Damages Calculator vs. Expert Economist: When to Use Which

By MedLegal AI Editorial · 8-minute read · Published April 18, 2026

Every plaintiff attorney running med-mal, nursing-home, birth-injury, wrongful-death, or serious PI cases hits the same decision point several times per case: who builds the damages number, and when? An AI-assisted damages calculator gives you a fast, defensible working number in minutes. A retained forensic economist gives you a report that survives cross-examination and plays to a jury. Those are different jobs. They are not substitutes. They are sequential.

This is a practical guide to which tool belongs at which stage of the case, and why treating them as "either/or" is how plaintiff firms either overspend on experts they didn't need yet or walk into mediation with a number that can't be defended.

What the two tools actually produce

An AI-assisted damages calculator takes the medical records, billing ledgers, wage history, and demographic facts you already have and builds a structured model: economic damages (past medicals, future medicals, lost wages, lost earning capacity), non-economic damages framing, and a present-value calculation using standard discount and growth assumptions. It shows its work. It cites the source pages. It runs in minutes and it costs essentially nothing per case once you have a subscription.

A retained forensic economist produces a signed expert report. They will often use similar inputs, but they add things the AI cannot: a credentialed opinion on worklife expectancy, a jurisdiction-specific view on discount rates, a defensible choice of wage growth data, household services valuation, and the willingness to sit for deposition and trial testimony. They cost real money. Typical ranges for a full report plus depo and trial time run well into five figures. In catastrophic cases they can go higher.

The trap is thinking these are the same product at different price points. They aren't. The calculator is a working model. The economist is admissible opinion testimony. If you confuse the two, you either pay an economist to do intake-stage triage or walk into a Daubert challenge with a working model dressed up as an expert opinion.

Stage 1: Intake triage

This is the cheapest and most valuable place to use an AI damages calculator. At intake you're deciding whether the case is worth the ten-thousand-dollar-plus investment of records acquisition, LNC review, and initial expert consults. You do not yet know whether you have liability. You have a client narrative, an incident date, and some basic demographics.

At this stage the right question is not "what are damages worth?" It's "does the damages ceiling justify taking this case on contingency?" A working model built in fifteen minutes answers that. If a seemingly severe injury comes out with modest economic damages because of the client's age, wage history, or medical trajectory, you want to know that before you order hospital records.

No plaintiff firm should be paying an economist to help them triage cases. That's the job of a fast, internal model. Hire the economist for the cases that survive triage.

At intake the calculator's job is triage, not precision. A number that's within 30% of the eventual verdict value is fine. You're making a go/no-go decision, not a settlement demand.

Stage 2: Pre-suit demand package

By the time you're drafting a demand letter, you have records, you have a working theory of liability, and you have enough medical picture to project future care. You do not yet have a retained expert economist in most cases.

Here the AI calculator does serious work. A demand package needs three things to be taken seriously by a defense adjuster: itemized economic damages with source citations, a non-economic damages framing tied to the medical record, and a present-value calculation of future losses. The calculator produces all three. It produces them in a format the adjuster's own internal tools can parse.

What the calculator does not do at this stage is provide an expert's signature. That's fine — defense doesn't need a signature to open settlement discussions. They need a number they can defend to their reserves committee. A well-sourced, well-structured demand number does that.

The firms that get this stage wrong fall into two camps. The first under-develops the damages model, hands the adjuster a round number, and gets a round low-ball in return. The second spends five figures on an economist's preliminary report before anyone has even offered policy limits. Both waste money. The right move is a detailed AI-built demand with every number sourced, and a note to the file flagging which assumptions the eventual economist will need to defend.

Stage 3: Suit filed, discovery open

This is the transition stage. Once suit is filed and you have defense experts identified, your damages number has to be defensible in testimony, not just in a demand letter. The AI calculator is still useful — it's now doing sensitivity analysis, not the primary model.

This is the stage where you retain your economist. Not before discovery, because you don't yet know what wage documentation, tax records, or collateral-source arguments will be in play. Not after expert disclosure deadlines, obviously. The sweet spot is after initial written discovery and before the first round of defense expert disclosures.

The economist's job at this point is to take the working model and make it their own. They will change assumptions. They will want different wage-growth data. They will swap out a generic worklife table for a jurisdiction-specific one. Expect this. The AI model is scaffolding for their opinion, not their opinion. A good economist appreciates receiving a clean, sourced working model because it tells them what inputs you already have and which facts they still need.

What to hand the economist, and what to withhold

Hand them the calculator output, the source records, the wage documentation, and the life care plan if one exists. Do not hand them the settlement demand that was built on the working model — you want them to build their own number without anchoring to a figure you already used with the defense. If the economist independently lands on a number close to your demand, that's useful. If they land somewhere different, you want to know why before the defense economist does.

Stage 4: Mediation

At mediation you want both tools in the room, conceptually. The economist's report is the anchor: it's the credentialed, admissible, trial-ready number. The AI calculator is your sensitivity layer — what happens to the economic damages if the jury finds a shorter worklife, if collateral-source arguments cut 20% of the medicals, if the jurisdiction's cap on non-economic damages applies?

Mediators respect numbers that bend without breaking. An attorney who can say "our expert has this at $X, and here's the range under the defense's best assumptions" is negotiating from strength. An attorney who can only say "our expert has this at $X" is negotiating from a single point estimate. The AI layer gives you the bend.

This is also the stage where life care planning becomes its own line item. A forensic economist values the stream of future medical costs; a life care planner identifies and prices the items themselves. In catastrophic cases both are required. The AI calculator can draft an initial life care plan outline — a defensible starting framework for the retained life care planner to refine — but it does not replace the planner's clinical judgment.

Stage 5: Trial

At trial, the economist is the witness. The AI calculator is not. No plaintiff attorney should be putting an AI output in front of a jury as a substitute for expert testimony. The calculator's job at this stage is preparation: stress-testing the economist's opinions against defense cross-examination, rehearsing scenarios, and running what-if calculations during trial when a ruling changes the damages landscape.

If the judge excludes a category of damages mid-trial, your economist is going to need to re-run numbers quickly. That's where the calculator earns its keep on day three of trial. Not as admissible evidence, but as a fast, internal tool for the lawyer doing the math on whether to rest early or recall a witness.

A decision table

Where attorneys get this wrong

Two recurring mistakes. First, retaining an economist at intake to "see if the case is worth it." That's spending expert dollars to do triage work. The calculator does triage. Save the economist for the cases that survive it.

Second, trying to take a demand-letter AI number to trial. The demand-letter model was built for persuasion with a defense adjuster. Trial requires a credentialed witness whose methodology can be defended in deposition and cross. The AI output is not that witness and does not pretend to be. Attorneys who skip the economist because "the calculator already gave us the number" are inviting a damages-evidence problem at trial.

The two tools coexist for the life of the case. The calculator does the fast, internal, iterative work. The economist does the credentialed, admissible work. Treating them as complementary — rather than as competitors at different price points — is the stance that saves expert budget on cases that don't need it and protects trial exposure on the ones that do.