I apologize for the lack of Padilla related posts since the New Year. There simply wasn’t much to report, perhaps due to the holidays. This has changed some of late, and courts have been issuing opinions discussing both Padilla retroactivity and Padilla ineffectiveness. As to the former, however, district courts seem to realize that the question of whether Padilla applies retroactively is one that, sooner or later, will be decided by the Supreme Court. And they have acted accordingly, either skirting the issue entirely or ruling on it but with little analysis. As to the latter question of what constitutes ineffective assistance of counsel under Padilla, the fight has mostly been over when a postconviction litigant establishes that he or she has been prejudiced by counsel’s unconstitutional performance — that is, given that counsel was ineffective, would it have been rational for a defense to opt for trial had he or she been given the advice that Padilla required, i.e., that your conviction will result in adverse immigration consequences and this is why. In this respect, courts have been been split as to whether the decision to go to trial should be based on one’s likelihood of success in prevailing at trial, i.e., lack of evidence of inculpatory evidence and the availability of viable defenses, or on one’s determination to “fight to the death” so he or she could avoid potential deportation that would certainly come with a criminal conviction. The Third Circuit adopted the latter test for Padilla prejudice in its landmark decision in Orocio.
Without further ado, I offer here, cites to a few recent decisions which consider the above issue with varying outcomes:
Pilla v. United States, No. 10-4178 (6th Cir. Feb. 6, 2012) (defendant failed to establish prejudice under Padilla because of “overwhelming evidence of her guilt”).
McNeill v. United States, No. A-11-CA-495 SS, (W.D. Tex. Feb. 2, 2012) (finding Padilla retroactive but denying relief because counsel was not ineffective and even if he were petitioner failed to establish prejudice)
United States v. Fajardo, No. 10-CV-1978, (M.D. Fla. Jan. 26, 2012) (finding Padilla not retroactively applicable after detailed Teague v. Lane analysis)
Yau v. United States, 11 C 8462 (N.D. Ill. Jan. 26, 2012) (granting 2255 petitioner an evidentiary hearing on Padilla claim after finding a sufficient threshold showing of prejudice where it would have been “rational under the circumstances for [the petitioner] to reject the plea agreement and go to trial had he known of the immigration consequences“) (emphasis added).
The decision in the cases cited above should be available on Google Scholar. If not, please feel free to email me and I will send you the decision.
Invisible Man 
I question the reality of immediately addressing the question of “would it have been rational for a defense to opt for trial?” as opposed to “Could (or Should) the Defense have structured the plea (charges) in a way to minimize immigration consequences?”
In real practice, it is often the case that the State has several charging alternatives, and will accept a plea to a lesser included with an equal (or greater) sentence with no immigration consequences. The fact is that the State often has an equivalent desire not to go to trial.
Thank you John for your comment. I agree in part with your comment. This was actually something that Justice Stevens addressed in the Padilla decision with the following passage:
However, this issue, I think, is more appropriately addressed under the performance prong of the Strickland/Padilla test. And an attorney who does try to plea margin for a deportation-safe conviction may avoid being labeled ineffective altogether, i.e., the court would find that he performed reasonably, even if the client still ended up getting deported. A similar situation can be seen in the recent Sixth Circuit decision of Pilla v. United States.