Bar insurance in Colorado

Colorado removed the common-law claim against a server before the legislature wrote a statutory one, which is an unusual sequence and it changes what a defense looks like. There is also a damages cap. This page explains the mechanism and deliberately does not print the cap figure, because a number quoted without its current source is worth less than none.

Most states arrived at their present position on service liability by accumulation: a decision, then another decision, then a statute reacting to both. Colorado legislated in the other direction. The statute abolishes any common-law cause of action against a vendor and declares consumption rather than service to be the proximate cause of an injury.

Having deleted the claim, the legislature wrote a narrower one back in. That surviving exception comes with two features an operator ought to know about while there is still time to act on them: damages are capped, and the action has to be commenced inside a single year.

The cap is where this page does something unusual. There is a figure in circulation and we are not going to print it, because it is the old statutory base and the statute has been adjusting it for inflation every two years for decades. The number you would find on most sites is affirmatively wrong. The mechanism is what you need, and the mechanism is what you get.

The coverage side of this — why a room that pours needs a policy of its own, and what that policy does about assault and battery — is on the liquor liability page. Colorado is what this page is for.

The legislature removed the claim before it wrote one

The structure here is worth taking slowly, because it is the reverse of what an operator moving from another state will expect.

Colorado’s statute abolishes any common-law cause of action against a vendor. That is not a limitation on an existing claim or a defense to one; it removes the claim. It then declares that consumption rather than service is the proximate cause of an injury caused by an intoxicated person, which closes off the causal argument the abolished claim depended on. A legislature can also subtract a claim and put nothing back — the repeal our Kansas bar page has to account for leaves an operator reading an entirely different file from this one.

Into that cleared ground the legislature put back a narrow statutory exception. What survives is a route to liability that exists because the statute says so, on the statute’s terms, and nowhere else.

The practical effect is that arguments in Colorado are textual rather than doctrinal. In a state where the claim grew out of case law, a plaintiff can argue that the doctrine should extend a little further. Here there is nothing to extend — there is only a section, and either the facts fall inside it or they do not.

We deliberately do not set out the elements of that surviving exception on this page. Our research on this state carries a limitation, and paraphrasing the conditions of a liability test is precisely the kind of confident sentence that gets quoted back to an operator when it matters. Read the section, or have counsel read it against your facts.

There is a cap, and we are not going to print the figure

Damages under the surviving claim are capped, and the cap is one of the few places where a well-meaning source will hand you a number that is straightforwardly wrong.

The reason is mechanical. The figure most often quoted is the statutory base written in the 1980s. The statute mandates a biennial inflation adjustment, certified by the Secretary of State, and that adjustment has been running for decades. The operative cap today is materially higher than the base — and when we went looking for the certified current figure, both of the Secretary of State paths we tried returned not-found.

So we carry the mechanism and we withhold the number. A capped exposure whose ceiling rises every two years is a different planning problem from a fixed one, and it is the mechanism rather than any single year’s figure that should shape a limits decision.

The second feature of the surviving claim is a clock. The action must be commenced within a single year, which is short by the standards of ordinary tort limitation. For an operator that cuts both ways: it shortens the window in which a claim can be brought, and it means a claim that does arrive is likely to arrive while the night in question is still recent enough to investigate.

That second half is the part worth acting on. A short clock rewards records that exist at the time — incident logs, camera retention that outlasts the retention setting somebody chose without thinking, refusal notes, staffing rosters for the shift. The value of those is highest in a state where the claim will land quickly or not at all.

Where limits are concerned, a cap sitting underneath an umbrella is worth thinking about deliberately rather than assuming the excess layer is idle. The cap governs the statutory claim; it is not a ceiling on everything that can be alleged after an incident in a licensed room.

A class list long enough to be a menu of business models

Colorado’s statutory on-premise classes are unusually numerous, and reading them as a list of business models rather than as administrative categories tells you a lot about how the state thinks about drinking.

They include beer and wine, hotel and restaurant, tavern, brew pub, club, arts, racetrack, optional premises, retail gaming tavern, vintner’s restaurant, distillery pub, entertainment facility, lodging facility and catering licenses. That is a regulator drawing distinctions between a room where people drink, a room where people drink while something else is happening, and a room where the thing being drunk was made on the premises. Who draws those categories matters as much as where the lines fall: the permit types our Connecticut bar page finds enumerated by the legislature itself reach the same opening question on a submission by another road.

For a placement, the class you hold is the first honest answer to the question of what your business is. It frames hours, it frames the expected revenue mix, and it frames whether an underwriter is looking at a hospitality risk, a manufacturing risk sitting inside a hospitality building, or an entertainment risk with a bar attached.

The brew pub and distillery pub classes deserve their own attention for exactly that reason. Production on the premises is not a flavor of hospitality; it is a separate activity with its own equipment, its own hazards and its own products exposure, and it should appear on the schedule as what it is.

One live caution: the regulator is the Liquor and Tobacco Enforcement Division within the Department of Revenue. Guidance still calling it the Liquor Enforcement Division is carrying an old name, and a source that has not kept up with the name may not have kept up with anything else either.

Where the kitchen is genuinely the center of gravity, the placement moves — the Colorado restaurant page carries the food-led side, including the classification argument that decides which of the two a taproom is being underwritten as.

Two peaks, two shoulders, and a room that is shut for one of them

Mountain-resort trade runs on a two-peak year — a ski season and a summer festival season — and the staffing plan and the between-season closures are designed into how the business runs rather than improvised when trade falls away. A twice-peaking year is regional rather than resort-specific, and the two short seasons our Idaho bar page sets against each other hand an operator a state away the same compressed staffing problem.

For a drinks-led room that produces a specific pattern. Both peaks bring crowds who are visiting rather than local, who turn over completely between weeks, and who are unknown to the staff. Everything a neighborhood room gets for free — recognition, reputation, the certainty that a customer will be back on Thursday — is unavailable, so it all has to be written down and everybody has to do it identically.

Seasonal staffing compounds it. The people working the busiest weeks are frequently the newest to that room, which puts the least experienced staff in front of the highest crowd densities. That is a training question and a supervision question, and it sits in the small set of things a mountain operator can genuinely fix between one renewal and the next.

The shoulders are where the liability point lands, and Colorado gives it a shape no neighboring state does. Shutting stops the takings and stops the pouring. What it cannot stop is a claim out of a night in the season just finished — and here that claim carries a deadline with it. The action has to be commenced inside a single year, so the exposure from a ski season expires while the room is still dark, or it does not expire at all.

That cuts two ways for a shoulder-period operator. The window is narrow and knowable: cover on risk through a season has to remain answerable across the year that follows it, and no further. But a claim will also land while the evening is recent, on a room that has been closed for weeks and may already have lost the roster, the footage and the person who was pouring. Ask which policy responds to that night — the one on risk when it happened, or the one on risk when the claim turns up — and ask it while cover is being moved rather than afterwards.

The door, the crowd and the person who has to say no

Whatever the statutory position, two controls carry the weight in a licensed room: checking identification properly, and refusing service without the refusal becoming the incident.

The first has to survive a full house, and a rule that relaxes as the queue grows was never a rule. What reads well is a standard producing the same answer whoever happens to be holding the door, a list of acceptable identification staff can apply without a judgment call, and something written on the night showing the standard was actually running. Where the surviving claim lives entirely inside a section of statute, the argument will be about whether a defined set of facts occurred — so a record either speaks to those facts or it does not help at all.

The second is about people. A refusal is the moment a friendly room stops being friendly, and how it goes depends far more on the temperament of whoever delivers it than on anything set down in a manual.

Door staffing brings a contractual edge in with it. Whether the person on your door is employed or engaged under contract decides who answers for their conduct, and the answer follows how the arrangement runs in practice rather than the name printed at the top of it. Where those two diverge, one incident becomes a liability matter and a payroll matter at the same moment — and the payroll half is the easy one to forget until it is the half that arrives. General liability is only part of what gets asked.

Entertainment is treated differently here than in most states, because Colorado licenses it. Entertainment facility, arts, racetrack and optional premises are classes in their own right, which means the state has already decided that a room where something else is going on alongside the drinking is a different sort of room. If your program has outgrown the class you hold, resolve that with the division before an underwriter resolves it by reading the two documents side by side. Whatever the class, the practical questions do not change: who books the act, who owns the equipment standing on your floor, who is in your room without appearing on your payroll, and who has agreed to answer for the number of people inside.

A year after the last night of a season

Colorado hands an operator an expiry date. The action has to be commenced inside a single year, so an evening in a finished season runs out rather than hanging over the business. Walk that year once, from the last night forward.

On the night, what will ever be known is recorded or it is not: who worked, who poured, what was refused, what the cameras still held, and how the count was held when the room was fullest. A venue that can answer the last is describing a control; one that cannot is describing an intention.

Then the doors shut, the roster goes into a drawer and the recorder keeps overwriting, so a season’s records want moving somewhere that outlasts the closure. The shoulder is also when cover moves — which makes it when to settle which policy answers a night from the finished season: the one on risk then, or the one on risk when the claim turns up.

A renewal in the same stretch is where the year gets described. Two peaks and two shoulders rather than an average that prices payroll wrong twice; when the room is busiest and who supervises the floor then, rather than a closing time; the class beside the alcohol-to-food split from your own records.

Then the anniversary passes and that night is closed. The building is not. Property and equipment breakdown answer to a kitchen working behind the bar, and where the admitted market will not write it, the residual structure and the post-burn flood question sit in full on the Colorado restaurant page.

Major Colorado bar markets

• Denver. More license classes in active use inside one market than anywhere else in Colorado, running from chef-driven rooms with a serious bar through taprooms and gaming taverns to places whose entire business is what gets poured. Programmed entertainment is common enough that capacity management is an ordinary underwriting subject here rather than an exceptional one.
• Colorado Springs. A market with a substantial military and institutional employment base, which spreads the late trade across the week and lifts the busiest hours off the usual Friday-and-Saturday shape. Neighborhood rooms serving a settled local clientele are a wholly different proposition from the resort venues to the west, and the difference is worth stating rather than leaving to be inferred.
• Aurora. A large residential base supporting neighborhood bars whose customers are local and largely known to the staff, which is the condition under which informal social control actually works. Many rooms sit in multi-tenant buildings, so the lease, the shared premises boundary and the certificate requirements a landlord imposes come into the placement early.
• Fort Collins. A university market layered over a dense brewery and taproom trade, which means academic-calendar demand and production exposure regularly turn up in the same file. Rooms that brew what they pour carry manufacturing hazards inside a hospitality building, and that belongs on the schedule as a separate activity rather than folded into a general description.
• Boulder. A compact walkable core holding an unusually mixed set of license classes within a few blocks — taprooms, taverns and rooms trading on an entertainment class — so the crowd on the street at eleven has been served that evening under several different permissions. Age checking is done on strangers here rather than by recognition, which is why it has to be systematic.
• Grand Junction. A Western Slope market with a more conventional trading pattern and a customer base that is resident rather than transient, giving a flatter week and a steadier payroll than the resort corridors. It is also a useful reminder that a Colorado file is not automatically a mountain-season file, and describing which one you are saves an underwriter guessing.