course is for courts to rule that the exclusionary rule doesn’t apply because it is not worth applying the exclusionary rule for that kind of violation, not because the rule wasn’t announced before the search occurred.

We can see this play out with the Fourth Amendment litigation on knock-and-announce violations — cases in which the Fourth Amendment required the police to knock on the door and announce their presence as police officers before executing search warrants. Before Richards v. Wisconsin, 520 U.S. 385 (1997), a lot of state precedents had allowed the government to dispense with this requirement in all felony drug cases. In Richards, the Supreme Court rejected a per se exception to the knock-and-announce rule and instead required a case-by-case determination of need for a no-knock warrant. Defendants with cases already in the pipeline tried to invoke the exclusionary rule in cases that could not satisfy the case-by-case standard. Two state Supreme Courts — Wisconsin and North Dakota — ruled that the good faith exception applied because the law had changed. See State v. Ward, 604 N.W.2d 517 (Wis. 2000); State v. Herrick, 588 N.W.2d 847 (N.D. 1999). The officers could not predict that the Supreme Court would overturn state court precedents in Richards, these courts ruled, and so the cost/benefit analysis favored letting in the evidence given the changing law.

But then a few years later, in Hudson v. Michigan, 547 U. S. 586 (2006), the Supreme Court examined how the good-faith cost/benefit analysis applied for knock-and-announce violations generally and concluded that suppression is not an available remedy for knock-and-announce violations. The Court’s reasoning is a bit odd, but the basic conclusion seems quite right to me: The knock-and-announce rule is too technical in nature for violations of it to warrant suppression of all the evidence found when the warrant is executed. The remedy would be extravagant for the nature of the wrong. With the holding of Hudson in mind, now go back to state Supreme Court decisions in Ward and Herrick. Now the cost-benefit analysis of the good-faith exception for changing law in those cases obviously looks different: The costs of suppression for changing law are exactly zero in all cases in which the good faith exception applies based on the technical nature of the violation. No one will go free based on the retroactive application of Richards v. Wisconsin because of the technical nature of the violation as recognized in Hudson v. Michigan.

4. Weighing Costs and Benefits. Now it’s time to weigh costs and benefits of applying the exclusionary rule for new law on direct review. We need to compare the harm caused to the functioning and integrity of the development of criminal procedure law if we apply the good faith exception for changing law to the harm caused by not punishing some people who committed crimes if we don’t apply the exception. This sort of weighing is normally hard to do, as it forces us to compare constitutional apples and constitutional oranges: It’s not like there’s a handy conversion chart for converting lost chances at punishment for unknown crimes with damage to the system of criminal procedure law.

Still, in this case I don’t think it’s close call. The benefits of the exclusionary rule here are simply enormous: As explained in Post #2, the exclusionary rule is the linchpin for our entire system of Fourth Amendment law. On the other hand, the costs of the exclusionary rule here are quite modest: freeing people from jail who shouldn’t have been prosecuted in the first place. It seems to me that this is just about the clearest case for the role of the exclusionary rule that you can find.

So that’s the cost/benefit analysis. In my next post, I want to turn to the toughest case against my position and explain why it is distinguishable. That case is Illinois v. Krull, 480 U. S. 340 (1987), and I’ll address it in my next post.