What you're looking at
Each year, roughly 1.5 to 2 million Americans file for bankruptcy. Each filing carries a court fee (set by federal statute) plus an attorney fee (varies by chapter and jurisdiction). Aggregated across 20 fiscal years and 34.9 million cases, the total flow of money paid by financially-distressed Americans to the bankruptcy industry adds up to an estimated $63 billion on the conservative end, and as much as $124.9 billion on the upper end of the raw federal data, depending on how you handle a known double-counting issue in the underlying schema.
The visualization above renders that money as a stack of $100 bills, drawn to true vertical scale alongside well-known altitudes (commercial cruising altitude, Mt. Everest, the Kármán line that defines the edge of space, and the orbit of the International Space Station). Even the conservative figure produces a tower 43 miles tall, six times higher than any commercial flight will ever reach, and high enough that the deepest hole on Earth (the Mariana Trench) could only hold about 16% of it before the rest overflowed above sea level.
Why a range, not a single number
The Federal Judicial Center's national.db dataset is the federal courts' own bankruptcy database, the most comprehensive public-data source for U.S. bankruptcy filings. It contains roughly 37 million case records spanning 2000 through 2026, with one row per case-disposition event.
The schema, however, captures certain events as separate rows that arguably represent the same underlying case:
- Case amendments. When a case is amended (chapter conversion, schedule revisions, etc.), some amendments produce a new row alongside the original.
- Joint-filer dual records. Married couples filing jointly often appear as two records, once for each spouse, even though there is one case and one set of fees.
The raw row count produces $124.9 billion. Halving for documented double-counting produces ~$63 billion. The conservative figure is the safer estimate to lead with; the upper bound is preserved in the visualization (as the dim "$124.9B raw upper bound" label at the top of the tower) so readers can see the full range.
Both bounds are based on the same per-chapter fee assumptions, listed below.
Per-chapter fee assumptions
Each case is multiplied by an average fee estimate that combines the statutory filing fee with a representative attorney fee (national averages):
| Chapter | Cases (FY 2005-2024) | Per-case fee | Total fees (raw) |
|---|---|---|---|
| Ch. 7 (liquidation) | 16,003,690 | $1,838 | $29.4B |
| Ch. 13 (wage-earner repayment) | 18,515,709 | $4,813 | $89.1B |
| Ch. 11 (reorganization) | 377,617 | $16,738 | $6.3B |
| Ch. 12 (family farmer/fisher) | 30,299 | $2,738 | $0.08B |
| All chapters | 34,927,315 | $3,577 avg | $124.93B (raw) |
Filing fees are statutory and set in 28 U.S.C. § 1930. Attorney fee assumptions reflect national averages reported in industry surveys (NACBA, Lawyers.com, FindLaw); local variation is significant. Ch. 11 figures use a conservative consumer-scale estimate, not large-corporation cases (which run into the tens of millions per case but are rare).
Stack-of-bills math
A U.S. $100 bill is 0.0043 inches thick (Bureau of Engraving and Printing reference). At this thickness:
- Conservative ($63B): 625 million bills × 0.0043 in = 2,687,500 in = 223,958 ft = 42.4 miles (rounds to 43).
- Upper bound ($124.9B): 1.249 billion bills × 0.0043 in = 5,370,700 in = 447,558 ft = 84.8 miles (rounds to 85).
True scale in the visualization: 1 mile = 8.235 SVG pixels. All altitude reference points (Kármán line, stratosphere ceiling, cruising altitude, Mt. Everest, Mariana Trench) are drawn at this same scale, so the eyeball comparison is honest. Liberty, Empire State, and Burj Khalifa are too short to draw at this scale (sub-pixel to a few pixels) and are shown as ground-level markers with leader lines.
Caveats and limits
- Fee assumptions are estimates, not measured values. Actual attorney fees vary significantly by district, case complexity, and whether the case is no-asset versus asset.
- The FJC schema's deduplication issue is the single largest source of uncertainty. The 2× range reflects a credible but not yet formally validated halving estimate; the true figure may be anywhere within that range.
- Filing fees for Ch. 7 are sometimes waived under 28 U.S.C. § 1930(f) for indigent debtors. This visualization does not subtract waived fees from the totals; the actual cash flow is somewhat smaller as a result.
- Chapter 11 includes commercial reorganizations. Ch. 11 represents 1.08% of cases but ~5% of fees in this aggregation. The per-case fee estimate ($16,738) is intended as a consumer-scale Sub V baseline, not a large-corporation case (which can run into tens of millions). Excluding Ch. 11 entirely would reduce the totals by ~5% and not materially affect the visualization.
- "Bankruptcy industry" includes the courts and the bar. The fees aggregated here include statutory filing fees paid to the U.S. Bankruptcy Courts, not just attorney fees. Both are paid out of the debtor's pocket (or out of a Ch. 13 plan's distributable funds, which would otherwise go to creditors).
Interpretation
The visualization is not an argument that bankruptcy fees are too high or too low, that's a normative question that depends on what you think attorneys, courts, and the system as a whole should be paid. What it does demonstrate, empirically, is the scale of the system. Roughly 35 million Americans have paid an average of $3,577 each over the past two decades, totaling somewhere between $63 and $125 billion. Forty-two percent of those filings end in dismissal, the debtor pays the fees and walks away without a discharge of debt.
That number deserves to be seen at the right magnitude. Most coverage of bankruptcy treats fees as a per-case line item ("Ch. 13 attorney fees in this district run about $4,000"). Aggregated across the actual scale of the system, the total is on the order of five years of NASA's annual budget. We think that's worth visualizing.
Data and code
Everything used to produce this visualization is openly available:
- Source data: Federal Judicial Center,
national.db(free, public, requires email registration with FJC) - Aggregator script:
_build_fjc_money_data.py, available on request via research@openbankruptcyproject.org - Aggregated state-by-year JSON: available on request
- Visualization source (SVG): view-source on this page
- Bill texture: Series 2009 $100 obverse, public domain via Wikimedia Commons
How to cite
Related research
- 1328(f) Discharge Bar Screening, statutory compliance failures in the discharge process.
- Section 727(a)(8), Ch. 7 discharge-bar analysis.
- Section 109(g), 180-day refiling-bar screening.
- Full OBP methodology, how the project's screening tools work end-to-end.
- All OBP research
Open Bankruptcy Project is a 501(c)(3) public charity. The Internal Revenue Service recognized OBP as tax-exempt under Section 501(c)(3) of the Internal Revenue Code by determination letter dated April 6, 2026, effective March 27, 2026.
- EIN: 41-5159631
- Public charity classification: 170(b)(1)(A)(vi), publicly supported organization
- Mission: publishing free, open-source tools and empirical research on the U.S. federal bankruptcy system
- Funding: tax-deductible public donations only. We accept no attorney advertising, no referral fees, no commercial legal-tech sponsorships
- Contact: research@openbankruptcyproject.org | press@openbankruptcyproject.org