Today’s case is interesting in no small part because it shows just how far class-action arguments have come in the last 18 months. In the latter half of 2010, most defendants faced with a class action would look primarily at adequacy (are the named plaintiffs good representatives?), typicality (do the named plaintiffs have the same injury and proof as the class?), and predominance (are there more individual issues than common issues?). Now, however, given recent trends in case law–particularly the increased focus on a "rigorous analysis"–more defendants are looking at numerosity  (can the plaintiffs show that there are enough class members?) and commonality (will the class action yield common answers?) as well.
Take, for example, the recent case of Khalif L. v. City of Union City, 2012 U.S. Dist. LEXIS 64567 (N.D. Cal. 2012). In this case, the named plaintiff sued the Union City Police Department, alleging that it had unlawfully discriminated against its African-American population by denying young blacks who had been victimized by Latino gangs any police assistance.
Under the best of circumstances, this would be a difficult case to prove on a classwide basis. (How, for example, does one really prove non-responsiveness on a common basis? Unless the police systematically never responded to calls from African-Americans, there would always be individualized issues.) When the plaintiffs moved for class certification, the defendants opposed them on both numerosity and commonality grounds, and did so successfully.
For numerosity, the plaintiffs had alleged that, based on 2000 census data, coupled with several affidavits alleging non-responsiveness by the police department, the court could conclude that there were enough African-American high school students in the area to constitute a class.
Defendant, however, asserts that plaintiffs’ reliance on the 2000 census data is unduly speculative, and furthermore, that plaintiffs’ supporting declarations are insufficient to satisfy numerosity requirements, since the testimony therein exposes a paucity of contact between certain putative class members and the UCPD (and thus, lack of membership in the class).
Defendant’s objections are well taken. Plaintiffs have failed to connect the census information they primarily rely upon — and that establishes the presence of 301 Black or African-American students enrolled in grades 9 through 12 in Union City in 2000 — to their own class definition or claims.
(Emphasis added.) Similarly, to establish commonality, the plaintiffs asserted that the common issue was whether the Union City Police Department had violated the civil rights of the class members. The defendants responded with an argument that is becoming more common after Dukes: anyone can pose a "common" question, but will it yield common answers?
What plaintiffs here must posit are not questions that within them presuppose a finite legal conclusion that will apply to all class members (e.g., whether the UCPD acts to violate plaintiffs equal protection rights), but instead questions of law or fact that will generate a common answer among members of the class that will aid in determining questions of liability (e.g, allegations that all plaintiffs were subjected to the same particularized UCPD response, or conduct by the same UCPD officer, as a result of similar underlying incidents/reports). This plaintiffs fail to do.
The takeaway from this case is pretty simple. In the wake of cases like Dukes, courts are beginning to recognize that certifying a class means certifying it for a classwide trial. Under those circumstances, even issues like numerosity and commonality are not foregone conclusions for the plaintiff. This is the new normal in class action defense.