The Data

Bankruptcy Mill Statistics

National data from 4.9 million federal cases across 94 districts. The numbers tell a story that individual anecdotes cannot.

4.9M
Cases in the FJC Integrated Database (2008-2024)
94
Federal judicial districts with case data
~26%
National Ch. 13 formal dismissal rate

National dismissal rates

Chapter 13 bankruptcy has the highest failure rate of any bankruptcy chapter. The formal dismissal rate -- cases where the court enters an order of dismissal -- averages around 26% nationally. But that figure dramatically understates the problem.

When you include all cases that close without a discharge -- voluntary dismissals, conversions to Chapter 7, and cases closed for other administrative reasons -- the true failure rate climbs to an estimated 60-67%. In other words, roughly two out of every three Chapter 13 cases filed in the United States do not result in the discharge the debtor was seeking when they filed.

Why the headline number understates the problem: The FJC's formal "dismissal" count captures only cases dismissed by court order under 11 U.S.C. section 1307(c). It does not count voluntary dismissals under section 1307(b), conversions under section 1307(d), or cases closed without entry of discharge. Each of these represents a case where the debtor did not achieve the intended outcome.

The attorney variation problem

The most striking finding in the data is not the national average -- it is the range. Within a single federal judicial district, where all attorneys practice under the same local rules, before the same judges, with access to the same trustee, attorney-level dismissal rates vary enormously.

Top-performing attorneys
~15%
District average
~30-40%
High-volume mills
~70-85%+

This is not a small variation. In some districts, the gap between the best and worst attorneys exceeds 70 percentage points. An attorney with a 15% dismissal rate and an attorney with an 85% dismissal rate are practicing in the same courthouse, under the same rules, drawing from the same debtor population.

When the variable is not the court, not the judge, not the local economy, and not the law -- the variable is the attorney.

Caseload and outcome correlation

Data analysis reveals a consistent pattern: as attorney caseload increases beyond a certain threshold, client outcomes tend to deteriorate. This is not a universal rule -- some high-volume attorneys maintain strong outcomes -- but the trend is clear and statistically significant.

Annual caseload tier Typical dismissal rate Notes
1 -- 25 cases/year 18 -- 28% Solo practitioners, often selective intake
26 -- 75 cases/year 22 -- 35% Established consumer practices
76 -- 150 cases/year 28 -- 45% High-volume firms, outcomes start diverging
151 -- 300 cases/year 40 -- 65% Mill territory -- volume exceeds capacity for individual attention
300+ cases/year 55 -- 85%+ Extreme volume -- the majority of clients do not receive discharge

The inflection point varies by district and by firm structure, but the pattern is remarkably consistent across jurisdictions: once an attorney or firm is filing more cases than they can individually manage, the data shows it in the outcomes.

District-level variation

Not all districts are equally affected. Bankruptcy mill activity tends to concentrate in larger metropolitan areas where advertising reaches more potential clients and where the debtor population is large enough to sustain high-volume operations.

15-20%
Dismissal rate in lowest-rate districts
35-45%
Dismissal rate in highest-rate districts
2-3x
Gap between best and worst districts

Some of this variation reflects differences in local legal culture, trustee practices, judicial approach, and economic conditions. But within high-dismissal districts, you can almost always find individual attorneys who achieve discharge rates well above the district average -- proving that the local environment alone does not explain the poor outcomes.

Fee patterns

Attorney fee data, available through PACER docket review, reveals patterns consistent with the mill model:

The prior-filer signal

FJC data includes a field indicating whether the debtor has previously filed for bankruptcy. National analysis shows that the prior-filer rate is approximately 27% -- meaning roughly one in four bankruptcy filers has filed at least once before.

Among high-volume mills, the prior-filer rate is significantly elevated, sometimes exceeding 40%. This is consistent with the repeat filer cycle: cases are dismissed, and the debtor -- often represented by the same attorney -- files again.

What 27% means: If one in four bankruptcy filers has filed before, and a significant percentage of those prior filings were dismissed (not discharged), then the system is recycling the same people through the same process -- collecting new fees each time -- without addressing the underlying problem. The 11 U.S.C. section 1328(f) discharge bar was designed to limit this pattern, but it only prevents discharge -- it does not prevent filing.

What the data cannot show

It is important to acknowledge the limitations of this analysis:

These limitations are real but do not undermine the core finding: the variation in outcomes is too large and too consistent to be explained by anything other than differences in the quality of representation.

Data sources and methodology

Primary data: FJC Integrated Database, which records case-level data for all federal bankruptcy filings including chapter, disposition, filing date, close date, and prior-filing status.

Supplementary data: PACER (Public Access to Court Electronic Records) dockets for attorney-level analysis, fee applications, and docket entry detail.

Coverage: 4.9 million cases across all 94 federal judicial districts, filed 2008-2024.

Tools: Analysis performed using open-source Python tools. Code available on GitHub. Discharge eligibility screening available at 1328f.com.

The next question is not whether the pattern exists -- the data answers that clearly. The question is why the system fails to act on it: Why Nobody Stops Them.

PACER charges $0.10 per page. Every document we access becomes permanently free through RECAP. Help fund this research.

Support this research