Walter has generally been a critic of the erosion of traditional restrictions on maintenance and champerty, dating back to his wonderful discussion of contingency fees in The Litigation Explosion (the chapter is available here). Like Walter, Lester Brickman, and others, I'm generally troubled by the perverse effects contigency fees can generate (for more, see my discussion with Alex Tabarrok here), though I wouldn't go so far as to eliminate such arrangements entirely, even assuming such were politically feasible.
In his post, Larry notes that we should distinguish among types of litigation and that outside funding helps to "eliminate the potential conflict of interest between a corporate client with diversified investors and a risk-averse lawyer who may have an incentive to settle cases that could be productively litigated," excellent points that shouldn't be ignored. I also think Larry is largely correct in stating that legal-system problems would best be remedied "directly by rules constraining improper litigation practices [than] indirectly by constraining firms' ability to pursue the litigation."
The question remains, however, to what degree "outside litigation financing might increase the amount of socially inefficient litigation." As Walter perhaps would, I would tend to believe that the answer could be, "significantly."
In her December paper on loser pays, my colleague Marie Gryphon describes how our current system of litigation encourages what she calls "abusive litigation." She defines "abusive lawsuits" as those that have "have little legal merit, regardless of the magnitude of the recovery sought." Abusive suits in turn break down into "lottery suits" -- those that combine "low legal merit and very high stakes" -- and "nuisance suits" -- those that combine "modest stakes and little legal merit."
Lottery suits include class action and birth-defect litigation: the stakes are so high that such suits have a substantial settlement value, even if the probability of ultimate success is very low. By reducing barriers to entry, outside litigation financing would probably encourage more such suits.
What I'd worry about even more, though, are nuisance suits, which consist primarily of small-value "shakedown suits" and mass-tort "manufactured suits." As a theoretical matter, the existence of such suits at all is a bit of a puzzle, as Marie explains in her paper. The mass-tort context is perhaps easiest to see. If a lawyer has a portfolio of some cases -- say, asbestos claims -- that are valid, but others that are bunk, he can collect on the bunk cases precisely because it is too expensive for a corporate defendant to litigate each case through to final judgment, since his costs are never reimbursed under the American Rule. Outside financing improves the lot of legitimate plaintiffs in mass-tort situations precisely because it would get rid of risk aversion that leads plaintiffs' lawyers to settle such claims on the cheap; but for the same reason it improves the credible threat presented by manufactured claims and thus increases their settlement value. So the total cost of such litigation rises. How one might view this problem depends on how one views such litigation, but the evidence from asbestos, silicosis, and Fen-Phen suggests that the ratio of bad to good cases is actually quite high.
So where are we left? If we limited large-award contingency fees and adopted loser-pays principles, there's much to be said for outside litigation financing. Indeed, Marie's proposal itself calls for eliminating maintenance and champerty barriers to outside insurance, an important access need for a loser-pays reform. (In separate conversations, Tony Sebok has expressed to me a tentative embrace of loser pays combined with outside financing.) Absent such reforms, however, I share Walter's worry about the real-world consequences that outside litigation financing brings.