The grand bargain

Workers' compensation is a no-fault statutory system. An injured employee gets medical care and indemnity benefits without proving employer negligence. In exchange, the employee gives up tort remedies against the employer. This trade-off is the founding logic of WC and shapes everything that follows.

Before workers' comp existed (in the United States, before the 1910s and 1920s state-by-state adoption), an injured worker could sue the employer for negligence in tort court. Employers had three powerful defenses: contributory negligence, fellow-servant doctrine, and assumption of risk. The result was that most injured workers got nothing, and the small minority who won at trial got large verdicts. The grand bargain replaced that lottery with a schedule. Benefits became automatic, predictable, and capped. Lawsuits became rare. The modern WC system is a state-by-state implementation of that bargain, with each state setting its own benefit levels, dispute procedures, and provider networks.

Part One vs Part Two

The standard WC policy has two parts. Most readers know Part One. The interesting underwriting work happens in Part Two.

Why EL matters

Excess casualty programs sit over EL, not over Part One. The umbrella does not extend WC benefits (those are statutory and unlimited by definition); it extends EL limits. So when a buyer shops a $50M tower, the relevant attachment under WC is the $1M EL primary, not the WC system. EL also catches the cases that fall outside the exclusive remedy: a contractor's employee injured at the buyer's job site sues the buyer (third-party-over), or an injured employee's spouse sues for loss of consortium (consequential BI), or an employee alleges the employer knew of a hazard and intentionally exposed them (intentional act exception in some states).

NCCI, state bureaus, monopolistic states

The National Council on Compensation Insurance (NCCI) is the rating bureau for the majority of US states. It develops class code definitions, loss costs, and the experience rating plan. State bureaus replace NCCI in larger states with their own regulatory frameworks: California (WCIRB), New York (NYCIRB), Pennsylvania (PCRB), Delaware (DCRB), New Jersey (CRIB), Indiana (ICRB), Massachusetts (WCRIBMA), Michigan (CAOM), Minnesota (MWCIA), North Carolina (NCRB), Wisconsin (WCRB), Texas (TDI). Each runs its own class plan and experience modifier methodology.

Monopolistic states

Four states do not allow private WC insurance. Coverage is provided exclusively by the state fund:

• Ohio · Ohio Bureau of Workers' Compensation (BWC)
• Wyoming · Wyoming Workers' Safety and Compensation Division
• Washington · Washington Department of Labor & Industries (L&I)
• North Dakota · Workforce Safety & Insurance (WSI)

Private carriers cannot write Part One in these four states. They can and do write the EL stop-gap. Buyers with multi-state operations thus carry a private WC policy for non-monopolistic states and a separate state-fund policy in each monopolistic state where they have payroll.

Competitive state funds

Several other states (California, New York, Pennsylvania, Maryland, Texas, others) operate state funds that compete with private carriers. The state fund is the market of last resort and the assigned-risk pool administrator. California's State Compensation Insurance Fund is the largest single WC writer in the US. These are competitive, not monopolistic, markets.

Class codes and payroll-based rating

WC premium is calculated as payroll multiplied by a rate per $100 of payroll, by class code. NCCI maintains roughly 700 class codes, each describing a kind of work, with a loss cost reflecting expected losses for that class. State bureaus do similar. The rate per $100 ranges from about $0.10 for low-risk clerical (NCCI 8810) to over $30 for high-hazard classes like roofing (NCCI 5551 in many states).

How a quote builds up

1. Estimate annual payroll by class code and state
2. Multiply by the loss cost (per $100) for each class/state
3. Apply a Loss Cost Multiplier (LCM) per the carrier's filed factor
4. Apply the experience modifier (mod) at the policy level
5. Apply schedule rating credits or debits
6. Add expense constants and assessments

Inclusion and exclusion of payroll

Not all compensation counts as payroll for WC. Bonuses count, overtime usually counts at straight-time (the OT premium is excluded), tips usually count, severance usually does not, owners and officers may be excluded by election, sole proprietors are typically excluded unless they elect in. The audit at year-end reconciles estimated payroll to actual payroll and adjusts premium up or down. Payroll classification disputes are the single most common audit issue.

The experience modifier

The experience mod, or e-mod, is the single most studied number in WC. It compares the insured's actual losses to expected losses for its class and size, over the most recent three completed policy years (excluding the current one), and produces a multiplier centered on 1.00. A mod of 0.85 means losses ran 15% better than expected; a mod of 1.20 means losses ran 20% worse.

Mechanics

• Three-year window. Earliest of the three years drops off each annual recalculation. Current and immediately-prior years are not yet credible.
• Primary vs excess loss split. The primary portion of each loss (typically the first $5K to $20K depending on jurisdiction) is fully credible. The excess portion (above the split point) is partially credible. This dampens the impact of single large losses.
• Credibility weighting. Larger employers have more credibility and a tighter mod range. A small landscaper with one year of high losses gets a mod near 1.00 because their data is not credible enough to swing it. A 5,000-employee manufacturer with a bad year sees a sharper mod move.
• Promulgated by the bureau. NCCI or the state bureau calculates the mod from data submitted by carriers and publishes a worksheet. The mod is not negotiable.

Federal acts: USL&H, Jones, DBA, FELA

Most WC questions live in state law. A small but operationally important subset of workers are governed by federal acts that override or supplement state WC.

Underwriters identify federal exposure during submission review. A buyer with shoreside warehouse operations near a port, a contractor with offshore wind work, a defense contractor with overseas logistics, all need the right endorsement on the right form. Missing an endorsement that should have applied is a hard claim to defend.

Loss-sensitive programs

Above a premium threshold (varies, often $200K+ in standard market, much lower in alternative markets), buyers can buy WC structures that share risk between insured and carrier. These are loss-sensitive programs.

Retrospective rating (retro)

Premium adjusts after the policy period based on actual losses. The buyer pays a deposit (standard premium) and the final cost is calculated as: actual losses + a basic premium + a loss conversion factor + tax multiplier, subject to maximum and minimum premium caps. Buyers with good loss control and predictable losses pay less than guaranteed cost. Buyers with bad years pay more. Retros run on three-year settlement cycles (initial, first adjustment, subsequent adjustments).

Large deductible

The buyer pays the first dollar of every claim up to a deductible (often $250K, $500K, $1M, or higher), with the carrier handling claims and seeking reimbursement. The carrier provides a claim-handling service and the regulatory wrapper; the buyer keeps the bulk of the loss-cost economics. Cash flow is favorable to buyers because they pay claims as incurred, not premium up front. Common in middle market and large commercial.

Captive WC

A captive insurer owned by the buyer (or by a group of similar buyers) writes the WC. Often paired with a fronting carrier that issues the statutorily required policy and reinsures most of the risk back to the captive. Used by buyers with stable losses, sophisticated risk management, and tax / capital efficiency goals.

Self-insured WC

Buyer takes the risk directly, posts collateral with the state regulator, manages claims through a TPA, and complies with state self-insurance regulations. Available in most states above a size threshold. Genuinely large employers often run self-insured WC programs.

Audits and the audit cycle

Every WC policy is auditable. The carrier estimates premium up front based on the buyer's payroll forecast, then audits at year-end to reconcile against actual payroll. The audit can either return premium (if payroll fell short of estimate) or charge additional premium (if payroll exceeded estimate or class assignments shifted).

What auditors look at

• State quarterly tax filings (form 941, state withholding)
• General ledger payroll accounts
• 1099 vs W-2 status of contractors (1099s may need to be picked up if the worker fails the independent contractor test)
• Job descriptions and class code assignments
• Overtime (extracted at straight-time)
• Excluded compensation (severance, certain bonuses)
• Subcontractor payments (in many states, the buyer's policy responds to uninsured subcontractors and absorbs their payroll)

How underwriters evaluate the risk

1. Class mix. What kinds of work, by what payroll proportions, in what states? Class mix is the single biggest pricing input.
2. Loss history. Five years of loss runs, frequency and severity, with attention to lost-time vs medical-only ratios and any shock losses.
3. Mod history. Three or four years of mods. Trend up, down, stable?
4. Operations. What does the buyer actually do? Manufacturing process, contracting type, healthcare specialty.
5. Risk management. Safety committee, return-to-work program, post-injury protocols, OSHA logs (300/300A/301 for prior years), training programs.
6. Federal exposures. USL&H, DBA, FELA, monopolistic state stop-gap.
7. Subcontractor controls. COI verification, hold-harmless agreements, audit history.

Ideal submission pack

Where IDP earns its keep

WC submissions are heavily structured but arrive in unstructured packages. The application is an ACORD form. The losses come from prior carriers in proprietary formats. The mod worksheet is a bureau document with consistent structure. The OSHA 300 is an OSHA-prescribed log. The audit reconciliation runs against the buyer's payroll system. Each of these is a clean extraction job, and the reconciliation across them is a clean validation job.

The WC workflows where Indico shows up

• Loss run normalization. Multi-carrier prior loss data, claim-level detail, lost-time vs medical-only, body part, cause of injury, status, paid/reserved.
• Mod worksheet extraction. Bureau worksheet line items pulled to structured fields with arithmetic validation.
• Class/payroll reconciliation. Match application class/payroll to prior policy declarations and audit results, flag drifts.
• OSHA 300 log parsing. Recordable injury lines extracted, incident rates calculated, year-over-year trends.
• Subcontractor COI tracking. Inbound certificates parsed, expirations tracked, gaps surfaced for compliance dashboards.