The $655.6 Billion Mirage
What the U.S. Chamber’s “Tort Costs in America” actually counts, and why Georgia’s $6,247-per-household figure is mostly car insurance, insurer overhead, and the money paid to people who got hurt.
On November 3, the U.S. Chamber of Commerce’s Institute for Legal Reform released the fourth edition of Tort Costs in America, prepared by The Brattle Group.1 The headline: the “costs and compensation paid into the U.S. tort system reached over $655.6 billion in 2025 - equating to 2.12 percent of U.S. GDP and over $4,900 per U.S. household.”1 Georgia gets singled out. The report puts Georgia’s bill at $26.4 billion, or 2.82 percent of state GDP (third-highest in the country), and at $6,247 per household (fifth-highest).2
Those numbers will turn up in op-eds, in committee testimony, and in the preamble of the next “reform” bill. So we read all 40 pages of the report itself, including the methodology appendix and all 49 endnotes. Then we checked its claims against the sources it cites and against the government data it leaves out. Here is what the number actually is.
- Commercial general/professional $336.5
- Personal auto $208.5
- Commercial auto $82.3
- Medical liability $19.5
- Homeowners liability $8.8
Auto, personal and commercial combined, is $290.8 billion, or 44.4 percent of the total. Medical liability is 3.0 percent. Of the commercial total, $186.4 billion is an imputed “self-insured” estimate rather than premiums anyone paid. Percentages are our arithmetic from the report’s tables.2
1. It’s an insurance-premium report in a lab coat
The report states its method plainly on page 6: “Our estimates rely primarily on liability insurance premiums, supplemented by estimates of self-insured exposures.”2 According to the appendix, a premium covers “expected claims, claims administration and defense costs, insurer operating expenses, and compensation for assuming liability risk.”2
Look at that last item. “Compensation for assuming liability risk” is the insurer’s price for carrying the risk, which is to say its margin.2 The 2018 first edition of this same study said it more plainly. The share of the total that did not go to claimants, it said, “covered the cost of litigation of both sides, operating costs for the insurers, and profits to effectuate risk transfer.”3 In other words, insurer profit is counted as a cost of the tort system.
That matters in years like these. According to Verisk and the American Property Casualty Insurance Association, property-casualty insurers’ “Net income after taxes declined to $148 billion from $169 billion in 2024.”7 The report mentions insurers’ “underwriting losses,” but it offers exactly one piece of evidence on insurer returns. Endnote 39 says insurance “returns are lower than those of other industries,” and it supports that with two NAIC reports on 2016 data.2
So when the Chamber says Georgia “pays” $26.4 billion a year for “tort-related costs,” or $6,247 per household, much of what it is counting is money paid to insurance companies, including their expenses and their margin.1,2 The foreword even names the channel: these costs “are ultimately borne by consumers and businesses through increased insurance premiums, prices for goods and services, and reduced resources for investment and growth.”2
2. Paying the injured isn’t a cost of the system. It is the system.
The report’s definition of “tort costs” begins with “(1) compensation paid to claimants through judgments and settlements.”2 So every dollar that pays a crash victim’s hospital bill, replaces a disabled worker’s lost wages, or compensates a family for a death is booked as a “cost.”
The Congressional Budget Office explained what is wrong with that in 2003, in a primer on tort liability that reviewed an earlier generation of insurance-data tort-cost estimates. CBO wrote that some direct “costs” “merely shift money from injurers to victims and thus are not true costs to society as a whole,” and that “the portion of a settlement or judgment that goes to the plaintiffs is a transfer payment.”8 The harm happened when the truck crossed the median. The tort system only decides who pays for it: the party at fault, or the injured person, their family, their health insurer, and Medicaid.
The report knows this. Its introduction says the system “serves two primary economic functions: compensating parties for harms incurred and deterring conduct that may cause future harm.”2 Then the methodology appendix concedes: “We similarly do not quantify the benefits associated with compensating injured parties and deterring harmful conduct.”2 A ledger with one column is not an “empirical analysis.” It is an invoice with the receipt torn off.
How much of the $655.6 billion actually reached injured people? The study says it calculates exactly that. It applies NAIC “loss and loss-adjustment-expense ratios to estimated liability premiums to distinguish amounts paid as compensation, costs of administering and defending claims, and costs of transferring liability risk to insurers.”2 Yet the fourth edition never publishes the result. The 2018 edition did: “57 percent of the tort system costs and compensation was paid in compensation to plaintiffs.”3 If that share held in 2025, about $374 billion of the headline figure would be compensation to injured people. The new report doesn’t say, so treat that as an illustration (our arithmetic), not a finding.
For scale, the CDC estimated that injuries in the United States cost $4.2 trillion in 2019, “including $327 billion in medical care, $69 billion in work loss, and $3.8 trillion in value of statistical life and quality of life losses.”9 The Chamber’s own 2019 figure for the entire tort system was $418.0 billion, roughly a tenth of that.2 Not every injury is anyone’s fault, so this is a sense of scale rather than a like-for-like comparison. Still, the tort system is a small slice of what injuries cost this country.
The foreword’s attempt at sympathy works against it: “According to Bankrate, more than half of Americans cannot afford a $1,000 emergency expense.”2 That is true, and it is the whole point. For those households, an injury nobody pays for is a catastrophe. The people this report counts as “costs” are the people the checks go to.
3. Forty-four percent of it is car insurance, no-fault included
Automobile liability accounts for $290.8 billion of the $655.6 billion. Personal auto alone is $208.5 billion, which the report says is “96 percent” of everything it attributes to individuals.2 For personal auto, the study takes the NAIC’s auto “liability” premium totals and assumes they “represent 100 percent of personal automobile liability.”2 It also lists what those premiums cover. Alongside bodily injury and property damage, the list includes “personal injury protection,” “uninsured/underinsured motorist,” and “medical payments.”2 The NAIC’s own technical notes confirm that the “liability” premium figures include “No Fault or Personal Injury Protection” and “Medical Payments.”10
Personal injury protection is the opposite of tort. In Florida, PIP “covers 80 percent of all necessary and reasonable medical expenses up to $10,000 resulting from a covered injury, no matter who caused the crash.”4 Florida’s no-fault statute goes further. It says a covered driver “is hereby exempted from tort liability for damages because of bodily injury” to the extent PIP benefits are payable.11 So the report treats the premium for a no-fault coverage designed to keep cases out of the tort system as a cost of the tort system. It describes uninsured-motorist coverage, in its own words, as “a form of first party insurance that covers the insured’s own losses.”2 It counts that too.
Auto premiums are driven by crashes and by what crashes cost. NHTSA’s landmark study put the economic cost of motor vehicle crashes in 2019 at $340 billion. That figure includes “legal and court costs” and “insurance administration costs,” and NHTSA found that “the total value of societal harm from motor vehicle crashes in 2019 was nearly $1.4 trillion.”12 The Chamber’s figure for all auto “tort costs” that year was $194.3 billion, personal and commercial combined, with insurer overhead included.2 In other words, everything the Chamber counts on its auto line, overhead included, is smaller than the economic cost of the crashes it pays for.
Why the auto line grew is no mystery either. Traffic deaths nationally rose from 39,007 in 2020 to 42,939 in 2021, an increase of 10 percent.13 In Georgia, according to the Governor’s Office of Highway Safety’s FARS-based series, deaths rose from 1,492 in 2019 to 1,809 in 2021, up 21 percent.14 Repairs got more expensive too. Between 2019 and 2025, the BLS price index for motor vehicle maintenance and repair rose 45.8 percent, against 25.9 percent for all items (annual averages, our arithmetic).15,16
And here is the tell. The report’s marquee finding is that “Total tort costs grew 7.8 percent annually from 2019 to 2025.”2 Over the same years, the BLS consumer price index for motor vehicle insurance grew 7.7 percent a year: 56.4 percent in total, against 56.8 percent for the report’s “tort costs” (our arithmetic).5,2 Within a tenth of a point, the Chamber’s “tort cost” growth rate is the growth rate of the price of car insurance. That is what you would expect from a number built mostly out of insurance premiums.
4. More than a quarter of it is a model’s estimate
Of the total, $186.4 billion, or 28 percent, “reflects self-insured exposures.”2 This figure is not observed data. Verisk’s MarketStance business “estimates the fraction of tort costs not covered by insurance and prices this self-insured liability cost using insurance rates.” The split between insured and self-insured exposure is estimated “Based in part on survey data.”2 Put simply, this is a hypothetical premium for insurance nobody bought, priced at insurance-market rates by a commercial model whose inputs the report does not publish. The report then points to growth in that imputed figure, from about $141.7 billion in 2021, as evidence that “businesses are bearing a growing amount of liability costs directly.”2
5. Ransom insurance and health-plan reinsurance, apparently
Endnote 6 concedes that “tort costs” “include legal claims beyond those governed by the law of torts including sources of statutory liability and some contract claims.”2 The appendix is specific about what that means: “Insurable statutory liabilities include claims arising under labor and employment, consumer protection, civil rights, antitrust, and securities laws.”2 Shareholder suits and antitrust claims are not what most people picture when they hear about frivolous lawsuits. To the extent they are insurable, they are in the number anyway.
Then there is the note to Table 7, which lists the “types of insurance” in the commercial-liability estimate. The list includes “17.0013 Kidnap & Ransom Liability,” “17.0025 Provider Excess Stop Loss,” and “17.0026 Excess Stop Loss.”2 The NAIC’s own product coding matrix defines kidnap and ransom coverage as coverage “for payments demanded by kidnappers for the release of an insured held against his or her will.” It defines excess stop loss as “Insurance coverage extended to a self-insured employer plan to insure against the risk that any one claim will exceed a specific dollar amount or that an entire plan’s losses will exceed a specific amount.”17 That is ransom insurance and reinsurance for employer health plans.
The same matrix assigns those products to annual-statement lines 26 and “13 & 15,” not to line 17, “Other Liability.”17 So one of two things is true. Either the study folded ransom and health-plan coverage into “tort costs,” or its appendix lists coverages it did not actually count. Brattle should say which.
6. Its own sources don’t say what it says
Malpractice. Endnote 41 hints that malpractice suits are baseless. It says a closed-claims study “found that in a large percentage of lawsuits filed, medical errors were not responsible for the claimed injuries,” citing Studdert and colleagues in the New England Journal of Medicine (2006).2 Here is what that study actually concluded: “portraits of a malpractice system that is stricken with frivolous litigation are overblown.” Most claims that did not involve error, “370 of 515 [72 percent],” were paid nothing. And “claims not involving errors accounted for 13 to 16 percent of the system's total monetary costs.”6 In fairness, the same authors also called the overhead costs of malpractice litigation “exorbitant.”6 The Chamber kept the study’s premise and dropped its conclusion. For scale, the report puts medical liability at $19.5 billion. That is 3 percent of its own total, and less than four-tenths of one percent of the $5.3 trillion the country spent on health care in 2024 (our arithmetic).2,18
The trucking “case study.” The report’s only case study is about the defendant. It describes RCX Solutions, an Arkansas carrier whose president said its “insurance rates had tripled” after an $11.3 million verdict, and which “ceased operations in March 2020.”2 It says nothing about the person who was hurt. The Fifth Circuit’s opinion fills that in. RCX’s contract driver “swerved across the median into oncoming traffic, hitting Alexandro Puga’s truck.” Mr. Puga “suffered a variety of injuries, including burns on large parts of his body and fractures in his spine, legs, pelvis, and fingers,” and he “was still undergoing treatment and surgeries two-and-a-half years after the accident.”19 The court affirmed on liability and sent back only the past-consortium award and a settlement credit for recalculation.19 That judgment is what the Chamber presents as a cost.
Old evidence, in-house evidence, industry evidence. To show the system is inefficient, the report leans on a RAND study from 1986, alongside a 2007 study of Texas closed claims. For auto coverage it cites NAIC’s Auto Insurance Database Report 2014/2015, published in 2017. For profitability it cites the 2016 data mentioned above.2 Its international comparison is ILR’s own July 2026 study. Its evidence on how litigation affects where businesses locate is ILR’s own 2019 “Lawsuit Climate Survey” and a 2007 ILR paper, plus a 2008 Commerce Department report.2 Its evidence that litigation is pushing up liability premiums comes from the Swiss Re Institute, the Council of Insurance Agents & Brokers, and the Insurance Research Council, which describes itself as “supported by leading property and casualty insurance companies and associations.”2,20 And according to endnote 4, the whole study “was developed for the U.S. Chamber of Commerce Institute for Legal Reform.”2
7. Florida: the showcase in which costs went up
The report’s star exhibit is Florida’s 2023 tort package, HB 837. Here are its own numbers: “From 2022 to 2025, Florida’s estimated tort costs rose from $48.2 billion to $56.2 billion,” and costs per household “rose from $5,768 to $6,149.”2 That is up 16.6 percent in dollars and 6.6 percent per household. The claimed improvement is that Florida’s economy grew faster than its tort costs, so the share of state GDP fell from 3.35 to 3.11 percent, and that Florida slipped from third to seventh in the per-household ranking. The report itself concedes that a change in rank “cannot, by itself, establish the reforms’ effect.”2 Even after that decline, Florida has the highest tort costs as a share of GDP of any state in the report.2
Florida does have a better number, and we’ll give it to them. Its dollar growth from 2022 to 2025 (16.6 percent) was slower than the nation’s (24.0 percent; our arithmetic from Table 1).2 And the Verisk/APCIA 2025 release reports that “homeowners and auto insurance losses and rates in Florida in 2025 declined significantly following legal system abuse reform.”7 That is good news for Florida policyholders.
Now look at how Florida got there. HB 837 moved Florida from pure to modified comparative negligence. According to the Florida Senate’s bill summary, which is the same document the report cites,2 “a plaintiff who is found to be more that [sic] 50 percent at fault for his or her own harm may not recover damages from any defendant.”21,22 The law also cut the limitations period for general negligence claims from four years to two.21 Every injured Floridian who now recovers nothing shows up in the Chamber’s metric as “progress.” A measure that cannot tell a more efficient system from one that simply stops paying injured people cannot tell you whether reform worked. And under the report’s own method, Florida’s figure includes its mandatory no-fault PIP premiums.2,4
8. What this means in Georgia
The report’s Table 4 breaks Georgia’s $26.4 billion down as follows: $13.9 billion in auto (53 percent), $11.4 billion in general and commercial liability (43 percent), $710 million in medical malpractice, and $362 million in homeowners liability.2 More than half of Georgia’s “tort cost” is auto insurance. That is in a state where traffic deaths peaked at 1,809 in 2021 and then fell to a preliminary 1,422 in 2024.14
The report explains Delaware’s chart-topping figure as “partly reflecting its concentration of corporate headquarters,” because “certain commercial risks remain concentrated at headquarters.”2 It does not say whether the same effect reaches Georgia, which had 17 companies on the 2025 Fortune 500 list, including Home Depot, UPS, Delta and Coca-Cola.23 Georgia readers deserve that answer before the $6,247 figure is read into the record.
The foreword also sets up the sequel. Governor Kemp signed SB 68 on April 21, 2025.24 It makes seat-belt evidence admissible on issues such as negligence, causation and apportionment of fault. It restricts “anchoring” arguments about noneconomic damages. And it lets defendants show juries “not only the amounts healthcare providers charge for such expenses but the amount actually needed to satisfy those charges.”25 ILR says the impact of SB 68 and SB 69 “is not meaningfully captured in this study’s estimates, but future editions will provide an opportunity to assess their impact.”2 In plain terms, the next edition will report Georgia “improving.” When it does, ask the one question this method cannot answer: did Georgians get hurt less, or did they just get paid less?
9. In fairness, and what an honest study would measure
There is real evidence that some liability costs have outpaced inflation, and we won’t pretend otherwise. RAND’s 2024 study of “social inflation” found that, controlling for case characteristics, inflation-adjusted personal-injury “trial awards per plaintiff grew at a 7.6-percent compound annual growth rate (CAGR) between 2010 and 2019.”26 But the same study also found that “The overall decline in frequency between 2001 and 2019 was large enough to nearly or fully offset the increase in severity over that period for all the market segments considered.”26 And a 2023 article in the NAIC’s own Journal of Insurance Regulation concluded that “there is not enough data for a regulator or legislator to confidently embrace further legal system reform,” warning that “there is significant risk that the reforms would do more harm than good.”27
An honest accounting would put injuries in one column and compensation in the other. How many people are hurt by someone else’s negligence? How much of their loss gets paid, by whom, and how fast? What share of each premium dollar reaches the injured? What is deterrence worth: the recall that happened, the truck that got inspected? CBO framed the test in 2003. Indirect costs, it said, “are not necessarily bad: from the perspective of efficiency, the question is whether they yield benefits larger than the costs.”8 Tort Costs in America never asks that question.
The bottom line
The $655.6 billion figure is real arithmetic, but it isn’t what it is sold as. Most of it is what Americans and their employers pay insurance companies. Added to that is a model’s estimate of what self-insured businesses would have paid them. Nowhere in it is the person at the other end of the check: the driver with the fractured spine, the family with the funeral bill, the malpractice patients in the Chamber’s own cited study, where “Sixty-three percent of the injuries were judged to be the result of error.”2,6 Count all of them as “costs,” and any law that stops paying them will look like savings. That is not a measure of a legal system. It is a sales pitch for cutting one.
Sources
- U.S. Chamber of Commerce Institute for Legal Reform, “Tort Costs in America – Fourth Edition” (Nov. 3, 2026). Read the source (uschamber.com)
- David McKnight, Paul Hinton & Adrienna Huffman (The Brattle Group), Tort Costs in America: An Empirical Analysis of Costs and Compensation of the U.S. Tort System, 4th ed. (U.S. Chamber Institute for Legal Reform, Nov. 2026). Read the source (uschamber.com)
- Paul Hinton, David McKnight & Lawrence Powell, Costs and Compensation of the U.S. Tort System (U.S. Chamber Institute for Legal Reform, Oct. 2018) (copy hosted by Law.com). Read the source (images.law.com)
- Florida Department of Highway Safety and Motor Vehicles, “Florida Insurance Requirements”. Read the source (flhsmv.gov)
- U.S. Bureau of Labor Statistics, CPI-U series CUUR0000SETE, Motor vehicle insurance. Read the source (data.bls.gov)
- David M. Studdert et al., “Claims, Errors, and Compensation Payments in Medical Malpractice Litigation,” 354 N. Engl. J. Med. 2024 (2006). Read the source (nejm.org)
- Verisk & American Property Casualty Insurance Association, “Strong 2025 Underwriting Income Masks Persistent Property/Casualty Insurance Pressures” (Mar. 25, 2026). Read the source (verisk.com)
- Congressional Budget Office, The Economics of U.S. Tort Liability: A Primer (Oct. 2003). Read the source (govinfo.gov)
- Cora Peterson et al., “Economic Cost of Injury — United States, 2019,” 70 MMWR 1655 (Dec. 3, 2021). Read the source (cdc.gov)
- National Association of Insurance Commissioners, 2024 Auto Insurance Database Average Premium Supplement, Technical Notes. Read the source (content.naic.org)
- Fla. Stat. § 627.737 (2025) (Tort exemption; limitation on right to damages). Read the source (flsenate.gov)
- National Highway Traffic Safety Administration, “NHTSA: Traffic Crashes Cost America $340 Billion in 2019” (Jan. 10, 2023). Read the source (nhtsa.gov)
- NHTSA, Traffic Safety Facts: 2021 State Traffic Data, DOT HS 813 509 (Sept. 2023). Read the source (crashstats.nhtsa.dot.gov)
- Georgia Governor’s Office of Highway Safety, Preliminary Motor Vehicle Traffic Fatalities in 2024 (Jan. 2026). Read the source (gahighwaysafety.org)
- U.S. Bureau of Labor Statistics, CPI-U series CUUR0000SETD, Motor vehicle maintenance and repair. Read the source (data.bls.gov)
- U.S. Bureau of Labor Statistics, CPI-U series CUUR0000SA0, All items. Read the source (data.bls.gov)
- National Association of Insurance Commissioners, Uniform Property & Casualty Product Coding Matrix (effective Jan. 1, 2026). Read the source (content.naic.org)
- Centers for Medicare & Medicaid Services, “NHE Fact Sheet” (historical NHE, 2024). Read the source (cms.gov)
- Puga v. RCX Solutions, Inc., 922 F.3d 285 (5th Cir. 2019) (substituted opinion filed Apr. 17, 2019). Read the source (ca5.uscourts.gov)
- Insurance Research Council, “About IRC and its Mission”. Read the source (insurance-research.org)
- The Florida Senate, 2023 Bill Summaries, CS/CS/HB 837 — Civil Remedies (Ch. 2023-15, Laws of Fla.). Read the source (flsenate.gov)
- Fla. Stat. § 768.81(6) (2026) (Comparative fault). Read the source (flsenate.gov)
- Miguel Legoas, “Fortune 500’s list of the biggest corporations includes 17 in Georgia,” Augusta Chronicle (June 4, 2025). Read the source (augustachronicle.com)
- Georgia General Assembly, SB 68 (2025–2026 Regular Session), status history. Read the source (legis.ga.gov)
- Jones Day, “Georgia Enacts Major Tort Reform Legislation” (Apr. 2025). Read the source (jonesday.com)
- Lloyd Dixon, Nicholas M. Pace, James Davidson & Jamie Morikawa, What Is the Evidence for Social Inflation? Trends in Trial Awards and Insurance Claim Payments (RAND Corp. 2024). Read the source (rand.org)
- Kenneth S. Klein, “The Case for Pausing Any Immediate Embrace of the Social Inflation Argument for Legal System Reforms,” Journal of Insurance Regulation (NAIC 2023). Read the source (content.naic.org)
Further reading
- Center for Justice & Democracy, “False, Inflated Drivel: What’s Wrong With the U.S. Chamber of Commerce’s ‘Tort Cost’ Report” (critique of the 2022 edition). Link (centerjd.org)
- Lawrence Chimerine & Ross Eisenbrey, “The Frivolous Case for Tort Law Change: Opponents of the Legal System Exaggerate Its Costs, Ignore Its Benefits,” Economic Policy Institute Briefing Paper No. 157 (2005). Link (files.epi.org)
- Michelle M. Mello et al., “National Costs of the Medical Liability System,” 29 Health Affairs 1569 (2010) (estimating medical liability system costs, including defensive medicine, at 2.4 percent of health spending). Link (pmc.ncbi.nlm.nih.gov)
- Tort Costs in America, 3rd ed. (U.S. Chamber Institute for Legal Reform, Nov. 2024). Link (uschamber.com)