These are fascinating results. However, I wonder if there is more going on in terms of regulatory capture than just employment opportunities. Defendants in major markets are (from a regulatory perspective) repeat players, and often have access to bigger law firms with specialized securities law attorneys, who frequently know the SEC enforcement staffers, at least on a professional level. Defendants outside of a major market, by contrast, may not have as much previous experience with SEC investigations, may hire local criminal defense attorneys, etc. I can imagine this having an effect on the end result.

David, you are spot-on with regard to prosecutors. The skill set one acquires in being an aggressive prosecutor is exactly the skillset that a private law firm seeks when they hire a white collar litigator. It is no different from hiring the CEO of a competitor company to (now) be your CEO: sure s/he's been a jerk before, but now s/he's your jerk.

I only just started reading Gadinis' paper, but it seems that the difference between the treatment of big firms in "favorable locations" may also be attributable to the presence of strong federal (and even state) criminal offices in those same locations. Gadinis is looking only at broker-dealer failure-to-supervise cases and whether the SEC initiates civil or administrative proceedings. But it may be that the SEC pulls back in "big firm" cases in part because its enforcement staff members believe that other law enforcement agencies are examining the same firms.

Just a follow-on from above: Gadinis also may not be capturing those cases in which the SEC staff refer cases of wrongdoing (individuals including entities) to the US Attorneys offices. In fact, I'm not sure he could accurately do this if he tried because the SEC might refer a case to the USAO that an AUSA later declines to prosecute, and the rest of the world would never know.

It seems to me that if you really want to get at the "revolving door" question there is an easier and more direct way to do it. For example, in some USAO offices, it is quite common for prosecutors to enter government service for a limited period of time (say 5-10 years max), and then move on to the private sector. That's how it was at the SDNY when I worked there (and I assume still is). However, in other US Attorneys' offices, it is quite common for AUSAs to stay there for life. A similar dynamic may exist at other government offices across the nation. Therefore, it might be useful to compare firms' treatment between an office that has fairly "high" turnover" and an office that has "low" turnover. At least in the US Attorney context, I would bet that you would find the "high turnover" offices more aggressive than the low-turnover ones.

I doubt the SEC enforcement division would pull its punches if it thought an AUSA was looking at it, unless of course the activity in question did not fall under the SEC's jurisdiction. I would suspect the opposite would be more likely, given the difficulties criminal prosecutors face in securities cases versus what is required in civil or administrative proceedings. However, other elements that might go into an SEC official's calculation is the effect of the sanction. For a small fly-by-night operation, an industry ban might not be that big of a deal. For a Wall Street trader, however, it could be quite a big thing -- the difference between a $50K a year job and a seven-figure income. Likewise, there may not be that much to be gained by fining a firm that is currently facing private litigation by investors, or otherwise promising to make amends -- an industry ban could be sufficient. For a smaller firm, spankying them hard might make more sense and allow for quicker returns to investors. Hypothetically, of course. No idea if that's actually the case, since I can see it going both ways.

Agree with Zaring, but on grounds less cynical and consequentialist. Many government lawyers are true believers who wish to do the right thing.

These are good comments! And I don't want to sound too dark about prosecutors, many of them surely believe in their mission - I just think that the other incentives also line up in favor, at least in some ways, of active enforcement (esp at the high turnover offices, as Miriam suggested).

Thank you all for fascinating comments on my paper! The criminal law thoughts are particularly valuable to me, because so far I have heard mainly from securities and administrative law experts. Here are some first responses to your comments, with additional clarifications and data that I hope to put into the next version of the paper.

MDF points out, and I agree, that large law firms might able to hire better lawyers and lower the penalties they suffer in this way. Unfortunately, the SEC does not release data on who represents defendants in its proceedings, so I cannot examine this hypothesis directly. I try to address this indirectly, by examining sanctions at a subset of cases, those involving fraud, where a likelihood of an industry ban is high. If stakes are very high, small firm defendants are likely to hire as good a lawyer as they can afford, and thus the difference between big and small firm defendants in these cases should shrink. However, I find that the gap between big and small firm defendants remains stable in these cases. Although this is only an indirect test, it indicates at a minimum that the quality of legal representation probably does not account for the full results.

Miriam Bauer helpfully points out that the SEC has a choice about whether to pursue violations itself, or to refer cases to USAO. My sample actually includes cases brought by criminal law authorities where defendants were found guilty. In these cases, after the criminal proceedings end, the SEC typically bans defendants from the industry. In my sample I categorize this as a ban issued after a court proceeding (since these are “non-SEC” proceedings). I apologize if this was confusing, and I will clarify that point in a later draft.

I did not include the above cases in the sub-sample I used to examine the “revolving door” hypothesis because the SEC does not seem to have much discretion in shaping their outcome (it issued a ban in every such case in the dataset). Moreover, I limited tests of the revolving door hypothesis to big firms only. For those interested, here is a breakdown of the criminal cases. Of the 90 SEC actions in the dataset that are either concurrent or subsequent to criminal actions, 79 involve small firm defendants, and 11 involve big firm ones. Of the 11 big firm defendants, 7 were located in the New York/New Jersey area, while 4 were located outside that area.

I also agree generally with MDF that the SEC might like to remain involved even if USAO is looking at a case, because the investigation may reveal facts that, even if they do not give rise to criminal liability, could establish civil liability or regulatory violations. However, I have not collected any data about the timing of these actions, which would allow us to test the SEC’s conduct on this point.

Finally, I appreciate MDF’s point that the SEC enforcement officials may be eager to shut down fly-by-night operations, but refuse to penalize large firms. Thus, my most puzzling result concerns the differential treatment of individual employees in small and large firms. Certainly, the SEC should think twice before depriving a Wall Street trader of his seven-figure income. But, it is hard to see a public policy justification for treating individuals employed in big firms differently than individuals in small firms, when these individuals have committed the same violations and caused harm to investors that substantially exceeds seven-figures. This pattern, which emerges rather robustly from the paper, seems highly problematic.

What this study doesn't control for is litigated versus settled cases. The sanctions are different in litigated cases. As the article does note, big firms almost never litigate against the SEC (they believe they can't survive if they do), while small firms do. My experience is that in a settlement context big firms are often willing to pay a larger amount in fines to avoid charges (or limit their effect) on individuals at those firms. I think that dynamic explains a lot of the perceived difference in small versus large firms.

I'd also note that are a number of errors in this study. To choose just one example, there hasn't been a stock exchange in LA for years, and when there was, it was actually headquartered and had its main floor in San Francisco (which the study counts as a non-stock exchange city).