Too much stuff: the deadweight loss from overconsumption

In a simple supply and demand world there are two sources of waste.

Underconsumption is the first. It occurs when taxes, price ceilings, price floors or other interventions prevent the economy from reaching the free market equilibrium. Consumers would like to buy, suppliers would like to sell, but they cannot make a mutually beneficial exchange because there has been interference with the market.

The loss of consumer and producer surplus associated with departures from the competitive market equilibrium is called a deadweight loss. It measures the inefficiency, the loss of well-being, the waste associated with market distortions. The deadweight loss created by underconsumption is described in Wikipedia, and just about every ECON 1000 text ever written. If students know one thing when they finish first year economics, it's that a deadweight loss is that little triangle under the demand curve.

 

 

Yet there is another kind of deadweight loss, the waste associated with over consumption. Take parsley, for example.

I use a sprig of parsley here and a sprig of parsley there to add flavour and nutrition to salads, soups and stews - say two ounces per week. But the local supermarket doesn't sell two ounces of parsley. A consumer's only options are a large bundle, weighing about half a pound, or no parsley at all. The same is true of cilantro and other herbs. True, if parsley is treated like a delicate cut flower, put in a jar of water, and kept in the fridge, it will last a fairly long time, but inevitably half of any bunch of parsley I buy ends up as waste, in the compost heap.

Graphically, the situation is shown in the diagram below. When drawing the diagram, I've assumed that the supply curve, the price charged to the consumer, is equal to the marginal cost of production. That's not a 100% realistic depiction of the parsley market, but it makes life easier.

Given the marginal cost of parsley, the consumer would like to be at the point "just right", where the consumer surplus is maximized. However the consumer is only given a choice betwen "too little" and "too much." At the point "too little", the consumer surplus is zero. At the point "too much", the consumer surplus is equal to the difference between the blue triangle (the gains from consuming up to the point "just right") and the red triangle - the loss from the production of parsley that generates fewer benefits to the consumer than it costs to produce.

Overconsumption is pervasive. It happens when there are negative externalities, and market participants do not take into account the full social cost of their production and consumption decisions. It occurs in insurance markets, where consumers do not pay the full marginal cost of the goods and services that they consume. Firms encourage over-consumption when the market structure allows them to charge a price above their marginal cost - for them, more sales means higher profits. Information failures - as in the health care market, where consumers do not know the true benefits of the services offered - facilitate excess consumption. Over consumption can result from product standards - mandatory airbags, for example - or ensue from consumers' attempts to signal social status, the consumption arms race.

The question is: why does the standard economics undergraduate curriculum focus so much on underconsumption dead weight loss, and so little on the waste associated with overconsumption?

Here are some questions on the deadweight loss associated with overconsumption. They would be suitable for an undergraduate economics class or graduate public policy class: Download Overconsumption_questions

Comments

[Bob Smith]

All good examples, although I don't think it's fair to say that the standard undergraduate curriculum ignores overconsumption deadweight loss - I seem to recall intro-micro having at least discussions of the (in)efficiency implications of negative externalities or subsidies.

The over-consumption of herbs due to package size, though, is a clever insight on overconsumption deadweight loss. I don't think I'd every thought of it that way, but when you point it out, it's obvious. It's a pratice that drives me nuts. In your typical grocery store herbs like basil, rosemary or thyme comes in a 40 gram container, which is probably three or four times what you actually need (unless you're making pesto).

That said, at least in competitive markets, there are offsetting pressures which mitigate against the packaging deadweight loss - you'd think that selling more efficient package sizes would be a profitable strategy. As an example, I recently found a local regional grocer (Longos, for anyone in the Toronto area) that sells 10 gram packages of herbs for half the price of their 40 gram packages. Apart from better suiting the consumer demand, it's probably a profitable decision for Longos since (a) they may have lower costs (since they can generate the same revenue by selling half as many herbs - though granted, packaging costs might be higher) and (b) consumers may consume more fresh herbs if they come in more convenient packages (rather than substituting less profitable dried herbs for more expensive large packages). Still, competition doesn't always work - witness the enduring (and infamous) hotdog/bun mismatch.

[Phil]

There is a free market for parsley, is there not? Is this a market failure? Or maybe it costs just as much to sell three sprigs of parsley as 100? Or maybe it costs only marginally less to sell three sprigs, but consumers won't pay the price because they feel they're getting ripped off.

So, market failure, or buyer irrationality?

[Jacques René Giguère]

Producing parsley is cheap. Packaging, transporting and sorting the package is expensive. You,re not buyibg parlsey, you're buying the whole process. It's probably cheaper to grow, package and deliver twice as much parley.
Fred Smith became the billionaire owner of Fedex when he discoveres that it was cheaper to ship one large plane from Boston to Nashville and another to Washington than having a small one going the direct route.

[Shangwen]

Before this turns into a discussion of parsley biscuits (which probably taste fine), it's worth noting that both types of preconditions can exist in the same market. In healthcare you have both underconsumption (price controls, labor supply restrictions, etc.) and overconsumption (insurance, information failure, political intervention, etc.)

A good overview of both problems co-existing can be seen here. It is always interesting to see who defaults to one failure rather than the other as the main problem that we need to solve.

[Determinant]

There is a strong trend in economics of all sorts to see economics as the allocation of resources under conditions of scarcity, hence prices. That reasoning excludes deadweight losses from overconsumption outright because by definition overconsumtion does not occur. But that's not the real world.

[Frances Woolley]

Jacques Rene - interestingly that was exactly the explanation Nick gave when I talked to him about it. This makes the DWL calculation more complex, but doesn't eliminate the DWL, unless MC=0.

[whitfit]

Isn't this a problem with the way that the standard S&D graph works in such a context? The problem is that the quantity on the demand curve is really "bundles of parsley". Is there overconsumption of bundles of parsley? I don't think that is the issue with bundling. Somehow this has to take into account bundling, which I don't think is well described by the standard graph.

The bundling issue is probably well considered in the context of understanding the phenomena of Costco and its ilk, where they sell large bundles of things, with the unit cost of the content being much lower. Is that because the cost of the actual product is low in comparison to the shipping/sorting and the whole process as noted by Jacques René Giguère?

In terms of losses, I was also thinking about the disposal/cooling costs. You have to keep it cool in your fridge, and then unless you have your own compost, you are putting it in municipal compost that isn't cheap to run. I wonder how much the overconsumption of food because of issues like the bundling of parsley costs rate/taxpayers?

[Frances Woolley]

Bob: I agree with you that the Toronto grocery market is *much* more competitive than the Ottawa one - lower prices on a whole range of goods - so I'm not surprised you have a more competitive market for bundles of herbs, and more ranges of herb bundles also. (Irrelevant tangent: for rosemary and thyme bundles I put the extra in a paper bag in the refrigerator - after a couple of weeks you end up with a nice bundle of dried herbs with much better flavour than regular dried herbs. Basil I pour olive oil over the extra and let it infuse.)

"I don't think it's fair to say that the standard undergraduate curriculum ignores overconsumption deadweight loss"

Over the years I've taught hundreds of students whose economic background consists of ECON 1000 and not much else. Invariably they find the idea that a DWL could be anything other than an area under a demand curve extraordinarily difficult to grasp - sure, it's there in the curriculum, but not much emphasized.

Phil: "So, market failure, or buyer irrationality?"

Both. As Bob points out, if the market is competitive, the bundles offered will be the ones that provide maximum consumer surplus, while covering the fixed production/bundling costs. So the huge expensive parsley bundles in Ottawa grocery stories are probably a result, in part, of the lack of competition in the grocery market here - i.e., a market failure.

At the same time, I think buyer irrationality plays a role. Take, for example, another overconsumption irritant: excessively long textbooks. Why, students ask, can't the publishers make the book half as long and cut the price in half? The reason is that, if the publishers did that, they wouldn't cover their fixed costs publishing/distribution/marketing/royalty to author costs.

Now since the students only use half the book in any event, they would be better off with a shorter textbook, even if the cost was only reduced slightly. But somehow the typical reader figures that a 300 page book (with 150 pages of useless material) for $200 is a better value than a 150 page book (with zero pages of useless material) at a price of $150.

[Linda]

Frances: seems to me you have left out one substitute here. Someone who uses parsley in small amounts one a regular basis could/would buy a plant for the kitchen windowsill, or the garden.. Buying packaged or bunched parsley incorporates a price premium for convenience. As for the quantity in a package - I'd opt for the cost explanation others suggested: cost of packaging /bunches of parsley is not linear in quantity of parsley in a bunch, and those costs outweigh the cost of the herb itself. Now, why people are not willing to pay almost as much for half the quantity, say, is another issue. Certainly it is possible with celery - buy by the stalk, or a whole...head? What is the correct term?

[Frances Woolley]

Linda -

Now I feel guilty about not having a pot of parsley on my windowsill!

Sure, bunched parsley *could* be a second-best market outcome, a way of dealing with high fixed/low marginal costs. But it doesn't *have* to be - in the absence of effective competition in the grocery store market, and prices below marginal cost, retailers have every incentive to increase the size of the bundles and ratchet up prices, in an attempt to extract as much of the surplus from consumers as possible.

My point is simply that all of those other forms of deadweight loss - all of over-consumption-type efficiencies, seem to get a lot less attention than the under-consumption ones - at least judging from what my students know/don't know.

[Jacques René Giguère]

Frances: this blog is where great minds congrgate...
MC of growing a bunch of parsley is essentially 0.
So is the MC of shipping one more parcel on a big plane. The Samuelson core is empty. What Fred Smith understood is that the only way to overcome it is to fill the plane and then ask for the reservation price, which you can get by promising thst it would be absolutely, definitivel, positively there overnight.
Otherwise, you are confrontrd to the same problems trains and ships encoutered.
In the immortal words of Warren Buffet: If a capitalist had been at KitttyHawk the day the Wright brothers did their first flight, he should have shot the two ba!@#rds...

[Frances Woolley]

Jacques Rene - if the marginal cost of growing/shipping/transporting parsley was truly zero, then the optimal strategy in terms of maximizing social welfare would be to package parsley in the biggest bundle anyone could ever conceivably want, sell it at a price that was at least enough to cover firms' fixed costs. How the potential surplus - the area under the demand curve - is divided between producer and consumer is then just a question of the degree of competition in the market.

But the marginal cost, though low, isn't zero. Think of all of those fields used to grow parsley that no one eats, that could be put to better use. As for the bundle size - I'd really like to know what percentage of parsley buyers ever get through a bundle of curly parsley. Unless you make tabouli, I'd say it's pretty difficult for the average buyer.

[Paul]

Frances, I think Jacques Rene is right, perhaps we should look at this problem using our individual MC? I also discovered that growing my own cilantro, mint, sorrel and chives is far better than buying bunches at Farmboy, and then throwing most of it out after I use what I need. I already garden, so I have tonnes of pots and soil, and a packet of seeds is about a dollar or two, and will last me potentially forever. For the cost of one or two individual bunches of herbs, I can produce all the variety of herbs I need. I only harvest as needed, and I can also harvest the seeds as the plants bolt, and start the process all over again. The marginal cost of growing/shipping/transporting all my herbs is essentially zero...certainly it costs me no more than $5, maybe $6? And I have fresh herbs all year around? No other producer can beat that, especially sorrel, which few people in Canada seems to know how to prepare, and is not offered in bunches at any price (afaik).

"My point is simply that all of those other forms of deadweight loss - all of over-consumption-type efficiencies, seem to get a lot less attention than the under-consumption ones - at least judging from what my students know/don't know." - Perhaps because these situations are minor compared to the under-consumption? Even the airbag example, I can't see this being that expensive for producers to buy and install at the point of manufacture. Certainly in the aftermarket, they are expensive to replace, but not on the assembly line.

Your healthcare example is a different beast entirely. How do you convince a terminally ill person to stop expensive treatments when they are willing to pay any price to live?

[Bob Smith]

"How do you convince a terminally ill person to stop expensive treatments when they are willing to pay any price to live?"

I think Frances' point is that we don't know that the terminally ill person is "willing to pay any price to live", because they typically don't pay any price at all (i.e., they're receiving insured services). If a terminally-ill person actually bore the cost of their care, at some point they likely would say "this isn't worth it" (either because they run out of money, or because they think the cost is disproportionate to the possible health benefits - this is a decision that terminally ill patients regularly make when they decide that the non-monetary costs of future care, notably pain and suffering, make futher treatment undesirable). And if the terminally-ill person is willing to pay for the expensive treatment, than their consumption isn't "over-consumption", it is optimal consumption.

The tricky policy issue is that, if you don't want to deter over-consumption of health care services by pricing, what other mechanisms can you use. In practice, of course, we do use pricing to deter over-consumption of health care services, in that our governments say "we'll cover these services, but not those. Those services you have to pay for yourself". Needless to say, those decisions are unpopular with the persons affected, precisely because they are unwilling (or unable) to pay for "those" services.

[Andrew F]

Bundling extends to those features that sound interesting but you seldom/never use on items like cars, cameras, computers, etc.

[Frances Woolley]

Paul - your important point is this: "Perhaps because these situations are minor compared to the under-consumption?"

On this planet some of us are consuming impossibly, unsustainably, huge amounts of resources. One response is to say "yup, that's efficient, and when there are no more resources left, prices will go up, people will consume less, and that will be efficient too."

I look around me and I see people constantly being urged to buy. I see everything from mouthwash to housing being marketed and designed in ways that encourage people to buy and use more than they really need. Robert Frank and all that.

If right now our economy is experiencing a lot of under-consumption type inefficiencies, I'd had to see how much resources we'd be consuming when we reach an efficient market equilibrium!

On the health care example: it's complicated. See my post on lessons from the animal hospital.

[Bob Smith]

"the airbag example, I can't see this being that expensive for producers to buy and install at the point of manufacture. Certainly in the aftermarket, they are expensive to replace, but not on the assembly line."

Perhaps, but that only goes to the magnitude of the deadweight loss, not whether or not it exists (although I think airbag units themselves are actually surprisingly expensive). I'm not sure that the airbag example is neccesarily an example of "over-consumption". I would have thought that the deadweight loss in that case would come about because the higher cost of cars arising from the airbag requirement would cause people to buy fewer cars than they otherwise would have. In that respect, the airbag examples is more akin to the classic "under-consumption" deadweight loss arising from taxes (assuming, of course, that consumers don't value airbags. There's also the argument that by reducing the social cost of accidents, through reduced health care cost or lost productivity, airbags are socially efficient, but that's ultimately an empirical question)

[Frances Woolley]

On overconsumption of gadgets, see this prank iphone 5 video that's making the rounds http://www.maximumpc.com/article/news/jimmy_kimmels_iphone_5_prank_hilariously_depressing

[J.V. Dubois]

Frances: I am with Nick and Jacques here. If marginal cost of producing smaller package is higher then the marginal cost producing the larger package, then society will be better off with "overconsumption". Obviously I agree with externalities explanation of overconsumption, but that is a completely different point. It could be possible that the parsley-bundle producer could grow less parsley and more carrot, but it is just a pecuniary externality for me. I assume that he already made this calculation and found out that it does not pay off. They would probably lose more by creating smaller bundles of parsley and using resources freed in this way to grow and sell carrot then to just sell larger bundles of carrots.

Anyways, this reminds me of Nick's older article about joint products. If you want to consume just a lemon peel but you can only buy the whole lemon, does it mean that you are "overconsuming" just because you did not offer the rest of the lemon to all your neighbors who may like to have lemonade from the fresh lemon juice? If economist observes that this behaviour does not occur, she may just assume that marginal costs of walking over to your neighbor are higher then gains from doing so. Therefore there is no overconsumption. Maybe there is just an information assymetry - maybe you do not know that just now your neighbor craves for lemonade but not as much so that he will take a walk to the nearest grocery store - but he would gladly pay you for your lemon juice.

[Frances Woolley]

JV Dubois: Parsley is being tossed away. That's waste if the MC of parsley of growing/transporting/cooling/disposing of parsley is greater than zero.

Now it may be that it is necessary to waste parsley because there are insuperable obstacles to delivering parsley in a range of appropriately sized bundles, and the costs of wasting parsley are less than the costs of adjusting the bundle size. But, seriously, given modern technology and the mass scale of modern agriculture, you're telling me this can't be worked out?

I think that a much more plausible explanation is the absence of effective competition in the supermarket/grocery retail industry in much of Canada outside of Toronto/Montreal/Vancouver. Calgary is closer to Kelowna than Vancouver is, but Okanagan produce costs far more in Calgary than in Vancouver - the result of much more competition in one market than the other. Absent competition, there is less pressure for supermarkets to tailor products to consumers' tastes, so they produce and sell large overpriced parsley bundles in an attempt to extract rents from the consumer.

It's interesting that, when it comes to cilantro, this is much less of an issue, because Chinese and other ethnic supermarkets sell low-priced smaller bundles - competition works. The flat parsley market is also different from the curly parsley market, because of the Lebanese and other middle eastern grocery stores.

[J.V. Dubois]

Frances: "But, seriously, given modern technology and the mass scale of modern agriculture, you're telling me this can't be worked out?" It could be possible that this cannot be worked out effectively - maybe not in this particular example with parsley in your local grocery store, but certainly in other areas. But maybe I just did not express my thought good enough. What was trying to say is that in many ways the things we buy and consume are "discrete". This is very well a case for most consumer durables. If you buy a chess set but you extract only one ocassional chess game every once and then, does it mean that you are "over-consuming" chess sets compared to someone who is able to extract hours of fun from it in form of "chess set services" every day? Isn't there a technology that would enable occasional chess players sharing one chess set and redirect resources saved this way to more productive areas? Can we say that producers of chess sets deliberately prevent such an innovation so they may keep their sales high and extract inappropriate rents?

Let me use some other example, for instance the ammount of meal in restaurants for medium clientele. The cost of the meal consists of ingredients, services of cooks and waiters, rents for restaurant and amortization of cooking equipment and furniture. I observed that in developed world where the cost of services is higher compared to the cost of ingredients, people do not make much of a fuss about size of the meal. As an restaurant owner you can order large fillets or smaller ones, the biggest issue for you is the time it takes for the cook to turn them over and for the waiter to bring the plate to the customer. In poor countries it works exactly the opposite way - with services being very cheap compared to costs of raw materials, they are much more strict about how large a portion you are entitled to for your money. Is it an insurmountable technological obstacle for restaurant to keep prepare different sizes of meal or add few pages into the menu or to devise a more intricate pricing system? Probably not. But they also probably think it is not worth it. They are just better off throwing out whatever some customers did not manage to eat.

[Frances Woolley]

J.V. Dubois: "Can we say that producers of chess sets deliberately prevent such an innovation so they may keep their sales high and extract inappropriate rents?"

As Shangwen pointed out earlier, one's willingness to accept such a proposition is a pretty good litmus test of one's views on how the economy functions and the efficiency of competitive markets.

[Bob Smith]

I think we're talking past one another. There may well be cases where a single bundle is optimal, given the costs of trying to sell different bundles, notwithstanding that some people would prefer (ignoring the transaction costs) a smaller bundle. If you see that phenomenon in a competitive market, I think you're right that is the efficient outcome.

But I think Frances point is that in a non-competitive market, retailers might prefer to sell a sub-optimally large bundle to increase sales. France's observation about herbs is interesting because, as I noted, there's at least one regional retailer who manages to sell small bundles of herbs, and as she points out, other ethnic retailers may do the same thing. That suggests that that it isn't a case where small bundles can't be done in a cost-effective manner.

It could be the case that different herb retailers are catering to different customers which might explain the different pricing practices of different stores. For example, consumers at Chinese or other ethnic supermarkets may have different demand curves than consumers at, say, Superstore (in fact, they almost certainly do - it's reflected in the selection, quality and price of the produce and seafood at a T&T grocery store (one of Canada's largest Asian grocery stores) compared to traditional stores). I'd be willing to bet that your average Longos or the "ethnic" retailers that Frances observes sell a lot more herbs than the average Superstore. Maybe for them, it's cost effective to sell two difference sized herb packages.

The analogy with a restaurant is interesting, because we're all familiar with restaurants that DO offer different portion sizes - think McDonalds, at the low end of the spectrum, maybe the Keg higher-up the spectrum. What they have in common is that they sell so much of their product, that stocking different portion sizes isn't an issue (at McDonalds you can get hamburger, a quarterpounded, or a third-pounder burder, at the Keg you can get an 8-oz steak or a 12-oz steak). On the other hand, if I got to the little family place around the corner, they'll have a burger and a steak on the menu, but it's one size fits all. In part the difference reflects the reality that they're serving different markets.

[J.V. Dubois]

Bob I agree, but then we are probably talking about the realities of the world that is governed by monopolistic competition. If even with relatively homogeneous product such as parsley we have various suppliers that are able to extract rents by catering to different market segments then competitive markets in groceries will exist only in a fantasy world. Or maby what grocery stores really sell is not only "parsley" or "bread". They are offering a joint product called "shopping experience" and some people end up sticking with particular sub-options that are not ideal for them. The same is valid for many other products, like let's say PC processor suppliers. You may choose one bundle named "AMD" and other bundle named "Intel" even if ideal product for you would be a mixture of the two. Bad luck, you have to choose the next best alternative.

[Frances Woolley]

JV, Bob - you both make excellent points. It's also the case that, if firms were really good at profit maximizing, they would be choosing the optimal bundle size. But in a non-competitive industry, firms can easily go along without innovating, so as long as they're making a comfortable level of profits. History and inertia are powerful forces. E.g., why are eggs sold in dozens or, sometimes half dozens? Because they've been sold that way for decades, if not centuries. Is a dozen eggs still an optimal bundle? Probably not, given shrinking household sizes, but it's not so far off that some innovator can make instant profits by selling eggs in a per-unit basis (though, yes, there are various alternatives, e.g. those packages of pre-mixed eggs that you keep in the fridge and use as needed).

So it is definitely a case of market failures (i.e. lazy firms, irrational consumers) + market power here.

[Bob Smith]

Fair comment, although once we're talking about monopolistic competition, then we're getting into Frances's world of imperfect competition and market structure allowing producers to induce over-consumption. You're point about bundling shopping is a good one, parsley probably isn't a show stopper for most consumers. I'm not sure the AMD/Intel comparison is all that compelling, though, since the PC processor market is a duopoly with Intel having a dominant position.

[Michael Hollander]

Not just health care but education. Take primary education. I would love to send my anxious kid (he's in K) somewhere that had nice and conscientious teachers, less noise, and a teacher to student ratio below, say, 15. academics can wait. this would probably be a premium of say 30% over the normal education cost. if i paid for that myself that would probably be 10k per year. so say 15k per year. but no, if i am going to do private school in new york, i have to pay for a lot of stuff i don't value, and the yearly total bill is going to be something like 30k per year.

i'm sure you can come up with other examples in education of people having to buy more than they will find useful just to get a credential. you have to subscribe to a whole institution, and play by all their rules, no choosing from a menu.

[Frances Woolley]

Michael - ouch! The sunshine hurts when it's shining directly in my face.

[Alex]

Another sample is car servicing. I take my car to be serviced, they ring back and say that the spurge-wangler is nearly worn out and needs replacing. Is this really the case? I wouldn't know a spurge-wangler if it bit me in the bum. (Maybe it doesn't even exist). For all I know, the old one has years of life left. I can either let them replace it, or get a second opinion at another garage, which is a lo of hassle. At any rate, certainly lots of car parts are replaced unecessarily.

[Steve]

Frances: Slight nit-pick on the Kelowna to Calgary/Vancouver distances. Google has Kelowna to Calgary as 609km and Kelowna to Vancouver as 390.

[Frances Woolley]

Steve - Urgh. Should have fact-checked.