Arnold Kling  

The Economics of Fire Protection

My Speech in Dallas... History of Fire Protection...

Surely, someone has done this?

I assume that fire-fighting is an industry with declining average costs. Suppose that it takes $1 million in fixed costs per year to maintain the fire department (that includes normal profit, aka opportunity cost), the variable cost per fire is $10,000, and on average you put out 100 fires a year.

Your total costs are $2 million. If you try to make them back by charging $20,000 per fire, a competitor will come in and charge less. So you cannot sustain that price.

The cost structure implies natural monopoly. A two-part tariff is probably optimal. You charge a "membership fee" to cover the $1 million in fixed costs and then charge a fire-fighting fee of $10,000 to put out each fire, paid by the owner who has the fire. Some customers might want insurance against having to pay the fire-fighting fee, in which case they would prefer to pay a higher "membership fee" and a lower fire-fighting fee. Depending on the degree of moral hazard, the fire company might provide this insurance, perhaps even charging nothing for fighting the actual fire.

The more members a fire company has, the lower the membership fee. Hence the tendency toward natural monopoly.

The interesting question, in light of the recent controversy, is what happens when a consumer opts out. What would the fire company do about a non-member's fire? In theory, it could charge the variable cost of $10,000, but that would eliminate the incentive to be a member. So it has to charge a lot more than that to non-members.

Suppose that there are two fire departments that can obtain enough members to be profitable (perhaps their services are slightly differentiated, so neither drives the other one out of business). Now, a fire breaks out at a home where the owner is not a member of either fire protection service. The owner offers to pay to have the fire put out. Do the two companies compete the price down to $10,000? If so, then how do they convince people to pay membership fees? If not, then how do they enforce what amounts to collusion?

Is there a libertarian solution for declining-average-cost industries?

Comments and Sharing

COMMENTS (21 to date)
ChanceH writes:

Please include my negative utilitons generated by any non-voluntary fees (even ones that don't affect me) in your optimality calculation.

mdb writes:

I don't see how this is much different than car insurance. The more people you insure the more you will pay out. I'm sure there are probability tables used by home insurers that calculate the likelyhood of a fire. Use those, come up with a deductible scale and you're in business.

mdb writes:

You could also look at the sea tow model. They sell insurance to tow your boat if you get stranded, you can also call them if don't have insurance. The thing is they know the salvage laws, if anybody tows you, they can claim the boat (total loss for you) unless there is contract. You pay a LOT more if you don't have insurance.

OneEyedMan writes:

If there really were a firm with a cost and revenue structure like the one above what would the incentive be for anyone else to enter that business? The current firm is already making zero profits.

Mutual aid societies have done quite well in fire fighting without resorting to government monopolies, just the use of social pressure.

I also doubt that there are infinitely declining average costs of fire service. I don't believe it would be efficient to have one national fire fighting service the way we have one military.

In the housing business, the mortgage owning companies might be natural providers of fire fighting services because they have scale and an incentive to protect a diversified share of homes. The already require homeowners insurance on most mortgages so they could just require fire insurance as well.

In a rural area there might be a problem with having a competitive fire services market much like people don't know how to apply vouchers to rural education. That said, in any large urban area they could support multiple fire fighting service providers. 911 could even route you to your provider automatically and there could be contracts between providers when they are too far or busy. That's the ambulance / hospital model which is another industry with declining average costs.

david writes:

The companies do indeed bargain the price down to $10,000, and they predictably fail to convince people to pay membership fees, and their geographical coverage shrinks until they no longer compete and can then charge a membership fee.

Supposing their services are just slightly differentiated, expect further differentiation as another method to mitigate competition. Also a lot of expensive signaling and attempts to pre-commit towards protecting or not protecting an area.

Marty writes:

Establish either by law or contract a mechanic's lien on the house and underlying property, value to be the cost of putting out the fire, maybe with an adjustment if the building's value post-fire is less than the cost (i.e., lien would be cost plus X, X being a factor to account for some properties' value below the cost of fire-fighting.

Tricky because of the potential that the after-fire value is too low, but by including the land (generally an improved, developable parcel) this can be greatly mitigated.

If the mechanic's lien is done by law, it's a less libertarian answer than if done voluntarily, but an unconstrained fire can spread to adjacent property, and the effect on neighboring properties of a burned-out hulk, so the issue is not entirely private.

Also tricky if the owner is having major work done on the house, or had it done but not yet paid, because of competing mechanics liens. But in theory something like this might work.

RPLong writes:

The "libertarian solution" has existed for over one hundred fifty years throughout the rural Western states in the form of volunteer fire departments. They work really well.

kurt writes:

If a fire department puts out a non-member fire, it can reimburse the proceeds to the members, not?

Pavel writes:

But how would they differentiate their services sufficiently for there to be 2? A fire department's most important service is fighting fires. While they do a lot of other stuff, not much of it is interesting enough to really differentiate.
Also, the cost of finding the cheapest alternative when your house is on fire is pretty high. So the consumer is likely to take the first best offer.

mobile writes:

Selling bags of potato chips from convenience stores is also a declining-cost industry. Suppose there are two convenience stores across the street from each other with variable costs of $1 per bag. Both stores mark their price up to $2 per bag. Now a customer arrives who is willing to pay for some potato chips. Do the two stores compete the price down to $1 per bag? If not, how do they enforce what amounts to collusion?

Lewis writes:

What if there had been another fire 5 minutes after the first fire (non paying) fire had started? How would everyone's opinion of this whole thing change?

JPIrving writes:

Not perfectly libertarian, but why not have the government continue its role of providing universal fire protection, but contracting the fire fighting to private firms? In this way we still have the general problem of falling average cost, but through competition we can lower the fixed and variable costs over time.

They do this in Denmark.

William Polley writes:

Not an economic analysis like you're probably looking for, but a very interesting paper on the development of fire departments...

Steve writes:


Volunteer fire departments are not unique to the west, nor are they necessarily libertarian. Most New England towns have volunteer fire departments that are funded by the town.

@everyone else:

Firefighting is not the equivalent of fire insurance. Saying that the solution is merely to construct an actuarial table ignores the part of the problem that makes it interesting: i.e., that in order to fight fires a station must be maintained at a high fixed cost. If all you're going to do is pay to replace the structure, then the high fixed cost goes away (although the variable cost is much higher).

Maybe the solution is to point out that Arnold has already solved it: he has shown a logical contradiction, thereby refuting his premises. As david comments, regional monopolies may be the equilibrium that the market comes to. The only difference between this outcome and current is that if a fire department is poorly run, depending on the level of barriers to entry, the department could be supplanted by an upstart.

floccina writes:

I have heard that since the era of fire retardants, smoke alarms, sprinkler systems etc. we over invest in fire departments.

I wonder if outside dense cities and manufacturing areas considering the alternatives, fire departments are a good value. I wonder if an insurance company would charge more or less than the cost of fire departments to insure a home without fire protection. I wonder how much home value fire departments are able to save per year. (I have heard complaints that the fire departments did more harm than good in some cases due to water damage in a small contained fire).

Another thing, if a block home and a wood home are both taxed the same doesn't that create a moral hazard? What are the annual cost of sprinkler systems compared to the cost of fire departments.

Floccina writes:

Oh I forgot to mention how many lives more are saved by having fire departments verses those lost due to factors related to the tax, the moral hazard and other factors.

Robert writes:

I would second Mobile's slightly cryptic post by asking: Aren't declining average cost industries very common? Perhaps even a majority of industries?

Lots of things have high fixed costs and essentially no marginal costs, like computer software and games, ebooks and online music downloads, newspapers (perhaps anything where you pay for information).

Stefano writes:

I think the two companies would set non-member prices in advance, and stick to them. Those prices would be high enough that the current members have no incentive to quit. There's no need for the prices to be fixed amount, they could be an hourly rate.

But how would they differentiate their services sufficiently for there to be 2?

Perhaps one company uses only water, the other some less property-damaging chemicals. One could have faster trucks and can intervene earlier. Or have better gear, or better trained men and do a better job.

David W writes:

I think mobile's point is the right one - there are in fact a lot of industries with the cost structure you describe - telecoms, restaurants, gas stations, car dealerships, airlines, and so on. They've pretty much all figured out ways to compete and remain profitable.

For the fire case, a lot of the competitive differences between companies would come down to location and hence response time. They might also compete on equipment and cost - I'd have to decide if I'd rather subscribe to the firm with the bigger engines, or the one with the faster response time. And if I'm not subscribed to either, I still have little incentive to spend a lot of time negotiating the price - my house is burning down! Probably the one who gets there first, with a reasonable offer, gets my business even if they're not the absolute cheapest possible.

An analogy here would be McDonald's - there are several in town, but the one closest to me gets 95% of my McDonald's business, because prices are similar, food is identical, and they're convenient. Even if the further one is somewhat cheaper, I usually ignore that in exchange for convenience.

Mike P writes:

One thing to consider is that the vast majority of fire departmetn calls are not for fires. They are often sent to medical calls as well as to rescues or extractions.

And even their fire calls tend more often to be either false alarms or small fires that could often be contained and extinguished without their help.

Josh writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Comments for this entry have been closed
Return to top