Here's what we've got:
Does that distinction make sense? Do we decide whether the statutory policy is right based on what the framers put in the constitution, or should we be concerned with more contemporary issues, like workload on the court? Does congress have to have the same motivation for the statute as the framers had when they set down the constitution?
Art. III § 2 is a pool of potential power that can be given to federal courts-- congress doesn't have to allocate all of it. This goes back to Sheldon v. Sill. On the other hand, this rule extends a power to some defendants and not others, so there's an equality concern here.
Should congress, when it shapes jurisdiction, be trying to effect the intent of the framers? We don't hold the commerce power up to that standard.
A lot of what is in this chapter is pretty much reflections of scholars' bitterness about the current jurisdictional statutes. All that hand-wringing about why the framers put diversity jurisdiction in federal court is really a sign that some people think congress is doing it wrong, and the law should be changed. They're trying to work up to an argument that it's unconstitutional because it violates the framers' intent, but that argument is kind of a stretch.
Or maybe it's law-based, and not party based: maybe congress is using diversity to get the federal courts to apply state law (i.e., they want state law as interpreted by a federal court, and with federal procedure). Would it be problematic if federal lawmakers were skeptical of state judges? Not at the constitutional level, at least: they could just as easily federalize the substantive law. Which is the greater offence to federalism?
Of course, you've got this problem with choice-of-law issues as well: one state might be applying another state's law. So it doesn't make sense to get too upset that state courts don't have universal control over their state's laws-- it'll never get 100% sorted out, because of issues with personal jurisdiction and supplemental jurisdiction.
And in this sense, the fact that diversity jurisdiction gives rise to non-authoritative interpretations isn't wholly bad: there's some value to getting outside articulations of what the law is, because it contributes to the discussion. If we think of jurisdiction as "the law-saying process," we like the idea of getting multiple inputs, in a sense.
Does it make sense to use jurisdiction to address these problems, rather than some other congressional act?
Does this tip up Osborne? Well, we still have to start with § 1331: there has to be some original federal question jurisdiction. Osborne still comes up, in the sense that maybe some other statutory grants of jurisdiction qualify. But the bottom line is that there has to be some statutory grant of jurisdiction: lower federal courts have no jurisdiction out of Art. III alone.
So we don't have to worry about § 1367 until we at least have something that can get into federal court. Once we've got that, we can maybe get some other claims and parties in, subject to the joinder rules.
Now if we're joining things in relying on supplemental jurisdiction, the other parties and claims have to have some relation to the federal claim. Part of the same case. So the question becomes: what is a case?
We can easily see how the joining would be efficient and good, but we need to be sure that all the claims are closely enough related (they're all one case or controversy). Because anything beyond that would be in excess of what the constitution grants, and that's not allowed. The constitution allows ajudication of certain cases, so we need to make sure that whatever we join on, it remains part of the kind of case that federal courts are allowed to hear.
footnote: is there a difference between a "case arising under the laws" and a "controversy between citizens?" Not going to worry about that now.
Now a case is not the same as a claim.
So Gibbs has a federal claim (under § 1331), and a tort claim, in state law. And we get the rule that if the plaintiff's claims are such that he would ordinarily be expected to try them all in the same case, then we can come into federal court.
Π --(§ 1331) → Δ
\_(state tort)_/
§ 1367 is written against the backdrop of the caselaw. It's taking into account things like Finley. That's what the "such supplemental jurisdiction..." language is about.
Pendant party: plaintiff has a federal claim against defendant 1, and a state claim against a non-diverse party. Without that second question in § 1367(a), there would be no pendant party jurisdiction (i.e., as Scalia formulates the situation in Finley). The problem for the plaintiff in Finley is that there's no way the plaintiff can bring all her claims in federal court, and if she bifurcated them, at some point there'd be a judgment in one, and that would preclude the other one as res judicata. And thanks to FTCA, she HAD to bring one of her claims in federal court.
Again, our big policies here are fairness and efficiency (economy and convenience). Bifurcation has an efficiency problem (obviously), but claim preclusion is even worse.
Now § 1367(b). This refers only to actions brought under § 1332 (diversity). If you don't have a diversity case, you can go right to (c), but if you do, you've got to deal with this mess. Basically: you're not going to get supplemental jurisdiction as a means to an end-run around the complete diversity or amount-in-controversy requirements (and that's on a per-defendant basis, by the way).
Impleader: if A sues B, and C will owe B money if A prevails, B can "implead" C into the case. Kroger (a pre-§1367) case, a third-party defendant brought in this way, and the plaintiff asserted a third-party claim against that new party. This was called "ancillary jurisdiction," incidentally. And this new party was not diverse. So § 1367(b) says you A can't assert a claim against C in this situation, because of the diversity problems when Rule 14 is applied here.
Allapattah: two plaintiffs have claims against a single defendant, but one of them doesn't meet the amount-in-controversy requirement. Exploit! the language of § 1367 allows this: it only prohibits defendants joined by Rule 20, not plaintiffs. The court rules against this for reasons too complex to bother with. Anyway, no need to worry about diversity unless you absolutely have to.
Phew- § 1367(c). A discretionary factor: the court may decline to exercise its jurisdiction over specific claims in several instances. This is pretty much an expanded codification of Gibbs. When it's predominately a state law case, when the state law issue is novel or complex, when the federal claim evaporates, or in "exceptional circumstances."
Consider Rule 18, by way of example (parties can pile on as many claims against each other as they want and join them to the existing case). And Rule 20: we can join in multiple plaintiffs and defendants (who can then make use of Rule 18). And 13(g): there can be cross-claims as well (i.e., a plaintiff can sue a co-plaintiff; same with defendants). And then those cross-claimes against can resort to Rule 18 and Rule 20. In short, cases can quickly sprawl to enormous complexity.
Thankfully (20(b)), the court can order separate trials.
Anyway, say that the only federal claim in the case is the first thing that went in, and then we have dozens of other claims?
So is this a good idea, or even constitutional? Probably no to both. At the very least, there are enormous practical problems. But there's also the fact that these supplemental claims won't necessarily share any common nucleus of operative fact, so they're not part of the same "case," and so jurisdiction isn't supported by the constitution.
Still, hard-core Osborne-ists will try to claim there's a federal ingredient in this monster-child of the FRCP. Even if you think that, though, there's not a good reason to squander federal resources this way. On the other hand, if we're really concerned about people with complex cases who are worried about preclusion being shut out of federal court, you might think this is acceptable. And that gets us to...
The statute is a reaction to this.
Walking through the claims, looking for a common nucleus of operative fact. Jurisdiction over the tortious interference claim will depend on whether the malice was the result of his religious or romantic feelings.