An interesting paper has been making the rounds discussing how appellate courts react to caseload pressure. After September 11, 2001, the Second and Ninth Circuits had a large influx of immigration appeals that affect the other circuits, and the paper uses this as a “natural experiment.” In his paper, Mr. Shay Lavie characterizes the Second Circuit as resistant to changing its procedures, and claims that this led to the circuit reversing fewer civil cases as a way of adjusting to increased time pressure. He calls this a “disturbing” reaction to the increased caseload. By contrast, he praises the Ninth Circuit for using procedural flexibility to deal with the extra cases by slightly decreasing the rate of dissents and slightly increasing the number of published opinions. He urges appellate courts to follow the Ninth Circuit’s lead to experiment with more procedural changes.

Interestingly, an earlier article by Mr. Bert Huang analyzed the exact same data and found that both the Second and Ninth Circuits artificially decreased their reversal rates in civil cases in response to the extra cases. He argues that they may have spent less effort on each case, resulting in fewer reversals, as a rational response to having less time. Each article used different statistical techniques to predict what the expected reversal rates would have been without the influx of immigration cases—and then arrived at opposite conclusions.

I am skeptical about the analysis in both articles—certainly neither gives circuits a strong reason to revise their practices or procedures. (Mr. Lavie, in particular, gives very little evidence that changes in procedure produced the results he highlights.) In the Sixth Circuit, we have seen that the time it takes to decide appeals increases substantially as caseload increases, that it takes longer to reverse than to affirm, but we have not seen evidence of a connection between caseload and items like per curiam opinions that might indicate less effort spent deciding cases.

One additional reason to be wary of sweeping policy prescriptions based on changes in a single statistic over a handful of years is that there is so much variation year-to-year. Below are the percentage of civil cases reversed for the Second, Sixth, and Ninth circuits from 1997 to 2016:

Click here to view table.

Looking at this data, my chief question would be what happened in the Second Circuit during 2005 to increase the reversal rate in civil cases from less than 3% to over 11%! Whatever happened during 2005 would doubtless overwhelm any small effects caused by an increase in immigration appeals. And measuring an “artificial” dip in reversal rates during 2003 and 2004 would also be difficult because the rates were already vanishingly low. It is also interesting that the Sixth and Ninth Circuits appear to follow a very similar pattern for changes in reversal rates over time.