Economics and the art of automobile maintenance

I've been taking a mental health break from economics. I've been buying and fixing up a 1994 Mazda MX6. It's nearly fixed up now. And so am I. So here's a light Summer post.

Why do some people fix their own cars?

The meaning of a question, and what counts as an answer, depend on the theoretical perspective you approach it from. I loved reading "Zen and the art of motorcycle maintenance". What he says is important. But he's not tackling this question from an economist's perspective.

To us economists, it's a puzzle why anyone would fix his own car. Not a very deep or insoluble puzzle, but still a puzzle that needs answering. Because what we teach in ECON1000 is that economies of scale (Adam Smith) and comparative advantage (David Ricardo) motivate specialisation and trade. I should stick to economics, and pay someone else to fix my cars. But I don't. Usually I fix my own cars myself. Why?

1. Because I enjoy doing it. In simplest micro models of the supply of labour (which is just 24 hours minus the demand for leisure) there are two ways to spend your time: working; and leisure. Leisure gives you more direct utility than working, but working gives you income, which you can use to buy things, which gives you utility indirectly.

But there are many different kinds of "work", and many different kinds of "leisure", and each of those many ways of spending your time gives you different amounts of direct utility. And some "work", like fixing your own car, produces goods (a fixed car) rather than money, and (for some people) is more enjoyable than their regular work (or less unenjoyable), and as enjoyable as some forms of leisure. All this gets ignored in simple models of specialisation and trade.

2. To avoid taxes. Suppose that both I and the mechanic face a 50% marginal tax rate. (It keeps the math simple, and is also roughly correct, or even an underestimate, if you add sales taxes to income taxes.) I could teach his kid economics, for cash, and he could fix my car, for cash. Suppose I can teach twice as much economics per hour as he can. Suppose he can fix twice as many cars per hour as I can. Even though I have a 4:1 comparative advantage over him in teaching economics (and he has a 4:1 comparative advantage over me in fixing cars), the 50% tax rate on each of those two trades means we will be indifferent between trading or not trading. Each of us has to earn $400 gross to give the other one $100 net. In a monetary exchange economy, a 50% tax rate on cash income becomes the equivalent of a 75% tax rate on barter, because there are two trades that get taxed (economics for cash, cash for car fixing), and not just one barter trade (economics for car fixing). (I wonder; do microeconomists, with their tendency to use formal models of barter exchange, recognise this point?)

3. Moral hazard/asymmetric information. (These are really the same, because if there were symmetric information, so you knew everything the mechanic knows, you would know if he were slacking off and doing a shoddy job because it's not his car.) My experience suggests that the biggest problem here is not mechanics doing too little, but too much -- fixing things that don't need fixing? I think that's the main fear of people who don't understand how cars work. Because you can usually tell if something hasn't been fixed that needs fixing; but you can't tell if something's been fixed that didn't need fixing.

4. Monopoly power. The perfectly elastic demand curves facing perfectly competitive firms are a simplification. Most sellers face a downward-sloping demand curve, so price above marginal cost. This means that the quantity of trades that take place is less than socially optimal. In other words, under monopoly, there is less trade in car-fixing and more people fix their own cars. Some potentially mutually advantageous trades, motivated by economies of scale and comparative advantage, are not made. It would cost the mechanic less to fix my car than it costs me to fix it, but i do it myself because he charges me a price above his marginal cost.

5. What else have I missed?

6. Update: Transportation costs. I shouldn't have forgotten that transport costs nearly always limit the gains form trade. "The division of labour is limited by the extent of the market". Getting my car to the mechanic (walking back), and getting it back (walking to the garage) are a major cost. Plus, He fixes the car when he has time; I have to book it in. I can fix it when I don't need the car as much.

And a few anecdotes, that don't really belong in this post, but I can't resist telling them.

A1. The dealer quoted my $80 for a new passenger side window switch. This is a very simple 3-position rocker switch. A generic version would probably costs less than $10 at an electronics store, but wouldn't fit the MX6's door and connector. So I spent an hour taking the tiny switch apart, cleaning the contacts, and putting it back together. Why is the cost so high? Partly it's the cost of maintaining inventory of a very large number of very specialised parts. But I read you can buy the switches for around $20 on Ebay (I couldn't be bothered with the hassle of Ebay), and the Ebay seller must also have similar inventory costs. It must be monopoly power. If car manufacturers could precommit to keeping parts prices reasonable, and if buyers paid attention to future parts prices when buying a new car, competition between car manufacturers would probably keep parts prices lower. But neither of those two conditions seem likely to be fulfilled.

A2. There are significant interprovincial trade barriers to buying a used car. I live in Quebec. If I buy a used car that's already registered in Quebec I don't need a safety check. But I bought an Ontario-registered car, and the safety check at one of a few licensed garages was mandatory. It cost $80, which is not unreasonable. The trouble is, when your car goes in for the safety check and they find anything major wrong (nail in tire in my case!) you are not allowed to drive the car away from the garage doing the check. If they find anything minor wrong (spare tire not anchored, battery not anchored, one slightly worn ball joint, one sticking strut mount bearing) you have 48 hours to get it fixed or you can't drive the car. I could have fixed it all myself, but maybe not in 48 hours, over a weekend, so they had me over a barrel. I paid them to fix it.

A3. I used to fix cars as a kid. Then I got busy, and cars got complicated, and I stopped. I started again because computers have made it easier to fix your own car again. First, the car's own computer can often tell you what's wrong (you need to invest in a $100 code reader to diagnose a check engine light, or just use a bent paper clip in the case of a pre-1995 car). Second, because the internet has made it much easier to compare notes with other people fixing their Mazda MX6. Car forums (linking my two internet worlds) contain a wealth of information and advice. They are like blogs, but different.

Comments

[Frances Woolley]

"Some potentially mutually advantageous trades, motivated by economies of scale and comparative advantage, are not made."

Two thoughts:

I'm a little surprised you didn't refer to the new economics of happiness literature in your point (1), e.g. this speech by Ben Bernanke. He says

"Another factor in happiness, perhaps less obvious, is based on the concept of "flow."12 When you are working, studying, or pursuing a hobby, do you sometimes become so engrossed in what you are doing that you totally lose track of time? That feeling is called flow. If you never have that feeling, you should find some new activities--whether work or hobbies."

In general, there are far fewer economies of scale in repairing than in manufacturing - how many durable goods other than your car do you ever repair? If your fridge needs repairs, it's probably cheaper to buy a new one. (Though I did fix an electric mixer once.)

[Mike Moffatt]

(I wonder; do microeconomists, with their tendency to use formal models of barter exchange, recognise this point?)

I remember it coming up in one of Robin Boadway's PhD classes in Pub Finance at Queen's.

I don't think it's all that realistic, though, because it suggests that there should be more barter in a no-tax economy than a tax one. But one of the benefits of barter is that it's helping in avoiding/evading taxes, so it's likely the other way around. (i.e. where would you expect more barter - the USSR or the USA?)

[Tim Worstall]

Err, you like fixing your car. People do things they like.

Economist discovers utility maximisation.....film at 11.

[Luis Enrique]

Nick,

One of the things rich economies may choose to spend their wealth on is switching to modes of production that increase happiness. What I mean is that fixing your own motorcycle is probably less efficient than paying professionals, but it may make you happier, and if an economy is rich it might make sense to sacrifice some productivity for happiness. Perhaps as we get richer, the margins are moving, and maximizing output productivity is not longer as close to welfare maximizing as it is when poorer. Many individuals do this sort of thing already, of course, by quitting demanding jobs and choosing a combination of lower income but better lifestyle.

However, I think we can still ask whether competitive free market economies are very good at facilitating this transition. What I have in mind is working hours. People often say we'd be happier if we worked fewer hours, but there's a co-ordination issue. If you try to work a 35 hour week whilst your colleagues are putting in 50, you won't get ahead. In most jobs, it's far easier to negotiate a pay rise than it is to negotiate a pay cut plus move to a 4 day week, even though many people may prefer the latter if available. I know working weeks have been declining steadily for decades (or at least, I think so) but it's still possible that we work above the welfare-optimizing level of hours. Of course, it's really impossible to say, but I think the possibility is worth taking seriously.

Might it be similarly difficult to move to happiness-increasing-yet-less-efficient modes of production? Could it be the case the the competitive economy makes that very hard to do, because in most cases maximizing output productivity is necessary for commercial survival? So might the gap between the market outcome and the welfare maximizing outcome start to widen as we get richer? I don't know, this is speculative and tentative. But I think there might be something in it.

[Nick Rowe]

Frances: yes, the new happiness literature does seem relevant. I don't know much about it though.

Mike: I'm not sure if the point you thought I was making about barter was the same one I was trying to make. Let me have another try to be clearer:

If both I and the mechanic have a 50% tax rate on cash sales, then a microeconomist who models me and the mechanic as if we were doing barter should assume a 75% tax wedge in the Edgeworth box. In other words, maybe taxes are higher than microeconomists think they are?

Tim: U max sure. But subject to what constraints? And how do we model the direct and indirect effects of choices on utility? There's more to economics than "We do what we like doing".

Luis: but if fixing your motorbike yourself makes you happier, it *is* more efficient (in the strict economist's sense of "efficient"). (Yes, I know what you mean though, just being picky).

But the First Theorem of Welfare economics (all bow) does tell us that competitive equilibrium (absent externalities) will Patero-maximise human happiness. People in fun jobs will work for lower pay, ceteris paribus.

It's just getting light. So I'm off for a long drive to the cottage in my new car!

[Luis Enrique]

of course - stupid of me. I think I was misled by thinking that if anything artisans etc. tend to get paid more than factory workers, but that's because their output differs in quality.

I wonder, does one need special conditions to obtain the result that, because wages will be lower for jobs that are more fun, welfare optimizing modes of production will be chosen in competitive equilibrium? You mention absence of externalities, but I'm thinking about frictions or maybe some strategic behavior (departing from the world of perfect competition) that means the welfare optimizing modes of production are not an equilibrium?

[Peter T]

1. Where is the evidence that wages are lower for jobs that are "more fun"? How do you measure it, given that people rapidly adapt their expectations to their realities? I rarely meet people who admit they do not like their job.

2. Money is a means of exchange, but also a means of social power. The peasant is well-advised to minimise his or her money transactions - money is the weapon of the landlord, the tax-gatherer and other people with opposed interests. In other words, monetary exchange cannot be assumed to be equal in the absence of other social forces. In modern times, the unemployed find this out quite rapidly.

3. The trade-off is not between money and leisure, but between money and time. For most, money is scarce and dedicated first to the things only it can buy. Time is more elastic. Mothers know this well - you have to trade the opportunity to earn money (and hence future earnings) for the time needed to raise children. Think of the Mazda as your latest infant.

[Just visiting from Macleans]

I had a 72 TR6 when in university. You had to become your own mechanic to keep it on the road (the service/parts costs were outrageous- well for a university student - and those British Lucas electrical systems...).

So, I became a bit of a mechanic. And here's why I find it useful today - many vehicles are just big black boxes to most owners - just add gas and turn on the ignition. That's all you need to know. Got a problem- drop in at Canadian Tire - just like every other auto service centre - they're honest, efficient, and only repair what is required, right? "Yes madam, the repairs amounted to $1800 because the flaggle came loose of the electronic wigwam which caused the carburmuffler to become clogged and resulted in frying of the thermistplug injector - all had to be replaced".

True story - I once took my vehicle in to a Canadian Tire to have some of the ignition system replaced - distributor cap and rotor. So, the guy looks in his standard labour rates - 0.5 hr for distributor cap, 0.5 hr for rotor replacement (or something similar) x $75/hr = cost (if it takes less, you don't pay less). So, I asked him - "why does it take 0.5 hr to replace the rotor when you already have the distributor cap off? Its just a matter of pulling the old one off, and pushing the new one in place?" "Well, that's how we prepare estimates and charge for repairs - standard practice"

So, I purchased the parts, went out into the parking lot, popped the hood and did it myself in about 15 min.

I often hear of ridiculous fees being paid for simple or unnecessary repairs - part of the industry.

[Frances Woolley]

Nick, after attempting to install a 2010 bicycle rack on a 1970s bicycle and failing (I would need a metal file, or would need to re-arrange the entire gear shift mechanism to give me the extra millimetre I need to get the parts to fit) *and* attempting to put handlebar tape on said bicycle and failing I feel anything but zen! We live in the dark ages and I have no knowledge of how to fix anything.

[GA]

Excellent piece. While the enjoyment part of it may be the dominant one (in the case of your car/your skill set), I'd add (or perhaps this is just elaboration):
1) Transport costs can be broadened to include search costs/transaction costs. Specifically, the task of finding the 'right' mechanic takes time. I am using the term 'the right' mechanic here to mean appropriate to the task - in terms of complexity, trustworthiness, quality of work, specific skill set _for the task at hand_. This scales according to the complexity of the task - as an example, think about oil changes, which are now (generally) commoditised to such a degree that even if you have the skill set, it's not worth the time - no specific skills required and easily comparable. Most other types of car work are not easily comparable from one mechanic to another.
2) Incentives: the reason for the parts having a 'fixed' cost or repairs being charged at fixed book rates is largely due to incentives; mechanics generally have no incentive to find you a cheaper part or a used part that will do the job - they charge a fixed margin on parts. Yes, they will consider after-market parts to keep your trade, but usually only for standardised parts that they can obtain with little or no effort. But I bet if you found the right part on the internet and paid them to put it in most mechanics would agree, although they might decline to warranty the end result (see my point below). (Mechanics that will go to some effort to find you used parts, for example, are a specialised market, and hence higher search costs...)
3) Trust/risk/insurance effects: another reason for parts and the 'standard time sheet' approach is trust and insurance effects. Consumers distrust and want a fixed estimate up front, even if it costs them more (which it does, because it includes the cost of making the free estimates). They want their risk reduced (within certain bounds, of course - Crappy Tire has the out to say they discovered your gaskets are blown, but they can't say 'it turned out to be more complicated than we expected, we had to remove the engine to get at the part and we weren't familiar with your Saab, and so we charged you an extra five hours of labour.') On the insurance side, consumers expect a quasi-guarantee (although they don't value this much, I suspect) if the repair is done badly; since this has an implicit cost for the mechanic, they do not want to use e.g. used parts that you provide.
4) Preventive maintenance/costs: probably the best reason to do the work yourself (if you have the skill) is avoiding future costs. Everyone wants the mechanic who will spend the time to check that, for example, the timing belt/chain needs replacing, and if it's not done now, it will take your engine out and cost you $8,000 two months from now. This is very, very valuable to the consumer. And it is a very rare mechanic that you can trust to say "please check everything, only charge me for stuff that you actually do, and provide me with a menu or list of stuff that needs done now or will need done in the next several months." Big shops like crappy tire charge very good money for this, although the consumer has no way of knowing if the work is actually done. It seems to me to be similar to the problem of restaurants in tourist locales: it is only worthwhile for the mechanic to provide this service properly if there is an ongoing relationship (or similar, like word of mouth/reputation effects).

My lessons from this (as someone who does not do his own work at all): finding a good mechanic and establishing a good relationship is very important and not so easy to do. If you do find one, refer all the business you can to that mechanic, and keep going back, and make sure all the business you refer mentions your name.

[Nick Rowe]

Luis: if we depart from perfect competition, it could go either way. Are more fun jobs more or less monopolistic? We might get jobs that are too much fun, in equilibrium.

Peter T: "1. Where is the evidence that wages are lower for jobs that are "more fun"? How do you measure it, ..."

Good question. Different people have different ideas about what is fun, so we can't simply regress wages against "fun". Plus, people whose labour is in less demand face a worse choice set, and will generally choose jobs that are both lower paid *and* less fun. Again, a simple regression won't work to reveal the negative correlation between fun and wages, other things equal, because they aren't. The best practical solution to these problems I have seen is to look at the wage premium in jobs that have a high risk of death. This works because: few people want to die; if the risk of death is big enough, this effect will swamp all the other things that should be held constant but can't be in practice. IIRC, there does exist a big measured wage premium in jobs that have a high risk of death.

JVFM: Garages also face fixed costs of booking in a client, getting the car into the shop, etc. So very small jobs can get very pricey.

It's difficult to know to what extent garages do unnecessary work. I think GA's comment gives a great description of the problems mechanics face.

"My lessons from this (as someone who does not do his own work at all): finding a good mechanic and establishing a good relationship is very important and not so easy to do. If you do find one, refer all the business you can to that mechanic, and keep going back, and make sure all the business you refer mentions your name."

That is exactly what I used to do. Long-term reputation solves, or at least alleviates, many of these problems. Choose a mechanic, let him know he will get all your business, provided he's straight with you. And let him know you will trust him to do what's best, and not haggle him over costs, if he finds something that needs doing. My garage man retired, unfortunately. (But don't recommend him to too many of your friends, or he will be too busy to fix your car.

And now I'm at the cottage, 600kms from home, and my new baby has sprung a coolant leak! I'm hoping it's not water pump, because I don't really have enough tools on me to replace it here (V6, hidden under timing belt cover, means removing an engine mount). Crossing fingers it's just the radiator.

[GA]

I wrote: "but they can't say 'it turned out to be more complicated than we expected, we had to remove the engine to get at the part and we weren't familiar with your Saab, and so we charged you an extra five hours of labour.'"

I should have been more clear - my point here was that consumers do indeed pay this cost. Sometimes the shop screws up, the guy who knows Saabs is out that day, the first time Mechanic A does Job B it takes him twice as long, nobody in the shop really is actually very good at doing A/C system repairs, they get the car up on the stand and dismantled when they realise it's a slightly different part that they can't get for a day (and limited bays available), etc. So the 'standard rates' build in extra time, which I'd guess is probably twice the actual time it takes for someone who is good at that job on that car to do. But they can't manage infinite specialisation well.

People with standard jobs on standard cars with bog-standard parts available overpay.

And occasionally some entrepreneur recognises that this means EVERYONE is massively overpaying for e.g. oil changes, minor windshield repairs, and comes up with a commoditised offering that undercuts everyone (until they all are forced to match prices).

I think that sometimes Crappy Tire and others like them will also just refuse to work on old/weird cars for anything but simple stuff; the pricing model doesn't work and it hurts the overall business. (Or they just charge so much you have to take your old Alfa Romeo to the local Alfa guy, if there is one, because otherwise it just costs too much - same difference).

[Luis Enrique]

Nick,

I like that possibility - too much fun making it, not enough consuming it.

How do you think things actually have turned out? Do you think that real developed economies resemble a model in which welfare optimizing (more fun) modes of production aren't chosen, are not an equilibrium, because of some game-theory style problem, or maybe just wage inflexibility? If we had a magic co-ordination enforcer, could we have an economy in which people enjoy their jobs a lot more, but we're all a little poorer? Maybe there's even some behavioral story, and for some reason people don't take lower paid jobs that would actually make them happier (hence the job market hasn't supplied them).

Would you take seriously the possibility that the outcome in rich capitalist free-marketish economies has too many miserable jobs (and too high output) and we could be significantly better off with much less boring but somewhat less productive production processes?

[Frances Woolley]

This article has something on the economics of spare parts (the link will work if you're on-campus).

Accidents, Liability Obligations and Monopolized Markets for Spare Parts

Pio Baake, DIW Berlin

Abstract
We analyze the effects of accidents and liability obligations on the incentives of car manufacturers to monopolize the markets for their spare parts. We show that monopolized markets for spare parts lead to inefficiently high prices for spare parts. Furthermore, monopolization induces the manufacturers to choose inefficiently high qualities. The key for these results is the observation that high prices for spare parts entail a negative external effect inasmuch as liability obligations imply that consumers of competing products have to pay the high prices as well.

[Just visiting from Macleans]

Garages also face fixed costs of booking in a client, getting the car into the shop, etc. So very small jobs can get very pricey.

I'd expect fixed costs to be covered in the mechanic's hourly rates. Jacking up rates and time req'd is double dipping, in my accounting books.

Just a note about CanTire - all stores are franchises, and the auto repair centre is separate again - the store owner benefits from rent (or percentage of sales) as well as on parts (and probably the garage gets percentage of parts as well). There is an economic incentive to replace parts.

[Helen Wheels]

Hey, just visiting: 2 points from the other side of the counter

"...if it takes less [time], you don't pay less"

Conversely, if it takes MORE time, you don't pay more. Really. That's what's called "eating the time" and chalk it up to experience. Flat rate times are for average skill/average experience. You get good at a task, why would you penalize someone for beating the time when they are generally paid per labour hour?

"...So, I asked him - "why does it take 0.5 hr to replace the rotor when you already have the distributor cap off? Its just a matter of pulling the old one off, and pushing the new one in place?" "Well, that's how we prepare estimates and charge for repairs - standard practice"

It's standard practice for mechanical repairs but the insurance-influenced collision side is another story. They very carefully monitor what they are paying for and "suggest" ways to include times or deduct time where one process makes another one easier (like if you are painting adjacent body panels, some factor is subtracted from each individual time to account for the fact you've already brought the car in, loaded the gun, masked off other parts, etc.).

If mechanical repair insurance ever took off, I guarantee your cap time would be included in your rotor time.

And Nick...parts can be expensive because the OEM has a monopoly, but also because they have to amortize design costs; plus the aftermarket stuff isn't always the best quality:
http://abrn.search-autoparts.com/abrn/Training/Its-time-to-revisit-the-OEM-vs-aftermarket-parts-d/ArticleStandard/Article/detail/666794?contextCategoryId=501

[Just visiting from Macleans]

Conversely, if it takes MORE time, you don't pay more. Really. That's what's called "eating the time" and chalk it up to experience.

I think you're confusing two issues. If I go into my CanTire, they have a published set of rates for standard work - oil change filter, say $30. Rotate tires - $20. Balance tire - $12 per tire etc. Same for all vehicles. So, yes, the garage/mech will go under on some cars, over on others.

When they use the Chilton book for labour for non standard repairs (or is it Michelin - something like that) recognize that it is put together by industry as a guesstimate/guideline. The fact that it doesn't (or allegedly doesn't) have a separate labour entry for a very common routine item of a combined replace distributor/replace rotor (say 0.6 hrs instead of 0.5 + 0.5), and that the service manager was unwilling to adjust the time accordingly once an informed customer pointed out the discrepency suggests to me they are gouging - and this is a normal practice for them. So, stay clear of that particular service centre.

A neighbour across the street just paid $1800 (the number I originally quoted) to fix a rough idling/hesistant Mazda SUV whose engine light came on- so after the fact when they described the symptoms - I told them immediately what I thought the problem was - dirty /fouled mass air flow meter (which controls the air/fuel mixture). I knew because I once had a Mazda 626 - same problem - and found the fix on a Mazda forum - similar to what Nick pointed out. I had to purchase a set of special screwdrivers (star shape with a dimple in the middle - the manufacturers do this trick so the average person can't service their own vehicles) take out the MAF meter, and spray the wire filament with an aerosol brake cleaner. Put it back in - the engine light went out and the car ran perfectly. At Can Tire - they replaced her MAF as well as a few O2 censors (not necessary) - My repair costed maybe $20 total (including the screwdrivers). She got the bad news /estimate when the car was on the hoist after being w/o it for a day or more.

The day after she got her car back, the engine light came on again- so I instructed her to go down to the local Parts Source store where they will plug in the engine diagnosis tool and report the problem code for free as a service to sell parts- which she did. It was a bad thermostat - $10 plus $10 gasket. An expensive lesson.