The Shopping Cart Puzzle, or Intermediate Micro takes on Behavioural Economics

Today's typical grocery store shopping cart is much larger today the shopping carts of yesteryear. The question is: why?

A behavioural economist would observe that people buy more when shopping carts are larger. For example, an AER article by Wansink, Just and Payne notes:

....consumption can also be unknowingly influenced by environmental cues—benchmarks or reference points—that may subtly suggest a consumption norm that is appropriate, typical, reasonable, and normal. ....large packages, plates, serving bowls, and even pantries have all been shown to increase how much a person serves and consumes by 15 to 45 percent

If grocery stores figure that larger shopping carts translate into higher sales volumes, one might expect them to increase cart sizes. 

Eaton, Eaton and Allen take issue with this explanation in the 2009 edition of their Microeconomics textbook:

If you're thinking that store owners just increased the size of carts to fool customers into making larger purchases, then you are off track. Customers might get fooled once, but eventually they would learn that larger carts are bad things, and would only visit stores with small carts. As a budding economist, you want to avoid lines of reasoning that suggest people habitually do things that make them worse off... (p. 64)

The problem with this argument is that it flies in the face of the abundant empirical evidence that people habitually overeat, overspend, and do other things that make them worse off.  

Yet there is a good point here: an economist should start with the premise that people aren't stupid. So is there an alternative explanation? Eaton, Eaton and Allen say there is:

One of the major changes to have taken place in the past 50 years has been the continued entrance of women into the workforce. When women were more likely to be at home they had more time on their hands and had schedules that were flexible. Multiple trips to a grocery store had the benefits of fresher produce and reduced required storage space at home in terms of pantries and freezers. However, the cost was that it involved large amounts of time. As women continued to work outside the home, their time became more scarce and a substitution took place: fewer but larger trips to the local grocery stores. Larger shopping carts are not a marketing tool of exploitation, but just another example of substitution taking place. 

There is one sensible observation here: today, people buy in bulk, buying more when they visit the grocery store. The question is: why?

If the trend towards larger trips to the grocery store was caused by increased female labour force participation, I would expect to see big stores and big shopping carts predominating in areas where female labour force participation is high, and more traditional, small shopping cart shopping experiences in areas where female labour force participation is low. I would expect to see relatively few at-home parent pushing mega-carts around mega-stores. That's simply not the case.

Why not? Families with one income, all else being equal, have less cash than families with two incomes, hence are relatively sensitive to price. Groceries often cost less when purchased in bulk. Mega-stores have lower costs due to economies of scale, low-rent locations, and monopsony power that allows them to extract price concessions from suppliers. Some of these cost savings are passsed onto consumers in the form of lower prices. Moreover, bulk sales facilitate price discrimination - a store can charge the last minute shopper $1.79 for the can of tomatoes he needs tonight, and charge the price sensitive planner $12 for 12 cans of tomatoes. The family with a full-time homemaker shops in bulk to take advantage of price discrimination that works in their favour. 

Large shopping trips also require a concentrated block of time. How much time does it take to go to Costco on a Saturday? Two hours? Three hours? The family with the at-home parent has time to take long shopping trips, and the flexibility to visit Costco when it's not insanely busy. The time-strapped jobs+kids parents are the ones shopping for bread for tomorrow's sandwiches at 9:45 at night at the local small-cart supermarket.

What about the argument that, 50 years ago, people shopped daily to buy fresh produce? That was then, this is now. People have larger houses and better refrigerators. Food storage and transportation technology has improved. Apples, potatoes, and onions are harvested in August and September, then stored and sold year round. What can't be grown locally and stored is trucked or shipped in from southern climes - tomatoes from Mexico, grapes from Chile, asparagus from Peru. Avocados, bananas and pears are sold while still hard, so the customer can ripen them at home. For most produce, it doesn't matter whether a person shops daily or weekly - regardless of their labour market status. 

Truly fresh vegetables are now a niche market served by specialist stores. Yes, they do have relatively small shopping carts. Interestingly, my local fruit-and-vegetable store, Farm Boy, has a large and extensive deli counter, selling ready-made meals. This suggests that their target audience includes - without necessarily being limited to - people who work outside the home.

So what has caused shopping carts to grow? My preferred explanation is simple: people have higher incomes now than they once did. When people have higher incomes, they buy more stuff, hence need larger carts to put it all in. Fifty years ago, Pampers had only just come on the market. Thrifty housewives used rags instead of paper towels, mops instead of Swiffers, and ground coffee instead of Tassimo pods. Yummy single-serving yoghurts and cheese strings had not been invented yet, and bottled water was this strange European thing. A well-stocked refrigerator had one type of mustard, not four; a well-stocked cupboard had one or two types of vinegar, not six (white, red wine, white wine, cider, malt and balsamic). With all of this extra stuff, people need bigger carts.

It could be argued that Eaton, Eaton and Allen are right about the causal impact of female labour force participation on shopping cart size, just wrong about the mechanism - women entering the workforce prompted an explosion of convenience foods and labour saving goods, and this has caused the growth in shopping carts. That's an old story, but one that has been called into question. For example, Greenwood, Seshadri and Yorukoglu, in their paper Engines of Liberation, argue that it is not women entering the labour force that caused the revolution in household technology. Instead, the development of washing machines, refrigerators, microwaves and other household appliances made it possible for women to work outside the home and still cope with the demands of family life. Mega-stores and mega-shopping carts are part of that technological change - but which way the causality runs is hard to determine.

There is just one last possibility to consider. The shopping cart was only invented in 1937, and took some time to be widely adopted. The shopping carts of 50 years ago were still relatively new models. It's entirely possible that they were simply badly designed, and too small for the average shopper even then. 

So here's the question: what's the best explanation for the growth of shopping carts:

1. Behavioural economics: people buy more when shopping carts are larger
2. Rising female labour force participation causing people to take fewer but larger shopping trips
3. Rising incomes mean people buy more stuff hence need larger carts
4. The original carts were badly designed in the first place

Update: in response to the  comments, here is an additional explanation:

5. Supermarkets are larger today than in the past, for reasons not captured in (1) to (4) above, such as the introduction of universal bar codes and scanners. Larger supermarkets require bigger carts. 

Comments

[Nick Rowe]

Slightly off-topic: "...and bottled water was this strange European thing."

As I remember it, bottled water meant "we are now in a less-civilised country where it isn't safe to drink the water out the tap".

If the "let's dupe them into buying more than they want" theory was correct, would supermarkets still make little baskets available for those who want them?

I wonder if the size of supermarkets, or width of aisles between rows, might be part of the explanation? Though I can't tell whether that is a cause or consequence of bigger shopping carts.

[Frances Woolley]

Nick: "If the "let's dupe them into buying more than they want" theory was correct, would supermarkets still make little baskets available for those who want them?"

Yes. The person who just wants to grab a can of black beans for tonight's dinner will go elsewhere if there is no little basket available. It's the person who is buying $150 worth of groceries, but can be nudged into buying $200 with a large cart, who is the target of the megacart.

Size of store is highly correlated with size of shopping cart - think e.g. of the local hardware store v. megastores like Home Depot or Rona. (The existence of super-mega-shopping carts - in suitably masculine colours - in Home Depot is another reason to be skeptical of the it's-all-due-to-women-working hypothesis). The emergence of mega-shopping stores has all sorts of causes too - the way cities are designed, the way zoning laws work, the cost of transportation....

[Keshav Srinivasan]

Here's an explanation Steve Landsburg proposes, in his intermediate economics textbook: "Large shopping carts, and the wide aisles that are necessary to accommodate them, have always been desirable luxury items. They are also expensive, because wide aisles mean that stores must occupy more land. As shoppers have become wealthier over the last few decades, they have become increasingly willing to pay higher prices in exchange for wider aisles and bigger carts."

[Greg]

How about, "There is more different stuff available at the grocery store now than in the past, so I can buy more useful stuff in one go."

[Ryan]

What about the rise of direct wholesale-to-consumer sales such as Costco, Sam's Club, etc.? You can even find wholesale "bulk" packages in standard grocery stores these days. You need larger carts to accommodate whole bulk purchases. This isn't quite square with the other options you've listed, so I'm throwing it out there as possibility #5.

[Frances Woolley]

Greg - "There is more different stuff available at the grocery store now than in the past, so I can buy more useful stuff in one go"

That's not quite #3, because 1970s grocery stores didn't sell clothes, appliances, Christmas tree decorations, and so on. It's more a variation of the mega-stores enjoy economies of scale, hence lower costs idea. Possibly combined with some amount of nudging people towards making impulse purchases.

Keshav: "large shopping carts...have always been desirable luxury items."

Since shopping carts weren't invented until 1937, I think "always" is stretching it a little bit. Plus wide aisles/large stores don't equal higher costs if the megastores are built on the urban fringe. Empirically I think you'd be hard pressed to find a positive correlation between aisle width and prices.

[Whitfit]

"If the "let's dupe them into buying more than they want" theory was correct, would supermarkets still make little baskets available for those who want them?"

But the baskets have become bigger too! And many of them have little plastic wheels so they are more like mini carts than carry baskets.

[BigEd]

A much simpler answer is that today's supermarkets are generally larger and carry many more lines of products than was the case 30 - 50 years ago. Today they are also drug stores, bakeries, flower shops, gift shops, butcher shops, etc. Most new Walmarts are clothing, grocery, hardware, appliance, sporting goods, etc., etc. stores. There are also a smaller total number of these super supermarkets than the former mom and pop corner grocery stores; so the average shopper must drive farther to stock up with more items per trip.

It is simple economics that a typical shopper will make fewer total shopping trips and buy more per trip than before. The larger carts accommodate this rational new behavior.

BTW: at the supermarket I just visited today I noticed large carts, small carts, as well as even smaller hand baskets.

[Livio Di Matteo]

Frances:
Ultimately, the resolution is an empirical question. You would need data on shopping cart size over time as well as data on household composition, income etc...

[Frances Woolley]

BigEd -

I agree with what you say, but it's not getting at the underlying *why*. Why are today's supermarkets larger? Is it because of increasing female labour force participation? Rising incomes that mean people demand and buy more stuff? Do companies erect megastores to dupe us into buying more, edged on by marketers and behavioural economists? Or is it because large stores on the urban fringes are simply a highly cost effective form of retail distribution, given the existing retail distribution technology/price and wage structure?

Livio - I wonder if it's possible to get that kind of data anywhere.

[Sandwichman]

It's really a perceived safety issue. People pushing smaller carts are afraid they will be seriously injured if rammed into by one of the monster carts. Also, it is an investment in the future. As homelessness creeps up the income scale and captures more hoarders, the shopping cart people of the future will require larger vehicles to haul their possessions around in. Eventually, the free market and the supermarket will merge in the utopian superfreemarket.

[Frances Woolley]

Sandwichman - excellent ;-)

[Shangwen]

Sandwichman has clearly figured it out. With the coming housing bust, the oversupply of storage pods, and a carbon tax, large shopping carts may turn out to be the unexpected killer app that solves the problems of high housing debt and soaring transportation costs.

Supermarkets and shopping carts have also become more common in East Asia, though they aren't as large (see here for hard evidence). In those countries, you might explain the same development as partly due to the removal of large numbers of women from the labor market for domestic help (more demand for convenience). My grandmother and all her friends in China did not shop; the maids did. Would shopping carts be larger in Japan than Thailand, when domestic helpful is unaffordable in the former but still readily available in the latter?

There is a good pictorial history here. I suspect that, for a thorough understanding, no analysis of the shopping cart is complete without a whole chapter devoted to the children's seat.

[Bob Smith]

I don't know what it means, but I've noticed that some stores are now offering smaller mini-carts (the size of the old portable baskets, but in buggy form). One explanation might be that stores are responding to an aging population which doesn't need a big cart, and which might have difficult dragging a big cart or full basket around the store.

[Frances Woolley]

Bob - well, portable baskets definitely discourage purchases (I use one whenever I'm going to be carrying my groceries home, so I don't buy more than I can carry.) So I'm not surprised they're being replaced by mini-carts.

On population aging and grocery shopping - I've recently spent a fair bit of time buying groceries in downtown Toronto, including the Loblaws in what was once Maple Leaf Gardens, which has lower prices than comparable Ottawa stores, and free parking. The condo boom and the surge in the number of people living downtown has led to a big increase in the number of grocery options. Now that's a really fascinating market.

But you're right - to the extent that stores are responding to population aging by introducing smaller carts, it suggests that the large carts aren't purely a tool of exploitation by evil marketers.

Shangwen, great link to the pictorial history! Always enjoying hearing about your grandparents...

[Leo]

Larger shopping carts relate to the big box phenomenon, which was caused by single use zoning, minimum parking requirements, and subsidies for car ownership via road construction and failure to charge property taxes per square foot of road.

Grocery stores in medium density areas got saddled with oversized grocery carts because of bulk grocery cart procurement across grocery store chains.

[Frances Woolley]

Leo - I'll put that as a vote for explanation #5 - stores grew for other reasons not mentioned above.

[Bob Smith]

But you're right - to the extent that stores are responding to population aging by introducing smaller carts, it suggests that the large carts aren't purely a tool of exploitation by evil marketers.

And I've also seen the basket-buggies in urban grocery stores, no doubt serving the condo crowd who (a) are on foot and (b) don't have room to store a gross of turkeys.

[Andrew F]

Also, computerization and more efficient supply chains, along with improved access to cars and better road infrastructure, enabled the large stores, which enabled SKU proliferation (offering more product impressions) and larger carts. Don't forget barcode technology that greatly increased speed of checkouts, making easily the worst part of a shopping trip (waiting in line behind people with enormous baskets) more tolerable.

The causes are potentially pretty nebulous.

[Andrew F]

Frances,

"Bob - well, portable baskets definitely discourage purchases (I use one whenever I'm going to be carrying my groceries home, so I don't buy more than I can carry.) So I'm not surprised they're being replaced by mini-carts."

Theft might be another reason. Customers are likelier to walk off with hand baskets. One store I shop at has antitheft devices attached to their baskets.

[Donald A. Coffin]

One comment, which tends to support the suggestion that grocery carts were inefficiently small. I worked in a supermarket in 1964, 1966, and 1968 (summer jobs). It was extremely common then for a family shopping on Saturday to arrive at a checkout with 2 carts. Now, I would guess, that is fairly rare (at least I don't see it all that often, even at Costo on a Saturday).

A factor working in the opposite direction, that I may have missed, though, is that average family sizes were larger 50 years ago than they are now.

[Frank Restly]

From my own experience I would agree with #5 - larger carts are an outgrowth of big box stores with a larger product selection and wider aisles.

[Determinant]

Or is it because large stores on the urban fringes are simply a highly cost effective form of retail distribution, given the existing retail distribution technology/price and wage structure?

Yes, combined with the greater market segmentation of grocery stores. I remember in the 1980's when my family shopped at A&P, everybody in the north end of town did, it was the local store though it was fairly big. Since the early 1990's each of the national chains (Metro, Loblaws and IGA) has a "hard discounter" in their franchise lineup, No Frills, Price Chopper, etc. My family was driving to the centre of town to the No Frills location by the mid 1990's.

The middle segment, the A&P's and Loblaws have started to sport hardware and drug stores where they formerly confined themselves to groceries.

The Canadian grocery industry had been deathly afraid of the entry of Walmart into the grocery segment for 15 years. In the US Walmart gutted the existing grocery players. That hasn't happened in Canada and won't. The Hard Discount franchises prevent that, Walmart can't undermine the national chains on price, the chains are already there. Walmart's grocery operation in Canada is an also-ran.

[Patrick]

Slightly OT: It seems Canadians waste an enormous amount of food. People really *are* habitually making themselves worse off.

http://www.cbc.ca/news/canada/story/2012/09/28/food-waste-report.html

[fresno dan]

Kudos to Francis Woolley for posting an interesting pictorial history of shopping carts (where in the world would you get something like that???)
Look at the early two basket shopping cart designs and offhand they certainly look like they can carry as much as current shopping carts. And the advertizements really ....dare I say "pushed" ;) - the idea of filling your shopping cart to the brim.
I think the two basket cart evolved into the one larger basket (still with the ability to put some items underneath the big main basket)as a matter of convience - not having to bend over.
As I age, I get the smaller shopping cart primarily due to the fact that I don't have to bend over so far to put things in and remove items from the the larger cart - in a smaller cart the basket is much higher up.

I think both econmic camps are partially right.
Oh, and 3 could be further elaborated upon - more stuff or BIGGER stuff? A box of pampers takes up a lot of room...the one serve Keurig coffee containers/packaging takes up considerably more room than you can get from a coffee can that gives an equivalent number of coffee servings.

[Jciconsult]

We are time and travel challenged. Time, even if not working, has become occupied with many things. It is easier to by 8 rolls of toilet paper and store them then to make multiple trips. In our two person household, even though we have a Metro in our building, we walk to a No Frills (good exercise, cheaper prices, better veggies - (more immigrants?)) with large sacks or a cart. Time, money and personal choices matter. You don't have to fill the cart.

[Phil Koop]

I think you make a persuasive case for #3 as an explanation for larger carts. But the explanation needs to be tempered because the relationship between expenditure and cart volume is not straightforward. People buy far more prepared food-in-a-box than they used to, and this is both more expensive and bulkier (because of packaging) than buying ingredients. But many expensive items are smaller than their older counterparts; my mom, shopping for three teen-aged swimmers, probably spent about the same on an institution-sized vat of peanut butter at the discount store as I spend on a 100% peanut 750ml jar. Then there is Donald A. Coffin's point about family sizes.

Also, you have to respect explanation #4 because as others have observed, it is not only larger carts that must be explained. In one store, I have a choice of a hand-held plastic basket, a larger, 2-wheeled plastic basket, a small, columnar double-decked metal cart, and a large, triple-decked metal cart.

[Bob Smith]

"In the US Walmart gutted the existing grocery players. That hasn't happened in Canada and won't. The Hard Discount franchises prevent that, Walmart can't undermine the national chains on price, the chains are already there. Walmart's grocery operation in Canada is an also-ran."

I wouldn't be so sure about that. Walmart may not be able to undermine the discounters on price, but they can match them on price, and undermine them on quality. There's nothing particularly fancy about my local Walmart, but it's infinitely more pleasant than any No Frills or Price Chopper I've ever been in (and you don't pay for bags), and frankly, in some respects, better than the local Loblaws/Superstore (which company has, generally, been horribly run for the past decade). The fact that Price Chopper rebranded itself as Freshco and substantially cleaned up its stores at roughly the same time that Walmart started expanding its grocery operations might be a co-incidence, but ...

[Frances Woolley]

What I'm hearing is that people aren't really particularly interested in the behavioural economics explanation - there are a plethora of economic, technological and social trends that can explain the large cart phenomenon.

I'm also hearing that large carts are no longer a ubiquitous thing. Just as the latest US housing data shows that houses are actually getting smaller, shopping cart size seems to stabilizing. Just as people are trading in their monster homes for downtown condos, they're trading in their monster carts for mini baskets and daily trips to the grocery store. This jives with Eaton, Eaton and Allen focus on the importance of time and lifestyle choices - but no one has voted in favour of their (probably Doug's) idea that increasing female labour force participation is the specific cause of mega-carts - in part because many commentators have their own pet theories of the specific cause of mega-carts!

The variety of explanations reminds me of something Greg Ip said in a recent interview, about interviewing academic economics:

One thing that I have learned about academics is that – even when they are not partisan or biased in the direction the research goes – I think it’s the nature of the discipline that academia rewards people who develop very strong views on often narrow subjects. So if you developed the view from your research that banking crises always come from moral hazard, you will tend to interpret all events through that prism. When you’re talking with someone like that, you need to keep this in mind. And when they wander out of their field, they may be less knowledgeable. Somebody once joked that the worst thing you can do to an economist is give him a Nobel Prize, because it gives him license to opine on almost anything.

[Shangwen]

An interesting contrast between industrial operations and consumer behavior can be seen in shows like Extreme Couponing, or design programs that feature enormous amounts of home storage. Over a period (from the 70s on) when North American companies started adopting just-in-time processes, many consumers have done the exact opposite. Shopping in a store that has an advanced inventory management system (RFID, or Wal-Mart's private satellite network) ironically makes it easier for consumers to oversupply their homes for a given period. Of course, the trade-off between margin and volume makes this sustainable for the sellers.

So, one of the intangible goods provided by supermarkets and large shopping carts is that it facilitates subclinical hoarding and the associated feelings of security. I have often noticed at Loblaws/Superstore how, for a product available in 500g, 1 kg, and 2 kg, the 1-kg item has the lowest cost by weight. The margin is jacked up for the largest item because so many people assign inherent value to the greater quantity.

[Frances Woolley]

Shangwen: "I have often noticed at Loblaws/Superstore how, for a product available in 500g, 1 kg, and 2 kg, the 1-kg item has the lowest cost by weight. The margin is jacked up for the largest item because so many people assign inherent value to the greater quantity."

I don't think it's so much "inherent value to the greater quantity" as it is a form of price discrimination - I suspect that type of pricing is particularly prevalent for goods bought by time-poor parents, e.g. ketchup and peanut butter, who are focused on getting the shopping done and getting out of there.

The counter-example to this type of pricing is white vinegar, where 4 litre bottle sells for a few pennies more than a 750 ml or 1 litre bottle. Price discrimination again - selling to different markets.

That's a fascinating observation of the replacement of commercial storage with home storage... We're now keeping the firms' inventories for them.

[Frances Woolley]

Comment submitted via email (confirming Nick's and Donald Coffin's earlier observations):

I think Nick is on to something - men are often the worst for loading up the cart with "bargains". Also the design of carts has definitely changed - the big box isles are wider allowing for larger carts. In the old days it was not that uncommon to see people with 2 carts of food at the check-out counter. I find that I do at nearly half my shopping at stores other than the large super- market i.e Whole foods and the veggie store and sometimes the butchers as well

[Jacques René Giguère]

Frances:the other half of warehousing is one of the other component of just-in-time manufacturing: nighttime truck traffic. The interstate-interprovincial system is used by people-moving cars and buses during the day is is a giant bumper-to-bumper rolling warehouse at night. A lot of trucking companies ( as well as FedEx and UPS) now define themselves as supply-and-inventory managers. My students in International Trade are always amazed at the new dimensions of logistics.

[Shangwen]

Here is an interesting spin on the behavioral econ explanation: even a shopping cart with theoretically infinite capacity (and an anesthetized payment experience) is drastically underused.

[John]

Maybe someone pointed this out already.

I'd imagine that non-perishable food items have become a much bigger part of the American diet than they were in the past. Frozen meals, chips, cookies, canned food, and other types of packaged, processed food. You can load up on that stuff and not worry about it spoiling.

[Determinant]

I wouldn't be so sure about that. Walmart may not be able to undermine the discounters on price, but they can match them on price, and undermine them on quality. There's nothing particularly fancy about my local Walmart, but it's infinitely more pleasant than any No Frills or Price Chopper I've ever been in (and you don't pay for bags), and frankly, in some respects, better than the local Loblaws/Superstore (which company has, generally, been horribly run for the past decade). The fact that Price Chopper rebranded itself as Freshco and substantially cleaned up its stores at roughly the same time that Walmart started expanding its grocery operations might be a co-incidence, but ...

Except the two Walmart grocery stores I have been in are more expensive than No Frills. They're playing at being a mid-line grocery store and not beating the discounters on price.

[Abhay]

I think authors chose #1 because this makes more sense from retailers point of view.

#2, 3 and 4 would be true if female or rich shoppers started complaining to store managers that carts are small? But why would retailers invest in larger carts? In fact, they have more incentive in terms of increased sale if they force people to come often.

What about obesity? May be it is a result of larger carts rather than a cause. But it makes sense to assume that fatter population eats more food.

Traffic congestion can be another explanation (same logic as female participation in workforce)

[Andrew F]

Walmart is about 2% more expensive than your typical hard discounter. They try to make up for it in service (hot deli counter, in-store bakery) and one-stop-shop service. On the other hand, other customers can get quite annoyed by a front-end clogged by grocery shopper when they want to buy just a few general merchandize items. I've found Walmart's checkout experience to be very poor, especially in stores with the full grocery section.

Walmart's grocery offer is almost exclusively about driving traffic to the rest of the store. Groceries bring in customers weekly, which is usually an improvement from what a discount department store usually gets from a customer.

[Reverend Moon]

According to Unarco Industries,

"The shopping cart has undergone many more changes as time has progressed. The shape has changed as ergonomic considerations were taken into account. For the most part, carts are bigger to hold more merchandise, thus enabling more sales."

http://www.unarco.com/carthistory.html

I would guess that since bigger carts enable more sales it was a combination of cost of carts falling and materials/technology changes (lighter) that allowed for bigger carts. They're not stored inside anymore either.

[OGT]

Slightly surprised at the near total resistance to #1here. Store owners would have to be irrational/ stupid to pay evil marketers all that money if they weren't producing something of value. I have a number of friends in the evil industry and they spend a good deal of their client's money to produce and exploit small asymmetries in information.

That said, I think the Reverend Moon is right (never thought I'd write that). It's a combo of factors allowing retailers to maximize people's rational inattention to shopping volumes.

[Frances Woolley]

OGT "Slightly surprised at the near total resistance to #1"

Commentators are a biased sample of WCI readers, who are a very biased sample of the world in general.

[ezra abrams]

This blog post is everything wrong with economics
It should *start off* with a chart showing the size of grocery carts over time.
or a table
or some *data*
yeah - you know data, that stuff that trumps theory ?
And, if the data were at all decent, like in what we expect from a second year grad student, there would be time series for same stores (we should have at least 50 years of data for some stores - maybe 100 in the sears or Mont gomery ward archive)
but no; the closest we come is an extract from a text, and it is not clear if the text actually has any empirical data.

The best you can say is that the authors had this cute thought, and being lazy, decided not blog it before spending a few minutes to find the data (a few minutes in net time; you put the thought on an index card on your corkboard, and let it sit for a year till you happen to run into the data)
so it is all just navel gazing

[Stephen Gordon]

I've had enough of your abuse, and I'm sure Frances has as well. That's it. You're blocked.

[Frances Woolley]

Ezra -

Complain to the Right Honourable Stephen Harper, and also to Statistics Canada: if there was any publicly available firm level data, we'd be analyzing it. The data isn't available. I wish it was, but it isn't.

Stephen -

Glad to know you've got my back!

[Donald A. Coffin]

John comments: "I'd imagine that non-perishable food items have become a much bigger part of the American diet than they were in the past. Frozen meals, chips, cookies, canned food, and other types of packaged, processed food. You can load up on that stuff and not worry about it spoiling."

Actually, going back (again) to my unscientific single observation of when I worked in a grocery store, fresh produce sections today are much larger than they used to be. This is a consequence (I think) of the reduced costs of transporting fresh items longer distances (refrigerated shipping used to be *relatively* much more exepnsive than it is now, and driving times used to be much longer. The variety of fresh food available literally year-round is considerably larger than it was even 25 years ago.

[Determinant]

Good to see Stephen is reading.

Totally off topic, I noticed somebody tweeted you about looking at jobs.gc.ca for government positions. I am an expert on using that, but I have to say that hunting for federal Public Service jobs is not like the private sector. The resume format is completely different, the expectations are different and the stages are different. There is an extra layer of law (the Public Service Employment Act) and language requirements which has a practical impact on what you see as a job hunter.

If you approach the search without specific guidance, you can very, very easily lose an opportunity because you did something you shouldn't have and didn't know better.

That kind of advice your friend tweeted you is what I call a "shove", it looks useful but it isn't because you need to know a lot more to do anything useful.

I'd be happy to explain just how to give yourself a few legs up in Public Service of Canada job hunting, there are some easy was to make your life a whole lot easier.

[Stephen Gordon]

I read all posts and all comments.

[Alex1]

This is an unrelated question about empirics, but is that necessarily true:

"If the trend towards larger trips to the grocery store was caused by increased female labour force participation, I would expect to see big stores and big shopping carts predominating in areas where female labour force participation is high, and more traditional, small shopping cart shopping experiences in areas where female labour force participation is low."

Couldn't it be that there is two events going on at the same time:
1) Women enter the workforce and this trigger bigger carts
2) Cart producers standardize for the bigger carts, thus they are alvailable in all regions
Would there be any statistical method to account for such possible compounding factor?

[Frances Woolley]

Alex "Would there be any statistical method to account for such possible compounding factor?"

Not really. If there is no variation in cart size at any given in time (because of standardization), all you've got is a time trend, with shopping cart sizes increasing at the same time as female labour force is increasing.
And house sizes are increasing.
And availability of prepared food is increasing.
And road networks are expanding.
And car ownership is expanding.
Etc.

It's extraordinary difficult to sort out causal relationships with time series data, unless one can point to some sudden event, e.g. the birth control pill, and see if the time trend changes at the point when the extraordinary event occurs. Even then it's not always convincing.

[Mike]

I don't think these explanations are mutually exclusive.

[Determinant]

Further to my post to Stephen:

If hunting for federal Public Service jobs, these are the guidelines:

(1) Every job comes with a Statement of Merit, you absolutely must demonstrate that you meet all of it to be hired. Read it very carefully and repeat exactly what they say with specific examples from your experience.

(2) Public Service resumes are not brief. Further to (1) you have to be verbose, explicit and descriptive in your claims. Do not make claims you don't have evidence for. It is worthwhile paying a government-oriented resume writing service to get your resume into the proper form. It is an art.

(3) Learn French. Linguistic requirements matter, a lot. All jobs in the National Capital Regions are English/French Essential and BBB (mid level) in the other language by policy. It is so frequent to see and so easy to filter people out on. Ottawa is the centre of the universe for second language training and there are schools that run internet-based courses available anywhere in Canada. It is the one asset that is so easy to improve with a little effort and it pays such dividends. If bilingualism is not required it will be an asset, which means it is a tiebreaker at the very end, if needed to narrow the pool.

If French is your first language, learn English (same schools above), do not assume you speak it to BBB standard (you probably don't). The language tests are not easy and designed to filter people out, they are not a walk in the park.

I hope this helps your students and/or twitter followers.

[Peter T]

Empirical observation: here in Australia larger supermarkets introduced larger carts, but the smaller ones did not (in Canberra the big retailers have not quite driven out the small ones). But in the last few years the non-discount large supermarkets have down-sized the carts. So the supermarkets are driving the changes. the downsizing seems to accompany self-service but also an aging population buying smaller quantities and a shift in the supermarket strategy towards emphasising smaller quantities of higher value (fresh bread, delicatessen things and so on). At the same time their house brands have tried to move up market. This does not align with the behavioural economics explanation, but also not quite with the others. The retailers are driving the change partly in response to changes in consumer preference, but also on their own account.

[David Tufte]

Let me take you in a completely different direction. Retailers are, in part, warehouses. And homes are, in part, warehouses. When we shop we are choosing to not keep the 5 pound bag of sugar at the store, and to keep it at home. The cart size is an indicator of our willingness to warehouse at home.

And why would be more willing to do our warehousing at home? Because our homes have gotten larger and our households have gotten smaller. Space to warehouse stuff it home is at record high levels, and acting on that might necessitate bigger carts.

[Andrew F]

David, that also requires less frequent shopping. Even if the stock of goods in the home grows, the flow should be about the same, or even falling as household size has declined.

[Frances Woolley]

David - the point about warehousing is a good one - Shangwen made it earlier.

[Chris J]

My mom shopped once a week and *filled* a cart (layers deep). She shopped once a week Saturday morning because that is when the store had the freshest produce. That was 35 years ago. We shop now probably twice to three times a week from the one large grocery store within 15km. West of us is a more rural area, and my cousins when they come into town to shop, buy a huge amount of stuff.

This makes me think it is a bit like buying a minivan when 51 weeks of the year you need a car and only one do you need more space. If only a few customers need big carts you supply them.

And yeah, it is easier to toss in stuff you don't need when there is an empty spot in the cart. Grocery stores are consumers of behavioral ideas. Why aren't milk and bread together? To make you walk past the most stuff. There is no doubt in my mind that the data mining done on the rewards systems more than pays back the $220 in "free" groceries we have earned this year. When you need two items in a grocery store see how far you have to walk in that store. And they have used that data to work out what ought to be between those two items.

[Frances Woolley]

Chris J: "When you need two items in a grocery store see how far you have to walk in that store."

If the two items are high mark-up inessential purchases, not far at all; if they're low mark-up essential items like milk or eggs or frozen peas, as far away from the door as possible.

One thing that's really interesting is how different the lay-out of the local Chinese supermarket (Kowloon Market) is from the lay-out out of the local Loblaws. At Kowloon market, rice is right at the back, on the far side of the store, where frozen vegetables would be in the local Loblaws.

[Nathanael]

Nick Rowe: here's a problem. You aren't thinking scientifically in this post

All of these theories are likely to be simultaneously correct. I don't see any reason to believe that any of them are wrong. Ask the question, what evidence would *disprove* any one of the theories? As a scientist, you always look for the evidence which *hurts* the theory.

This looks like an argument against the "they're fooling us" theory: "Customers might get fooled once, but eventually they would learn that larger carts are bad things, and would only visit stores with small carts. " But it isn't really a good argument, because people are stupid in this particular way, and it doesn't harm them that much to be stupid in this particular way.

This looks like an argument against the 'two-income family' hypothesis:

"If the trend towards larger trips to the grocery store was caused by increased female labour force participation, I would expect to see big stores and big shopping carts predominating in areas where female labour force participation is high, and more traditional, small shopping cart shopping experiences in areas where female labour force participation is low. I would expect to see relatively few at-home parent pushing mega-carts around mega-stores. That's simply not the case."

This is an argument against the two-income family hypothesis, but it doesn't disprove the possibility of the two-income-family hypothesis *plus* a secondary explanation. (Is there anywhere where the female labour force participation is truly low?)

Other than that, I don't see any serious attempt to disprove any of these arguments. They could all be true. It seems likely that a whole bunch of phenomena at once led to the rise of large carts. Perhaps the most important is that it never seriously *hurts* a store with wide aisles to have large carts, so why not? (Many of them still have small carts.)