Why Are Discount Stores Full of XS and XXL Clothes?


My former Ph.D. student and frequent co-author Erik Snowberg sends along an interesting question:

Why do discount clothing stores (like Nordstrom Rack — and clothing sales in general) have an excess of really small and really large sizes?

I have to admit, I’ve always wondered. Erik continues:

The typical answer seems to be that there are more medium [people] in the world than XXL’s. Duh.

But the problem with this idea is that clothing buyers for stores obviously know this. And if they didn’t at first, they should have figured it out by now.

Hmm. He’s right — it is important to think about the supply side as well as demand. If there are twice as many mediums as XXL’s, then twice as many should be produced; and so it should be about as likely that a store will be left with excess mediums as XXL’s.

Here’s Erik’s explanation:

My best guess is that for some reason small and large people are, in general, less willing to pay for clothes. Maybe because they are outside of the norms for physical beauty, they believe that sharp clothes won’t help them that much.

Whatever the reason, a clothing store can’t set lower prices for different sizes, so it price discriminates by waiting a little bit and putting remaining stock on sale — which happens to be (surprise!) in the large and sometimes small sizes.

That’s a pretty interesting story, and it may well be right. But there must be readers with better information, or competing theories.

What are your thoughts?

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.



  1. Christopher says:

    Could there be a disproportionate amount of larger people in the group of less wealthy people? There have been many links to obesity and being poor. And, naturally, the poorer you are the more prone you are to look for discounts on clothing. The converse is then true, the richer you are the more you’ll spend on clothes.

    So the end result of your theory would be true- clothing stores stock the same amount of S, M, L as they do XL, XXL, XXXL… they just count on discounting the larger ones and waiting to sell them- knowing XXL customers will eventually buy them.

    Thumb up 0 Thumb down 0
  2. Ryan says:

    Hm… maybe people who are less well off or just plain cheap don’t mind buying ill-fitting clothing if it’s on sale? Sure they could put normal sized clothing on sale, too, but then alot of people would just wait until it’s on sale.

    Thumb up 0 Thumb down 0
  3. Adam says:

    Speaking from personal experience, overweight people are less willing to spend a lot of money on clothes because they (we) are always thinking that the big change is right around the corner. There’s no reason to waste money on fat clothes when we’re going to get our butts in gear and drop this excess weight soon.

    Thumb up 1 Thumb down 1
  4. eman says:

    How about this: they are targeting obese/overweight (probably negatively correlated with income) with the XXLs and foreigners (Asian, some Latin Americans, etc) with XSs?

    Thumb up 0 Thumb down 0
  5. Dennis says:

    Do discount stores like Costco and Walmart have the opposite proportions of sizes left over?

    Thumb up 2 Thumb down 1
  6. Elizabeth Anne says:

    Actually, I think it’s this: if you’re buying a clothing line in which you are the “XS” or “XL”, you probably won’t buy within that line – if you’re an XS you’ll most likely drop into a petite line in which you’ll have more gradation of sizes, and if you’re an XL you’ll head to the Women’s department for the same reason.

    Thumb up 1 Thumb down 0
  7. Marty says:

    Perhaps many of the XL and XS clothes have been returned. Many people who are of odd sizes will buy a lot of clothing items, then try them on at home where they feel more comfortable, then return them. Many of these go straight to discount, or they’ve been out a while and now the store only keeps them in discount.

    Thumb up 2 Thumb down 0
  8. Kevin says:

    I like the explanation provided above, but I’ve got another idea: a statistical explanation. Disclaimer: I have no knowledge of clothing industry practices and haven’t taken economics beyond introductory-level courses.

    Maybe clothing stores attempt to ensure that, with some level of confidence (90%, for example), every customer who wants to buy a given item in a given size at full price will be able to do so. Perhaps, then, there is a greater variance in the quantities demanded by small and people. This would force stores to buy more extra small and large shirts above what they expect to sell than medium sizes.

    Just an idea.

    Thumb up 2 Thumb down 0
  9. MattV says:

    I’ve always wondered why stores and manufacturers can’t seem to get their act together when it comes to clothing sizes. For instance, would it be that hard to make pants in one-inch increments for inseams? In jeans, 32s are always just a bit too short for me after drying, and 34s are usually a bit too long. I am right at 6 feet tall, so there have to be millions of guys who have the same problem. I was just at Target yesterday where there are sales racks full of S, M, and XXL. One has to wonder why they keep making so many of those sizes if few buy them, regardless of the reason that few buy them.

    Thumb up 1 Thumb down 0
  10. Ryan says:

    Frankly I think it just has to do with having inventory everywhere. A store needs to make sure that it can make a sale to everyone that walks in. So it stocks a bunch of sizes, more in the middle than at either end, but they’re all still in stock. Now because of the large population of medium people, they’re fairly well represented and distributed in any geographical area. So there’s always someone to buy out the stock. Now consider the population at either end of the spectrum is big enough to buy out the stock of any one store in an area, but not all the stores in an area, so that despite every store being able to sell to either extreme, the distribution of the extremes doesn’t favor enough people of either extreme in every area to buy out the stock of every store. I’m sure someone can state that better than me….

    Thumb up 0 Thumb down 0
  11. walter wimberly says:

    I would have to say it is from poor planning on the (store) buyers part. If you look at the demand side of it, you cannot simply look at the population size and say that there are 2 times the number of mediums as XXL, and therefore buy 1/2 for the XXL size, and then still consistently have to mark them down.

    The buyer needs to look not at the population size, but the purchasing size of that market segment so that instead of buy 1/2 of what they would buy for a medium, maybe they buy 2/5ths. (With enough data on what is sold, this would be more accurate – not a random guess.)

    You also have to consider that while there are rarely stores that specialize in “medium” there are specialty stores for “plus size”, “big and tall”, as well as “petites”. Each of these affect the supply side of the market in a way which a regular store should account for.

    Regular excessive amounts of discounted items relate to nothing more than either a) original prices being too high, or b) the store’s buyer not understanding the market well enough to know what to order.

    As a side note, it would be foolish for a store to want to wait to discount an item. Not only does their profit margin decrease on the initial sale, but at that point, they’ve probably used up the time on any of the net+30 agreements, and now have to either “rob peter to pay paul” or pay interest to the manufacture/distributor on that item.

    Thumb up 0 Thumb down 0
  12. GS says:

    Working for a corporate retail entity, there are lots of reasons outlets have off sizing like those mentioned.

    The main one is merchandising and sell thru – although buyers do buy things in less quantity for the less popular sizes, those sizes don’t necessarily sell – yet are needed in store for product presentation (retail merchandising calls for having a full size run) If every Nordstrom has a few extras in these low styles, they can then send them to the Rack. There are 105 Nordstrom and 51 Racks. I’m sure other retailers build their outlet partners based on a similar understanding of their sell through on off sized clothing. It’s a natural part of the business.

    Thumb up 0 Thumb down 0
  13. MNMike says:

    If the original clothing buyers estimate the distribution of sizes correctly but buy a bit too much then the full retail store will be left with the same number of each size. They’ll pass this inventory along to the discount store and the medium items will sell out immediately, leaving a rack full of XS and XXL clothes.

    Thumb up 0 Thumb down 0
  14. JamesM says:

    Yes, they CAN and DO have different prices for different sizes. I don’t see it a lot, but I do see it from time to time. Generally it’s a slightly higher price for a larger size. I guess you can justify this because of additional material cost. Different size = different product so it isn’t size discrimination, even if it is.

    Thumb up 0 Thumb down 0
  15. Otter says:

    This situation could make the average size customer (their main profit center) be more willing to pay full price. I know I’m discouraged when I go to discount stores and never find my size. I tend to immediately buy full priced or regular sale items that fit because I know they won’t be in my size when they’re on close-out.

    Thumb up 0 Thumb down 0
  16. Jeff H. says:

    My intuition is to look at the markets or XXL clothes and XS clothes separately. Nonetheless, I think price discrimination has a lot to do with both.

    For XXL clothes, I like Adam’s explanation that fat people are more demand inelastic for clothing because of wishful thinking.

    As for XS clothing, it’s worth keeping in mind that small sizes often blur the lines of segmentation. An XS men’s shirt on sale might, for example, entice larger young males or even women who for whatever reason aren’t willing to buy at the normal price.

    Thumb up 0 Thumb down 0
  17. Dan says:

    I am a XXLTall guy, I gave up on finding my size in retail stores (business and business casual) a long time ago. 99% of my shopping is done on-line or by catolgue. I am not sure how this choice is reflected in the discount rack scenario, but what I want or need is not generally avaiable.

    Thumb up 0 Thumb down 0
  18. Susan says:

    But isn’t this infinite regress? Shouldn’t the buyers have figured out that in addition to there being fewer XS and XL people, they buy less per capita, and thus they should stock accordingly?

    Thumb up 0 Thumb down 0
  19. Richard says:

    Another theory
    Shops need to have some stock at each size but as the number of stock required gets higher the % oversupply drops as you can’t have half a dress on a rack.
    These overstocked items are then transferred to discount stores but now the number of medium sized clothes is closer to the number of large and small sizes than the original stocking requirements.
    At this point the medium sizes are run out quickly and only the smaller and larger sizes remain

    Thumb up 0 Thumb down 0
  20. Brian says:

    I’m sure part of excess in XXL and XS clothing is because medium-sized shoppers, being more prevalent, buy out the discounted stock at a much faster rate. Knowing that if you leave the store to think about it you won’t find the same item when you return could lead to more impulse buying. You have to be on your toes to make good finds if you wear a popular size.

    Larger and smaller than average sized people know after a lifetime of experience shopping that they don’t have to hurry to purchase an item. The extra time to think over the purchase decreases their chance of buying clothing they don’t need.

    Thumb up 0 Thumb down 0
  21. Rev Matt says:

    Another possibility: people in sizes other than XXL may buy multiples of something they like figuring they aren’t likely to drastically change their size in the next few years, whereas people in the XXL range think that they well (this is related to the comment of Adam). This does not provide an explanation for the XS.

    Thumb up 0 Thumb down 0
  22. LB says:

    It may not be just the style of clothes or the price that leads to the remainders, it may also be the store’s style or reputation. It is for me, anyway. I’m an XXL-wearing woman and it would never occur to me to shop at Nordstroms, and therefore I’d never go to Nordstroms Rack. Nor to Macy’s, Bloomingdale’s, or any of that ilk.

    The times I’ve tried to buy large-size clothing in upscale department stores, I’ve literally been discouraged – by the store’s layout, by the looks and “assistance” from sales clerks, and by the looks (and occasional reactions) from other patrons. Not everyone or every salesclerk is rude, some are very helpful, but there have been enough incidents that I just don’t bother any more.

    The women’s section or large-size section in a department store tends to be in the back of the store. It tends to be further away from the fitting rooms. There tend to be fewer, if any, attendants available. It is generally not a welcoming message. There have been enough remarks like, “we don’t get much call for that [a business suit!] in this size range” delivered in an insincere tone to make it clear I wasn’t welcome. (Think of the first shopping scene in Pretty Woman, and the sneering “help” from the clerks – that’s about right.) Why would I return to a store that makes it clear they don’t want my money? If they’re then left with the few garments they bothered to buy in my size and have to sell them at discount, that’s their problem. If you drive away your customers, you can’t expect to sell your stock.

    Also, “XXL” means different things in different stores. In some stores and catalogs, XXL translates to a 20-22. In others, XXL is a 16 or 18. I’ve received some specialty catalogs where XXL is a 14!

    The average size of a woman in the US is a 14. Many stores stop carrying clothes at size 16 or 18. They’re ignoring nearly half the market potential.

    I make a good living (6 figures) and am quite willing to spend money on clothes that fit and look good. However, I’m going to do my shopping where I’m welcome.

    Thumb up 0 Thumb down 0
  23. mark says:

    I am on the business side of retail, and I believe the answer is similar to #8 above. When you do merchnadise planning (and this is built into the planning systems), there is uncertainty behind the demand you will experience for any product, and for any size of any given product. For smaller volume items, the standard deviation behind the predicted demand is obviously greater. So as you try to fill demand, there is a higher likelihood that you will be overstocked (and a similar higher likelihood that you will be on backorder) for low volume items. Since large and small sizes are out of the tails of the bell curve, you are always going to find that these sizes are either overtocked or on backorder with a much greater frequency than sizes in the middle of the bell curve.

    Thumb up 0 Thumb down 0
  24. Matt says:

    My thoughts, as someone working for a major apparel discounter, is we magnify the industry problem.

    A vendor usually will pack orders like this:


    There are many more S-M-L customers, than XS/XL/XXL etc…and the size breaks do not accurately reflect the actual size breakdown of the customers.

    When discounters buy closeouts or orders that didn’t sell elsewhere, they receive an inordinate amount of XS/XL etc.

    So when you look at a discount retailer, the small errors of many manufacturers are gathered in one location , skewing the size breaks even more.

    Thumb up 0 Thumb down 0
  25. Sam says:

    I have a friend who imports wholesale clothes into the UK and he explained this to me once. There is apparently a LOT of poor quality control on sizes when you order from cheap factories overseas. So its pretty common to have an entire batch of Medium shirts which turn out to be Large, no problem if they catch it as they just relabel them on ship them. But if the items have been made beyond the last size (ie smaller than XS or larger than XXL) then there is nothing they can do except sell it via a discounter. This is why you get dispropotionate numbers of the extreme sizes, some brands only go up to XL and for those you’ll find a glut of XLs on the clearence rails. This is why you really need to try on the actual item you’re buying in discounters like Ross and TJMaxx. Even if you try on one XLL shirt you may find that antoher doesn’t fit if you buy two….

    Thumb up 0 Thumb down 0
  26. Talyssa says:

    I’ve never been a big clothes spender. I have a small set of outfits for work and jeans and stylish tshirts for the rest of the time. When I started this job 3 years ago and was 40 pounds lighter (yes, really) I enjoyed looking at clothes in stores, putting together my few outfits, and so on.

    40 pounds later, I walk into a store, find a pair of pants that’s comfortable, buy it in two colors, a couple of shirts that will be appropriate for work, and thats about it. I don’t want to own more than that in this size, and although I think Adam’s comment applies a little bit, its more that when you’re heavy clothes are just a way to not get arrested in public for a lot of us. But the stores have to carry at least a few of that size in every section so there’s bound to be leftovers.

    RE: XS sizes I think there aren’t really that many people who are that small, but again they have to carry at least a few of that size in each style.

    BY THE WAY would you all PLEASE stop saying its because fat people are poor? Yeesh. Lower income people are MORE LIKELY to be overweight, but it doesn’t mean that overweight people are of lower income.

    Thumb up 0 Thumb down 0
  27. Bill says:

    My own experience, from being a male 2x and larger, is that we don’t shop for clothes in mainstream stores. I have been surprised to find small quantities of larger sizes in mainstream stores; but often only in “super-center” stores. So when in need of new clothes, I will head for the “big and tall” stores that cater to me rather than waste my time going to Target.

    If we’re not shopping there, then they ain’t selling what they got!

    Thumb up 0 Thumb down 0
  28. DanC says:

    One I’m not sure that discount stores are full of XXL and XXS clothes.

    Still, it could be that medium size close sell out faster at discount stores then more extreme sizes. People in the middle just buy more clothes.

    It could be that medium size people are more fashion conscience on average and buy hot trends while people at the extremes are less likely to follow changing tastes.

    Medium people are more alike. Large and small people are large and small in different ways (apple shape Vs pear shape etc). So buyers may just have a hard time finding clothes which look good on medium people that will still look good on smaller or larger people.

    If much of the cost of clothing for manufacturers is in the choice of style and color and the marginal cost of making different sizes is small then you might risk that people on the extremes might buy your product and you can still make a profit even with deep discounts.

    Look at shoes. Even if bigger shoes cost more to make, manufacturers don’t charge more for bigger shoes (in most cases.) That indicates that the marginal cost of offering various sizes is very small.

    Lastly, it may be easy to price discriminate on the basis of size, as long as you don’t make it too obvious.

    Thumb up 0 Thumb down 0
  29. Anna Kuperberg says:

    As an XS person, I have often wondered about this. My guess is that it’s more expensive to make clothes in small runs than in large runs. So if they’re going to make odd sizes at all, it’s doesn’t cost much more to make too many of them.

    On the other hand, whenever I go into a retail store looking at regularly priced clothes, I have a HARD time finding my size. And if I’m looking for a designer brand, then forget it, they won’t even bother making my size. Designer clothes also tend to be for taller people. (Are there more short-length pants on the discount racks? I wonder).

    Thumb up 0 Thumb down 0
  30. Alex says:

    From my experience shopping in discount stores, if you hit them on the days that new clothes come in, the medium/large racks have about the same amount of newer stuff as the xs/xl/xxl racks do, but the clothes on those middle racks go MUCH faster as there are many more people shopping in the mid-range size category! I used to be an XL and am now a M/L — discount stores were my favorite when I was larger as I could always find something liked for a reasonable price — since I was convinced that I’d loose the weight and not be wearing the clothes for long (LOL)! Only took half a decade, but now, as a M/L, it takes more effort and luck as I have to go on days they get their new merchandise in, else it’s gone too quickly! Even regular department stores are tough sometimes!

    Such a conundrum!

    Thumb up 0 Thumb down 0
  31. Al says:

    I’m in the closing business and size run issues are a major variance that we deal with. A typical retailer will purchase a suggested “size run” that looks something like 1 Small/2Medium/2Large/1 XL etc, with extended sizes getting less play. In general, retailers on the coasts tend to buy more of the smaller sizes and retailers elsewhere buy more of the larger sizes.

    Elizabeth in #6 hit on part of the issue. Most customers are in the market for medium and large apparel. Unless the apparel is specifically marketed to the extended sizing, there is less demand, and the process of adding extended sizes offers only a marginal benefit to the manufacturer.

    There is also a cost associated with making larger clothing, (more fabric, shipping etc,) and a discount with smaller sizes, but retailers and manufacturers choose to keep their transaction costs lower by offering one price across the entire size scale.

    Thumb up 0 Thumb down 0
  32. Erika says:

    I like Erik’s answer, but I think he’s answering the wrong question. Thinner and better looking people are certainly more willing to pay more money for clothes, which is why upscale women’s stores rarely sell sizes above 8 or 10 – when the average woman wears a size 14. (Try this experiment sometime next time you are idling in such a store – look at the distribution of sizes on the rack – you’ll see a lot of 2, 4, and 6s and not a lot of anything else.)

    The excess of size XS and XL on sales racks and discount stores has always been explained to me as a product of clothing being sold in lots – typically retailers can’t just order certain amounts of clothing sizes. So I don’t think retailers can actually adjust their distribution of available sizes for their particular shopping demographic as flexibly as they would like. My guess is that you’ll find more larger clothes discounted if you live in a thinner region of the country, more smaller clothes if you live in Wisconsin.

    Thumb up 0 Thumb down 0
  33. Erik says:

    I might be saying essentially the same as Kevin, but since the sample groups are smaller at the extremes of the size curve, the buyers are going to guess wrong more frequently with very small or very large sizes when deciding how many to purchase.

    Thumb up 0 Thumb down 0
  34. charles says:

    Retail uses the wrong model, the bell curve. Behavior intercedes, and the bell changes shape by the time the people show up to the story clipping the tails.

    Thumb up 0 Thumb down 0
  35. Jessie says:

    I think Ryan is onto something in #2. I am an XS, and I don’t have any problem with trying things on at the store, wanting to buy nice things, etc. (Those explanations may be true for heavier people, but then wouldn’t the leftovers be only XL, and not XL and XS?) But the idea that the most price-sensitive customers would also be the most willing to wear ill-fitting clothes makes a lot of sense. Taking it further, a number of stores actually produce clothes specifically for their outlets, or overproduce deliberately; the clothes can be made so cheaply that even selling them at a “discount” is profitable. Customers still feel like they’re getting a deal, even if they’re actually just paying less of a markup. So it might makes sense for a retailer to overproduce the sizes at the extremes of the spectrum; that way they’re capturing the demand of the people who actually are that size, but they’re also profiting from customers who might not otherwise purchase the item.

    Thumb up 0 Thumb down 0
  36. Jessica says:

    I would think that, if anything, very small and large people would be more willing to pay for clothing that fits well but have a harder time finding it. Often a style that works on a medium-sized person won’t translate into very small and very large sizes, and these leftovers are the extremes that didn’t fit anybody very well — an XS that wouldn’t fit an Olsen twin or an XL that wouldn’t flatter anyone. I’ve always found that the clothes that make it to the sale rack or the discount store are much less likely to fit well, even in my very common size, and would guess that errors in design would be even more magnified at the XS and XL tails.

    Thumb up 0 Thumb down 0
  37. Angela Smith says:

    Much of the reason is due to the fact that, with many manufacturers, shops cannot reorder/restock just one size. They are limited to ordering a garment by the ‘run’, which is a set selection of a size range (say, 10 X-Smalls, 20 Smalls, 40 Mediums, 30 Larges, and 10 X-Larges). If a store runs out of M & L, they are forced to order a full run of sizes although they may not need any more XS or XL.

    However, the opposite is true of other retail goods. The latter is a chronic problem for me as a consumer, because my shoe size is at the end of the run, and shoe manufacturer ranges are even more extreme than those of clothing manufacturers. One manager at a local shoe store told me that their typical size run contains forty each of size 8 and 8.5, but only two each of size 11 and size 5. As a result, discount stores tend to have very few options in very small and very large sizes–customers who wear those sizes have fewer choices, and are less apt to be picky than a customer who has her choice of every shoe in the shop.

    I think Elizabeth Anne above has a good point, however. In plus-size stores, the clearance racks are disproportionately stocked with size 14 and size 28–generally the smallest and largest sizes the shop sells. I posit that this is because women who wear a size 14 can usually shop at both “normal” and “plus-size” shops, and are thus less likely to be in a specialty store, and women at the upper end of the range are more likely to turn to catalogs and online ordering (not to mention that there are likely many more women who wear an 18 than a 28).

    Thumb up 0 Thumb down 0
  38. Clyde McPhat says:

    #24….has to be the winner. It’s a perfect theory. You order based on the number of people in the various size groups. The chances of someone in that particualr size group buying the product in YOUR store in greater in the s-m-l groups than in the xs-xl, so when you go to the discount rack, the clothing left over will be from the lesser. And if a s-m-l is on the sales rack, the chances of it being snapped up are greater due to a heavier volume of costumer in that category.

    It’s the same reason that there are more accidents on non-rainy or snowy days. There are more of them.

    Thumb up 0 Thumb down 0
  39. Logical Extremes says:

    Most of these explanations don’t address the feedback problem. Over, whatever the reasons for different takeup of different sizes, sales data will be fed back into the model and inventories should be adjusted. So that leaves either distributor restrictions (forces certain amounts of outlying sizes on retailers), retailer constraints (e.g., they like their displays to look balanced and are willing to price discriminate via seconding excess inventory), or just plain high statistical variance (for many of the reasons folks have stated)… I doubt that ALL items will end up with excess XS/XL+ inventory, only certain ones with high variance (which of course implies that certain items will also be hard to find in XS/XL+, which I can personally vouch for).

    Thumb up 0 Thumb down 0
  40. Kevin H says:

    The authors point only makes sense if Nordstrom’s and others place their sales orders by looking at popluation demographics. It makes much more sense for them to look at past sales figures, in which case the buying habbits of the large and small would already be factored in.

    I’m going to guess it’s a minimum order problem. When you are ordering 10,000 shirts or something, no one wants to make just the 100 xxs that you’ll need. I guess that would be a rare case of inefficiencies of scale =)

    PS. If Matt (#24) is right, there is a good bonus in stock for the exec that actually starts placing orders accurately instead of just doing 100%/50%.

    Thumb up 0 Thumb down 0
  41. Jacques Rene Giguere says:

    It is 3rd degree price discrimination. Like cars in the 50s where models changed so that higher income consumers bought the new models to keep up with the Jones, while used cars were sold down the income ladder. The euro-japanese imports took the down market and now you design different car lines instead of pushing them down

    Jacques Rene Giguere
    Professor of economics
    College de Sept-Iles
    Sept-Iles Quebec Canada

    Thumb up 0 Thumb down 0
  42. Liam says:

    Two markets at work here.

    First market is new merchandise. I’m sure the product buyers are good at ordering. Given the standard stated by others above, each store wants to cover all sizes in the first market. So you have end of season left overs pretty well distributed among sizes.

    Second market is discounts and left overs. With Nordstrom, I would guess this market is in the normal Nordstrom store, at least for a while. If you start with even distribution of sizes, the distribution of buyers leaves you with a big hole in the middle. I doubt that the product buyers for the first market have a goal to optimize distribution in the second market.

    Nordstrom Rack probably constitutes a third market, and so it even more concentrated at the ends.

    Thumb up 0 Thumb down 0
  43. chris says:

    I think the original observation is wrong. I worked in retail and pay a fair amount of attention where ever I shop. I don’t notice this issue here in the NY metropolitan area. (Although everyone notices when the store doesn’t have their size in an item they want).

    Thumb up 0 Thumb down 0
  44. Logical Extremes says:

    One more a-ha… Stores want to keep an item on the shelves as long as they have a reasonable offering for the bulk of their customers, i.e., lots of mediums and at least a few of some other sizes. Once the mediums (and larges) are gone, it’s tough to justify the shelf space. Variance is still the key though. My hypothesis is that XS/XL+ run out just as often as they are surplus.

    Thumb up 0 Thumb down 0
  45. DanC says:

    Perhaps the discounts on medium sizes is too large and the discounts on extreme sizes too small.

    Thumb up 0 Thumb down 0
  46. Sam Thornton says:

    My weight and build have remained pretty consistent over the last 40 years. Forty years ago I was wearing Medium size sports shirts. When overseas manufacturers started making big inroads into the market, I had to wear size Large. Since the Global Economy came onto the scene I’m up to XL. Do the math.

    Thumb up 0 Thumb down 0
  47. Simon says:

    Off tangent, but I’ve noticed that what was a Men’s L 10 years ago is now a Men’s S. I think Americans are getting fatter but manufacturers want them to believe that they are still within an acceptable range.

    Same for women too. Lots of overweight women call thin (but normal) women waifs or twigs to compensate for their ego.

    Thumb up 0 Thumb down 0
  48. Anon says:

    There is not an excess of small and extra small sizes as far as I have seen. I wear small and the sales racks are always full of medium and larges. Are there any stats that there are a lot of XS and XL?

    I find that it’s always good to challenge the assumptions of economists from the get-go.

    Thumb up 1 Thumb down 0
  49. Emmanuel M says:

    A few european ideas.

    * Situation is identical in Europe. I presume it’s identical everywhere

    * One idea I had about it was that the clothes are manufactured in China, then sent worlwide. So a clothes container must be sellable in different part of the worlds were people have different physiques. A set sold in Holland should have more clothes for tall people than a french one. A set sent to the US should have more strong sizes than one sent towards Japan. This is too much of a hassle to handle. So sets must cover a wider population distribution

    * The base manufacturing cost is muliplied by 10 or more when selling to the final customer. So having extra unsold copies costs far less, for an average retail shop, than missing a transaction. Due to this idea, you may want 5 extra copies for each size, but your 5 extra M mean 20 %overstock while 5 extra XS mean 100% of excess.

    Thumb up 0 Thumb down 0
  50. Dimitrios Symeonidis says:

    I don’t think the “xs and xxl people buy less clothes” explanations are valid.

    I assume that the relative percentages of production of each size are calculated by past sales numbers (which are readily available), not the more complicated statistics of general population multiplied by preferences per size etc.

    Thumb up 0 Thumb down 0
  51. Josh says:

    I once had a job forecasting orders for blue jeans for a large discount retailer. I have an explanation that may apply to items such as jeans or khakis that do not change as frequently as other more seasonal items.

    Items such as these were packed in cases of 12 by size. We sold approximately 3pcs of 34×32 per week while we sold approximately 1pc of 40×32 per week, so we would order less of the larger size. But the problem is that when you send a box of 12 to the store, the box of 34×32 would sell out in 4 weeks versus a box of 40×32 selling out in 12 weeks. A new box would be sent to the store when they were down to 3pcs. Say on Monday, you sell both the 34x32s and the 40x32s down to 3 and resupply for that store is scheduled for next Monday. The 34x32s sell out by the resupply day on average whereas the 40x32s probably have 2 left on Monday. And since this is an average, half the stores in a chain would sell out of 34x32s before a week was up.

    Throw in the fact that the increased volume of the smaller size leads to increased shipping and inventory control errors and you have even more reasons that the more popular sizes would be out.

    I’m sure there are other reasons for this problem but this is definitely one for certain less seasonal items.

    Thumb up 0 Thumb down 0
  52. GW says:

    I’m a fairly tall man (6′-5″) who is not overweight (185 lbs.) and the vast majority of clothes simply do not fit me right off the rack. I need pants with a 35″ to 36″ inseam that do not have a huge waist, and shirts with 36″ to 37″ sleeves, 16″ neck, and athletic fit (i.e. not ballooning out in the torso like a tent). I look good in nice clothes and appreciate them a lot. Typically, I have to get clothes custom made or heavily tailored…except at some of the discounters like Nordstrom Rack. Mainstream stores aren’t interested in buying clothes for people my size, so I stopped even bothering to look with them long ago. Every once in a while I’ll go back, only to be disappointed (interestingly, Nordstrom…the Rack’s parent store and supposed source of the Rack’s merch…does not typically carry clothing in sizes that fit me, while the Rack does). The “Big and Tall” stores aren’t really…they over-emphasize the “big” and leave out the merely “tall.” I don’t bother with them, either.

    All of which is totally anecdotal, but I know for a fact that there are lots of men out there who share my frustration on this. I can look good in nice, fashionable clothes (better, in fact, than most “normal” sized people). But the clothing manufacturers are mostly uninterested in providing any of their wares in my size. They seem to think that someone who is in the mid-six-feet range must weigh at least 300 lbs. So, their largest sizes are ridiculously huge, while still ill-fitting.

    I imagine that people on the short range of the spectrum face similar problems, though with far less problematic tailoring challenges (you can always take away extra fabric…adding more isn’t often an option).

    Thumb up 0 Thumb down 0
  53. Miles says:

    Possible that clothes are designed for the common sizes, and don’t look as good in extreme sizes.

    Thumb up 0 Thumb down 0
  54. Anna says:

    I wonder if it has to do with the fit of the clothing. To my knowledge, patterns for clothes are first cut to a person of average, or slightly below average size. For women, this might be a size 8 or 10. The patterns are then scaled to other sizes. If body types change along with body sizes, then it’s conceivable that in some instances, a garment that fits a size 8 woman perfectly would not, even scaled down, fit a size 2 woman perfectly. (A “nonlinearity”, if you will.) As a buyer, you might not be able to predict this. You then buy what you believe to be appropriate amounts in each size. Even if you’ve bought fewere xs and xl garments, if those fit poorly, there will be fewer buyers that expected. Voila- overstock!

    Thumb up 0 Thumb down 0
  55. Jeff says:

    The government mandated that clothes makers make the same number of garments for all sizes to eliminated discrimination. I think it was called the Clothing Resizing Act (CRA).

    I blame the democrats.

    Thumb up 0 Thumb down 0
  56. Jim says:

    First, externalities are important! A clothing store wants all sizes represented so they attract those customers. If I’m an XXL and Walmart never carries my size, then I won’t shop there for anything, clothes or otherwise.

    Secondly, the number of people consuming these sizes are fewer (think of the edges of a normal distribution). As such this “demand” is likely to be more variable over time so whereas “medium” clothes will eventually go off the shelves (ignoring style), XXL and XXS will constantly have either too many on the racks or too few. Due to my earlier point, stores always want too much rather than too few so they will react but having MORE supply than demand.

    Thumb up 0 Thumb down 0
  57. Mustermann says:

    I just asked some years ago. The Shop (actually the Chain) has to buy the whole line.

    Thumb up 0 Thumb down 0
  58. ZB says:

    Assume a manufacturer makes sizes XS, S, M, L, and XL. Toward the end of the season, supplies run low (or, perhaps, demand increases because items begin to get marked down). There are two sizes of people who are one size off from S, M, and L. (If you’re an L, you can squeeze into an M or get away with an XL; M can wear L or S; S can wear XS or M. Put another way, an XS and an M can get away with an S; an S and an L can get away with an M; an XL and an M can get away with an L.) But there is only one size of person who is one-away from XS and XL (an S can go down a size, an L can go up a size). There are simply fewer people for whom it would be reasonable to settle for a size-off article in buying an XS or XL. (Of course, this would rely on distribution of size across a population, and number of articles produced in each size, among other things.)

    Thumb up 0 Thumb down 0
  59. xxl says:

    Really simple answer: if you are xxl as I am, the only remaindered clothes to be found are sized medium. This is based on years of observation. I’ve often speculated that store buyers always overbuy medium.

    Thumb up 0 Thumb down 0
  60. gene grossman says:

    Several years ago we operated a small screen-printing company that did nautical designs on T-Shirts, our stock of choice at that time being “Hanes Beefy Tees.”

    We weren’t interested in purchasing any of the ‘small’ size, and only a few of the ‘mediums,’ but our distributor refused, stating that the factory forces him to include those sizes in each dozen ‘blanks’ purchased, and this sizing requirement was passed on to us.

    The upside was that at the end of each year we had several cases of brand-new, small, screenprinted T-Shirts to donate to the Los Angeles Children’s Hospital.

    Thumb up 0 Thumb down 0
  61. Tiffany says:

    The same issue exists for shoe sizes.

    Thumb up 0 Thumb down 0
  62. j says:

    I like the statistical explanation better than the ‘norms of beauty’ one. One way to decide between them: very small guys may be unattractive, but very small women are not. So the norms of beauty explanation would imply that both very small and very large sizes would be common for men, but only very large sizes would be common for women. The statistical explanation would imply that both are common for both genders. Any women want to comment?

    Thumb up 0 Thumb down 0
  63. Ralph says:

    But is it true that “most” people are actually concentrated around the size identified as “medium”? Looking around me in America, I’d guess that m-l-xl is the core, a bigger slice of the market than s-m-l. As for the XS remainders, I have no idea, but in menswear I don’t think you see them. Look at a Banana Republic men’s clearance rack, for instance: S and M combined will usually take up one rack, L a second, and XL a third. (And good luck finding trousers on sale with a waist less than 36.) That actually seems close to the body-type distribution I see around me. Not that this answers the question. But as someone pointed out, there are often other dis-incentives for people of larger sizes to shop; perhaps by stocking up on XL, some retailers are trying to send a message to those shoppers, “See, we carry your sizes!” But for whatever reasons, it doesn’t translate into selling those goods. Sorry for offering random disconnected thoughts…

    Thumb up 0 Thumb down 0
  64. db says:

    I think Dennis asks the right question. We live in a very poor, rural area and can find a disproportionate number of “mediums” on the discount racks despite the fact that the size distribution is tilted to the Xs on the regular racks. You need a larger and more diversified sample than Nordstrom’s – which could be very interesting.

    Thumb up 0 Thumb down 0
  65. Paul G. Brown says:

    A simpler idea?

    There’s money to be made by buying the middle sizes in bulk from the discount retail outlets, and then selling them on E-Bay (for example) at close to retail?

    Thumb up 0 Thumb down 0
  66. Ramstone says:

    Just another datapoint: This phenomenon can be observed at thrift stores as well, where the “supplier” is the community. Certainly there isn’t a disproportionate amount of donation by the XSs. and XXLs.

    Thumb up 0 Thumb down 0
  67. Brian Drought says:

    Suspect the answer is very simple.

    The clothes stores put the ‘unusual’ sizes on the racks so that you and I walk in, see a discounted item we like but in the wrong size and then go to the regular store to buy it at the normal price.

    Thumb up 0 Thumb down 0
  68. BenK says:

    Could be a variance in relative rate issue; that is, you’ll never see the XL/XS sizes for clothes that are _sold out_ already.
    In a smaller population there is a greater impact of deviation of rate of sale from the mean, so even if you predict your median demand curve perfectly and provide clothes X and Y in all the sizes in appropriate proportion to the buyer population, you’ll see, by chance, the most extreme sizes of one sell out and the others go on sale (when the mediums are all sold out and it can’t be stocked on the main floor anymore because it’s presence would just frustrate the average buyer).

    Thumb up 0 Thumb down 0
  69. Kads says:

    Normally, the clothing in discount stores is produced for those stores. But with the XS and XXL issue you are talking about distressed stuff. In production, on the cutting table, sections of the garment that are extra small may be paired with extra large to make the cutting as efficient as possible ie they may be cut together, even if prime store orders don’t match-up. Plus when buyers order they statistically over order what they buy least of. (BTW, they order according to what sold last year, not how many people of what size are in the population.) For purposes of illustration only, for a particular SKU for a particular store they may order 3 XS, 6 S, 12 M, 15 L, 2 XXL. But statisticaly they only should have ordered 1-1/2 of the XS and XXL. So in a big chain that all gets compounded. The over cuts and the unsold goes to the discounters.

    Thumb up 0 Thumb down 0
  70. Vikas says:

    Medium sized clothing leaves shelves faster.

    Thumb up 0 Thumb down 0
  71. Derek says:

    I wonder if the stores even send those sizes to the real store first. I know that certain manufacturers make special clothes that are different for outlet stores. May the same happens with their versions of xxl and xxs sizes, they may go straight to the outlet and may even be of a lower quality. Some one needs to compare the quality of the ones from the actual store to the xxl at the outlet to see if they are truly the same as the construction may differ. A good reason for stores to do this would also be that because so many outlets have xxl and xxs sizes that the xxl and xxs small crowd is used to going to the outlet completely skipping the trip to the actual store. Another point that is a bit out on a limb is that it could be psychological and a marketing tactic. If the large and small crowd does indeed skip the store and go straight to the outlet that would set up a possible benefit for the store as well. People like people that are like them. If the store separate the sizes they will keep the people who look similar together making them see other people like them buying the same products this theory opens a lot of psychological possibilities for setting the system up this way. The stores have a great deal of incentive to send xxl and xxs small to a separate store to begin with as I see it.

    Thumb up 0 Thumb down 0
  72. anonymous says:

    Well, we’re taking this person’s word for it that discount stores have an excess of Small and Large sizes. Do they really? Or do they seem to when this person shops?

    Thumb up 0 Thumb down 0
  73. Matthew L. Wong says:

    Selection bias?

    Maybe retail stores do buy their clothes with a normal distribution of sizes, but because there genuinely are more M and L people in the world, those sizes are sold/bought first and most frequently.

    Therefore, when you go to the store and observe a non-normal distribution, the tails of the distribution are more heavily represented on the shelves because the mediums and larges are the first to go.

    Thumb up 0 Thumb down 0
  74. Adam Caper says:

    The explanations that have to do with a structural misfit between merchandising requirements and size distributions in the customer base make a lot of sense, as do the ones that refer to a concentration of variance ending up at the discounters (i.e. there’s a recursion in which the middle sizes sell more quickly in each merchandising cycle, which amplifies what was originally an acceptable margin of error). And I agree with those who’ve pointed out that the question is poorly-stated (because the data would feedback and the system should self-correct).

    Even more, Erik’s answer troubles me a great deal — as an economist, he should have a stronger instinct to gather data before making an extreme claim to the effect that people at the tails of the bell-curve in body-types are inherently cheapskates.

    But I think that there are two major flaws in the question as posed. The first is that it assumes that discounting is undesirable for the parent entity. That’s obviously inaccurate — the fact that almost every major brand in retail (both labels and retailers) has an eponymous off-price outlet is major evidence that it’s not a bad business to be in. That’s even more evident when you stop to consider that they do so at the risk of their primary brand and there are plenty of other ways to dispose of extra stock (Filene’s Basement, Marshalls, etc.) So they have to be in it for the money. So if they know they can make money with overstock, they have an incentive to overstock at the tail end of the curve in order to catch those — perhaps rare, but nonetheless profitable — instances where a really big or really small person walks into the primary outlet.

    The second flaw is that the answer doesn’t consider that the industry may, over time, have signaled to really big or really small people that outlets are the place to go to find the widest selection of styles. In that case, you’d expect to see less sell-through at the primary POS in favor of the outlet one. So demand for clothing is the same, but people have been taught to game the system — or perhaps more accurately, they’ve been trained by the retailers that the outlets are the most efficient way to buy their clothes. In order to test that theory, one would have to gather data on the ultimate disposition of the odd sizes — that is, if they all eventually sell out at discount, then it’s true that the demand is there but just being shifted. If the discounters end up throwing away a fair amount of those off-sizes, then we’re back to a mismatch between the # of off sizes that are ordered and the proportion of the population that sits in the tails of the curve.

    Thumb up 0 Thumb down 0
  75. Karen says:

    I once asked a salesperson in a dept.
    store why they didn’t stock bigger quantities
    of the sizes that more people actually wear.
    The answer was, basically, that the supplier sends them these irrational batches on a take
    it or leave it basis.

    Thumb up 0 Thumb down 0
  76. Dave Diamond says:

    When are you sampling the shelves of the discount stores? When I go to Building 19 looking for something cheap in size large, I often find more of the very-small and very-large sizes more available than any other. And I always think to myself, I should have gotten here sooner. Maybe there were more of all sizes but the more average-sized people shopped in higher numbers and bought up those items.

    Thumb up 0 Thumb down 0
  77. APR says:

    Facetiously, since the explanations above make so much more sense in the aggregate, maybe average-sized people (like me) have average taste and will buy anything the average retailer puts on the shelf whilst the large and the small are far more particular for reasons of both fit and having had to pay more attention to clothes (and, as a result, having developed a superior sense of taste in clothes.)

    Thumb up 0 Thumb down 0
  78. GBM says:

    Could it be as simple as the original store buys according to the distribution curve, but that implies that an even amount of each size will be left at the outlet.

    Therefore, when that same uneven distribution shops at the outlet, it will leave the extremes.

    Thumb up 0 Thumb down 0
  79. WholeMealOfFood says:

    I think the explanation from OR is related to the idea that the opportunity cost of a stock out is more expensive than the cost of keeping an extra item in stock, especially when it has salvage value (i.e. selling clothes to or through discount stores). Stores balance the risk of a stock out against the carrying costs of inventory.

    Low volume SKUs tend to have higher variances in their sales. So the right balance of inventory to protect against stock outs leads to a higher abundance of inventory in the aggregate (some stores that picked the right fashions may sell out, but when everyone takes this strategy, there will be excess inventory – especially if we assume that the number of SKUs offered to the extreme sizes isn’t too much different than for the more common ones).

    If we assume the population’s clothing sizes follow something close to a normal distribution curve, this would make XS and XXL sizes the low volume SKUs in fashion. So you see more of them in discount stores as a result of the original stores following a profit maximizing strategy.

    (disclaimer: I’m not an OR person, but I took a few OR classes)

    Thumb up 0 Thumb down 0
  80. DAM says:

    I would challenge the premise of the question–are the extreme sizes actually over-represented at the discounters, or is it a sampling error that has created this perception? Maybe the buyers are actually pretty good at what they do (at least as far as picking sizes). In my limited sampling of discounters and clearance racks, I sometimes see many S items, sometimes many XL and XXL items, sometimes mostly M and L, but usually a continuum with the middle sizes somewhat more numerous.

    Thumb up 0 Thumb down 0
  81. Ignatius Flacogordo says:

    By definition, XS- and XXL-labeled clothing are mavericks! They are different, they don’t “hang out” with the other clothing, and are, therefore, both inclined and more able to wander off on their own, be shunted aside, play with patterns of a different cloth, or hide beneath piles of more popular countenances. As they are generally regarded as either unsuited, unapproachable or hard-to-get, the elite floor walkers and clerks of Department stores regard XS- and XXL-labeled clothing as “special,” even though they won’t admit it, and, so, treat them differently than the mass-produced “middle-sizes.” Though exceptional, or perhaps because they are exceptional, XS- and XXL-labeled clothing usually end up in “special stacks” off the floor, left to dance with one another in out-of-the-way places; homogeneously grouped so that only the persistently committed and desperate might find themp; knowing that patience will reward them with a small or large bodily home sometime soon via the racks of the discounted open market.

    Thumb up 0 Thumb down 0
  82. buck says:

    as an former apparel analyst….Commenter 24.October 10th,2008 11:50 am is correct. (packets have portional size allocations that are not representative of the consumer…)

    as an economist…your assumption of rational agents (that is that sellers have learned) is incorrect.

    Why are sellers not maximizing profits? ask milton to look into his black box and tell us. Actually, there is a valid and reasonable answer. In the apparel trade, having a standard packet of units reduces manufacturing costs (consider the alternative of having to count for each style a different number of each size…and then packing those items efficiently for shipping). Although one wonders whether the cost savings from standardization is greater than the benefit derived from the flexibilty of varied production runs. To answer that one, I gather Adam Smith puts forth a reasoable, yet unoriginal assertion.

    by the way, where is my autograph?

    Thumb up 0 Thumb down 0
  83. Martin Saavedra says:

    Suppose n people of normal size (small, medium or large), and m people of abnormal size (extra small, extra large) walk into a store. Assuming that a shirt of his/her size is in stock, then the probability of that a customer purchases the shirt is p. Then, the number of shirts sold can be modeled as a binominal distribution, with p being the chance of a success, n being the number of trials for normal shirts, and m for abnormal shirts. Also assume that p is identical for people of normal and abnormal size.

    Since there are more people of normal size than abnormal size, it follows that n>m.

    The expected number of medium shirts sold equals p*n; the expected number of large shirts sold p*m. Consequently, the store should stock p*n medium shirts, and p*m abnormally sized shirts, if they want don’t purposefully under stock or over stock the shirts.

    Notice however, that the variance for the number of medium shirts sold equals n*p*(1-p); and for abnormally sized shirts, the variance is m*p*(1-p).

    Notice that the variance in the expected number of shirts sold is higher for medium shirts. That makes selling medium shirts more risky, because the store is less certain whether it will be under stocked or over stocked. Since it is a risky investment, the store then purchases less medium shirts than the expected number sold. Thus making it more likely that medium shirts will sell out.

    The net result is that medium shirts are more likely to sell out than to not get sold, and thus fewer end up in discount stores. Since abnormal size shirts are equally likely to sell out or not get sold, more of them, at least in comparison to medium shirts, end up at discount stores.
    Indeed this the observed situation is exactly what my model predicts.

    Now all I need is data to test the model. I sense a future publication.

    Thumb up 0 Thumb down 0
  84. Jennifer says:

    As an xs wearer and bargain hunter, I have not found an abundance of xs items on the clearance racks.

    Perhaps it’s different for men’s clothing. If men’s clothing comes in xs (which I haven’t often seen), it’s possible that some men are loathe to identify themselves as ‘extra small’ while in general many women may take pride in that.

    However, I have noticed that the xs clearance selection varies significantly by geography.

    Banana Republic’s clearance rack in Sioux Falls, South Dakota has a much wider selection of xs items than does the Rockefeller Center flagship store.

    This is indicative of both the amount of store traffic and portion of ‘extra small’ people among their target shoppers.

    Thumb up 0 Thumb down 0
  85. Justin says:

    comment #52, GW:

    This is the exact problem I have. Although not as thin as you (6’6″ 220lbs), finding anything wearable at a Big and Tall store is impossible. A merely Tall store would be welcomed.

    The majority of my clothes are found at Nordstrom Rack, Marshalls, TJ Maxx. Otherwise things have to be custom tailored (expensive).

    Shoes, same problem, although it is much easier to find size 15 shoes than size 34×36 pants or proper length sleeves.

    Living in a cold weather climate makes it especially tough because you need jackets as well as long sleeve shirts.

    There has to be enough demand out there to make affordable clothes for our body type, and I do see others like me walking around very well dressed. Where do they get the clothes?!

    Thumb up 0 Thumb down 0
  86. Shiv says:

    It seems to be related to high forecast error associated with low probability events. It is difficult to forecast a low probability event (extra small or extra large sizes) precisely. One prediction of this hypothesis is that stores run out of “extra” sizes more often than “medium” sizes.

    Another related explanation is that if stores want to keep the risk of running out of stock same for all sizes than you will see more extreme sizes.

    Thumb up 0 Thumb down 0
  87. Ron Tomlinson says:

    From my personal and limited viewpoint I believe there is a discretionary income difference between people of different characteristics such as size because there are employment selection processes which tend to favor peopele who are otherwise equally well qualified on non employement related criteria.

    The criteria which I believe to be an asset are: Size; Taller men and women new hire engineers according to one study benefitted more from being taller than from good academic grades. Physical attractiveness; numerous television and other journalists have found a bias in which the better qualified but less attractive applicant was not hired for an advertised position while an attractive candidate was offered a more highly paid positon than the advertised one. Nationality of name origin; In an organization where the management tends to have most names of a certain type of origin they will tend to promote within that group. A short last name easily pronounced is an asset. Alphabetical preference; when a group of equally well qualified candidates for a limited number of promotions has been produced, some management types look to the top of the list which is often sorted by alphabetical order. In the award of research contracts many researchers get their funding by adding an early in the alphabet low level research associate to the research title, and then listing the associate first.
    Size and discretionary income are also related as some food types which may contribute to obesity are inexpensive. When I have seen food stamps used in a grocery they are often in the hand of an overweight person.

    The next question is who shops in a discount house? If a person has a high descretionary income they don’t need to shop in a discount house. So this higher discretionary income advantage among the taller reduces their propensity for discount shopping. If good diet and time away from work for exercize and self care are also correlated to a high discretionary income the smaller sizes among the female members of the higher discretionary income group should also be a greater proportion.

    A buyer for discount has a better measure to project future sales from past discount house sales than the measure based on statistics available for the general population which includes upscale as well as thrift store shoppers.


    Thumb up 0 Thumb down 0
  88. scott says:

    We often look for one cause to a problem when there are many. Which generally adds to the problem. But if I had to place a bet, #51 is the closest.

    Thumb up 0 Thumb down 0
  89. RWC says:

    It is quite simple. Clearance racks are filled from an aggregate of outlets. The surplus is an illusion of being sold in one store representing many (and, yes, they sold ALL of the averages or they are still hanging in those same outlets and will eventually sell).

    No one is losing money.

    Thumb up 0 Thumb down 0
  90. Dan T. says:

    Fairly sure it’s a merchandising explanation (as per #8 and #24), and not Erik’s self-esteem explanation. Why? Because I shop at big and tall stores, and the same thing happens there — only the clothes left over are the XL and 5XL sizes.

    Thumb up 0 Thumb down 0
  91. Joel says:

    This can be a product of completely rational behaviour:

    You’re a store owner ordering stock. Based on your retail experience, you estimate (accurately) that for each person who walks into your store, there’s a 5% chance that they will want XS, 30% to want S, M, L and 5% to want XL. You also expect to sell 20 items.

    As a store owner, you decide to attempt to ensure that there is a 95% chance that any customer that walks into your store will be able to purchase the item in their size.

    Lets look at an X-size first. The probability of no-one wanting an XS is 0.95^20 = 35.9%. The probability of 1 person wanting an XS is 0.95^19*0.05 * (20 choose 1) = 37.8%. The probability 2 XS is 0.95^18*0.05^2* (20 choose 2) = 18.9%. The probability of 3 XS is 6.0%. 0-2 is 92.5% and 0-3 is 98.4%. So to guarantee that you keep 95% of your customers happy, you need to stock 3 XS. Similarly for XL.

    The same calculation for S,M or L reveals that there’s a 88.7% chance of demand for 8 and a 95.2% chance of demand for 9. So you stock 9.

    The expected left-overs are then 2 XS and XL and 3 S,M and L. The ratio of the demand for the “normal” to “X” sizes was 12:1, but the discount store gets stock in a ratio of 1.5:1. Clearly, they will sell out of the S,M and L more quickly.

    In general, this is a consequence of the central limit theorem. The standard deviation of the distribution of the number of customers wanting a particular size will grow as the square root of the number of customers. If you want to satisfy 95% of customers, you need 2 standard deviations worth of extra inventory (over the expected sales) for each size. If the proportion of customers wanting a particular size is p, the expected end-of-season surplus inventory (which is what ends up in the discount stores) will therefore be proportional to sqrt(p). The ratio of stock to demand (assuming the demand profile is the same) in the discount store will be (sqrt(p)/p). So small p => high ratio of discount stock to demand.

    Thumb up 0 Thumb down 0
  92. Nick says:

    It’s amazing how much class- and weight-based prejudice this post elicited (and contained). I think it’s incredibly telling that this discourse reached first to answers of the form “because poor/fat people are cheap/dumb.”

    Thumb up 0 Thumb down 0
  93. Caleb says:

    Almost by definition, the items that end up on the discount rack are the items for which there is less demand. If you as a shopper fit pretty well into the demographic targeted by a given store, there’s a good chance that you won’t find a lot of discounted items that appeal to you, because there are lots of other shoppers just like you also shopping at the same store. Of course, you look at the discount rack anyway, because sometimes you get lucky and because everybody likes to find a good deal. Naturally, most of the things you find there are unappealing either because they’re the wrong size or they’re just ugly, damaged, or whatever. Speaking for myself, I think my brain tends to put all those items in the same category, the “not for me” category. So, without some real data, I’m not sure I believe that the real issue is that there’s a disproportionate number of items in XXL and XS sizes so much as it’s that the items that I’m looking for aren’t available in the size that I want.

    Also, again by definition, the discount rack is a direct result of the error made by store purchasers in estimating how many of each item the store can sell. Most stores carry so many different items that the purchasers probably can’t reasonably estimate the demand for each individual item, and the purchasers are therefore likely to by different items in similar proportions. But some items appeal more to consumers of certain sizes. So it seems likely that you’ll find plenty of rugby shirts with horizontal stripes in size XXL on the discount rack because consumers of size XXL have largely figured out that horizontal stripes make them look even bigger than they are.

    It would be interesting to collect some real data so that we can tell whether discounted items really are mostly in the extremes of the size range, or whether that’s just an illusion. After all, most of us never pay any attention to non-discounted items in sizes other than our own, so our sense of how many items in the XXL or XS sizes there should be may not be reliable.

    Thumb up 0 Thumb down 0
  94. Steve O says:

    Interesting. I’ve always thought it was merely a production issue.

    Presumably production schedules, batches, and prices are optimized for the largest unit sales: M, L, and XL.

    But the XS and XXL are (over) produced using multiples the same schedules because that’s cheaper than setting different parameters for the smaller sales.

    Thumb up 0 Thumb down 0
  95. Glenn says:

    Ouch. I believe that people who do not look “naturally attractive” would probably spend More to compensate, than those who are seen as “naturally attractive”. Attractiveness is a form of social capital and people seek to optimize their “worth”… the same as any other kind of capital.

    Instead, I suspect that there is more variability in the buying habits of XS and XL folks. This would make it more likely for clothing manufacturers to overestimate or underestimate the demand. When the manufacturers underestimate demand, most medium people would not perceive the lack of XS and XL as a matter of course. Only the XS and XL do. Yet, when manufacturers overestimate demand, the M would notice the extra supplies on the shelf. I think that is sufficient to explain the perception of the phenomenon.

    But perhaps, in addition, this is also a marketing gimmick. Again, if the manufacturer can only poorly predict demand, and consumers go to a store and half of the time cannot find things in their size, they will likely shop elsewhere to avoid the hassle. If the manufacturer produces a slight excess, at least that may yield a more dependable consumer. Further, if purchases are made in groups, for a family for instance, then a scarcity of XS and XL may also reduce the sales of M, among hassle-avoiding consumers, who are seeking one-stop shopping.

    Thumb up 0 Thumb down 0
  96. AlleyGator says:

    This is an interesting phenomenon, but one that is visible in other areas of the market. For instance, I was flabbergasted to learn that certain companies which sell PVC Pipe Fittings make and sell them at a loss. If they are selling at a loss, why do they continue to produce them?

    The answer was that when customers show up to purchase an order, they wanted to get everything they needed and vendors HATE telling you to go somewhere else to get what you want. Both buyers and sellers want to complete business in one stop, so producing a product that did not earn them any profit on its own still helped out their bottom line and their customer satisfaction.

    As someone who wears “outlier” sizes (the only place to buy jeans with a 38″ inseam is at western stores) and who knows many other clothing outliers, I am far more likely to shop somewhere that makes me feel like they have a good selection. Keeping those XXXXL clothes on the rack probably makes big-boned consumers feel like there will be something for them so they can return back.

    Final thought: have you ever noticed that largest-size clothes tend to be marked up a few dollars? That’s probably because they end up with more spare inventory in the largest sizes. There’s no predictable stream of outlier clients to buy up inventory before the sale.

    Thumb up 0 Thumb down 0
  97. Greg says:

    Having worked in retail and procurement, I am inclined to suggest that there are a couple factors which contribute to this situation.

    First, the vendors who distribute the clothing probably require that the stores purchase X amount of XXS, and XXL, for every amount of S, Med, L. These vendors have contracts with their suppliers that have minimum required orders. So, although the vendors may know that these are unpopular sizes, they are forced to purchase them since the factories can only make a minimum amount for the production to be profitable.

    Yet, what I have also noticed is that there are far more XXL’s usually remaining than XXS. Many women/girls will try and wear the smallest size possible, to achieve a certain look or for their own self-image. I think people are generally more loathe, whatever their size, to wear larger clothing. So, even if the buyers purchase larger sizes in exact proportion to their target group, that group may choose smaller clothes, even if poor fitting, again for self-esteem/perception reasons.

    There are also those who may buy larger sizes, particularly for sportswear, since these clothes are often worn baggier, for both aesthetic or practical reasons.

    Thumb up 0 Thumb down 0
  98. gp says:

    My wife is from an Asian country and I did notice something interesting. She has a difficult time at Nordstrom or Macy’s, but it is easy to find her size in really expensive lines (Saks and boutiques). Most XS and petite stuff is made for thicker women. The obscenely expensive stuff fits her great in sizes 0 or 2. It sucks for me (prices are stupid), but there must be something to this.

    In addition, I notice that when I was young and broke I could wear a suit off the rack. Now that I am fatter and have some money I need to have my suits tailored extensively. I see tons of men with my gut, but most of them look worse than me in suits (I look bad too in casual clothing).

    Thumb up 0 Thumb down 0
  99. Sinead Mason says:

    I think Ryan hit on something–if all retailers have one extra shirts of each size, then there will be an equal number of all sizes that go the discount store, and the medium, large, and small sizes will sell out before the x-small and x-large.

    On the other hand, as someone who is xs, I can rarely find clothes that fit me in *any* store. It never occurred to me that discount places might be *more* likely to have items in my size. So I’m going to hit up the discount stores and do my part to reduce the excess of xs items therein!

    Thumb up 0 Thumb down 0
  100. Louis Anthes says:

    Larger clothes sizes means using more resources. Using more resources props up the economy — like larger food portions in restaurants (which exploded over the 90s and through the 2000s), and larger cars, homes, etc.

    Bascially, the cheap credit policies of the Clinton-Bush years made Americans fatter.

    This is not the first time this has happenedc in history. England did the same thing with forcing people to buy black woolen clothes in the 18th century — it propped up markets.

    Americans have been treated like cattle, since Reagan’s supply side theories won the day thirty years ago.

    And until they get off their fat asses and recognize it, they’ll get right back to it, after this “crisis” has passed.


    Thumb up 0 Thumb down 0
  101. Kate says:

    I know for a fact that many clothing manufacturers do not refit the outfit for larger sizes – they just have the factory make everything bigger – you get an odd fitting outfit with huge neck and armholes long arms and the waist still fits too tightly.

    Oversize people need garments fitted for their body types which aren’t simply bigger all over. They need fashions that disguise a large stomach or buttocks. Something that looks good on a medium person won’t look good on someone with rolls of fat here and there.

    When I go to the outlets, the only large sizes are found on hideous clothing that no one in their right mind would wear – the same with large shoes – I need classy well fitting pumps, not zebra stripe things with bizzarre narrow points that stick out 3 inches past your pants.

    So it’s about shapes not all fitting the same design, and buyers not bothering to take the simple step of having a large (or small) size actual person trying on the clothes in those sizes to see if they still are well fitting.

    Thumb up 0 Thumb down 0
  102. Jen says:

    I say it’s fashion and fit, pure and simple.

    Good fashion and good fit are all about body proportions. Most garments are designed around and “ideal” size 6 body form, and sized up and down from there.

    If you are considerably smaller or larger than the orginal size 6, the garment will not hang as well on your body. Our shoulders do not grow as wide with weight gain as our mid-sections, yet the garments do.

    Large and small sizes reduced or enlarged from a medium simply do not fit as well on small or a large frame. Buying stock in equal proportion to people’s frame type does not take fit into account, so these items end up on the sale rack.

    Thumb up 0 Thumb down 0
  103. L says:

    It seems as though most of the commentary offers perspective on the demand side, but very few have addressed the fundamental question relating to the supply side: why the difficulty in matching supply to demand for the more extreme sizes, especially when the market dictates that these sizes are more likely to sell at a discount? Kevin’s explanation, above, does seem a plausible alternative to that posed by the author…

    My other question, as a runner, is why do all of the races seem to run out of s/m/l sized t-shirts so quickly, when runners tend to be average sized people (if not smaller)? I mean, many of them even ask your t-shirt size when you register, so it’s not necessarily a lack of information.

    Thumb up 0 Thumb down 0
  104. Mike M says:

    #52 has a very valid point (more anecdotal evidence to follow). I am in a similar situation to GW, as I am taller (around 6’3) but not fat (built like a linebacker).

    I am not willing to pay $75-$120 for a shirt that does not fit me very well, which includes nearly all clothing at traditional retailers (just try finding an 18.5″ neck with the required sleeve length).

    I am much more likely to purchase a nice shirt that ‘mostly fits’ for $30-$50 realizing that there will be limitations to its use, such as not being able to wear a tie, or haivng to roll up the sleeves. I won’t fit any better in the clothing, but I feel a lot better about the purchase.

    Rather than making shirts that fit each body type in the XL/XXL size range, it’s probably easier to discount the product and hope the customer tailors it, or is otherwised satisfied at a lower price point.

    Thumb up 0 Thumb down 0
  105. dia says:

    I’ve heard that it’s because retailers get the clothes for less if they buy the same number of items in every size. Naturally the mediums and larges sell out first, leaving the extra-smalls, smalls, and XLs. As a size medium I find it extremely annoying.

    Thumb up 0 Thumb down 0
  106. Lola says:

    Perhaps your premise is wrong. The truth may be that the regular sizes may sell out quicker in the discount stores and what is left are the odd sizes, thus, the XXL or XS.
    It just seems to you that there are always a disproportionate amount simply because these are the ones that don’t sell or sell slowly.


    Thumb up 0 Thumb down 0
  107. Stuart Eugene Thiel says:

    Suppose XX people buy size M clothes when the price is low merely to indulge their optimism or are in denial. That’s not enough of an answer, because manufacturers and retailers should have learned to take that into account.

    The explanation is more likely to be on the supply side — economies of scope and scale, or on a simple tale of self-fulfilling expectations.

    Let’s suppose that the proportion of size-M and size-XX in the shipment to the discount store is the same as it was in the original shipment to retailers. These clothes simple don’t hew to popular taste. But there’s a range: some clothes just aren’t quite right, and some are, as George Harrison might have said, simply “grotty — you know, grotesque.”

    Now. If it is generally believed that size-M clothes are disproportionally scarce in discount stores, size-M customers will be less finicky — as my mother used to say, beggars can’t be choosers — and more likely to grab the least grotty bargain clothes because they have no confidence that there will be a similar bargain next time. This exacerbates the shortage of small-grotty-factor size-M clothes, making it yet more likely that the next size-M customer will grab the next least grotty size-M bargain clothes. . .

    And I don’t see where anyone on the supply side has an incentive to correct it. It may be a prisoner’s dilemma, sort of like when everybody stands up at a football stadium.

    Thumb up 0 Thumb down 0
  108. daz says:

    “My best guess is that for some reason small and large people are, in general, less willing to pay for clothes.”

    I don’t think so, at least not for large sizes. I’m a large guy (6’2″ and over 250 lbs.), and I would like very much to dress in the style of my choice . . . but regular clothing or department stores rarely carry them.

    They do carry a *few* that fit me, but the selection is so teensy that . . .

    . . . short of paying for custom-tailored clothing, which I can’t afford on a regular basis, I have to go to the “big & tall”-type stores, which almost always charge a premium due to little competition, and have a bigger selection that other stores.

    Here’s my guess about why many ordinary stores end up with large & small sizes, but no mediums: They deliberately buy more extreme sizes than they will need, since if an XXL-sized customer repeatedly finds next to nothing of their size — when there are plenty of the medium sizes — they will stop shopping there. (Only some clothing styles even *have* XXL sizes.) The stores don’t have to worry about medium-sized customers running out of choices since all their styles come in medium.

    Thumb up 0 Thumb down 0
  109. Erik says:

    I think is has to do with costs. A store that does not have your size in stock incurs two losses. One, the loss of the income from the sale. Two, the potential loss of a customs if the perception is that a given store caters to average sized customers. Te second cost is real “Oh, they never have my size”. Nobody that is a size M would really think that way. But an XXL likely would. Therefore is makes sense that a store would buy extra small and large sizes to avoid the second cost.

    Thumb up 0 Thumb down 0
  110. Pepe says:

    Because America is STRESSED OUT, we don’t know what to do with ourselves anymore!!!

    Thumb up 0 Thumb down 0
  111. Doug Nelson says:

    People with low self-esteem are less willing to spend full-price on clothes. The “discount” market for clothes is actually the primary market for people that have been marginalized by Hollywood, PE class, etc. Therefore the “discount” market has responded correctly by increasing their inventories in these sizes.

    Thumb up 0 Thumb down 0
  112. RCE says:

    “My best guess is that for some reason small and large people are, in general, less willing to pay for clothes.”

    Tell that to the thousands of tiny, fashionable, petite Asian women/clotheshorses.

    I don’t shop American stores anymore because everything is vanity sized. What used to be a 2 is now a “0,” and what used to be 0’s are nonexistent.

    If stores actually carry xs sizes, that would be news to me.

    Thumb up 0 Thumb down 0
  113. Chris H. says:

    Perhaps it’s because the standard sizes sell more quickly. Since there is a limited supply at discount stores once the most popular sizes are gone, they are gone. This leaves only the XS and XL sizes left. It is possible that the same amount of medium items arrive at the store as other sizes.

    Thumb up 0 Thumb down 0
  114. AndreaS says:

    We should consider the volume of XS and XXL leftovers of high-end market clothes and of simpler, humble clothes. It could give a measure to the assumption that “sharp clothes won’t help them that much.”

    Thumb up 0 Thumb down 0
  115. CJT says:

    There are two different points on the demand curve at work. The original point (full price) is set at a level where supply meets demand. When the price is cut to sale price (half price) two things have happened: (1) Supply has dried up (the skew is no longer being produced) and (2) Demand at the half sale price is much greater than the remaining supply when customers actually observe the price. This is the critical observation – the change of the sale price being observed by any particular size is far larger for the center of the bell curve than the outside fringes. Thus, thus the medium sizes are all swooped up. Because there were far more of the medium sizes to start with, the suppliers are content that most of their inventory was cleared up. It’s not the price of the product that stops the XS and XXXL sizes from moving, it’s the advertisement of the sale and the overall volume of foot traffic that is insufficient to capture the rare long tale people.

    Thumb up 0 Thumb down 0
  116. Ana says:

    Are there really that many people in the middle? Maybe there are more XS and XL clothes on discount because manufacturers make more of them? Sizing has changed so much in the US (see the picture) that many many women who are just at a normal, healthy weight are shopping in the S/XS end of the clothing rack with the genuinely skinny. Due to the shocking levels of overweight and obesity in the US, many many women are shopping at the XL end of the scale. Maybe the bell curve has inverted and discount stores are just getting the expected distribution of clothes.

    Thumb up 0 Thumb down 0
  117. Dan says:

    I think we are over thinking this. If a store has 10 medium shirts and 10 XXL shirts left over the 10 discounted medium shirts should sell out reletively quickly because a signifigantly greater percentage of customers in the outlet are that size. This will leave us with an abundence of XXL shirts despite initial inventories being equal. It is interesting to note that XXS and XXL people must know that they can not only find a deal at the outlets but also have a greater selection than a typical store that intentionally keeps their size in low quantites. This should lead to abnormaly sized people seaking outlets at a greater rate than normally sized people.

    Thumb up 0 Thumb down 0
  118. Corinne says:

    Interesting variation: When I was in college my family lived in Hawaii, so my breaks from school I would visit Hawaii and scour the city for prices that were friendly to my college-size budget. There I found it was very rare to find a small or an extra small size at a discount, instead there were plenty of medium, large and extra large.
    Hawaii has a large asian poplulation and a lot of asian tourists, and the girls seem to be smaller than mainland girls.
    Which makes me think that at retail chains are not doing enough demographic store planning.

    Thumb up 0 Thumb down 0
  119. Ken says:

    Being in the garment industry in the past too, I can absolutely agree with my fellow “garmentos” and others above. Since this is an economics blog I think I can add value here with statistics! Apparently our government has published size statistics on our soldiers. While it (military) is not the best representative group, it shows the normal distribution with 95% of the group being within 2 standard deviations from average. Unless the product is an uniform in high volume for everyone, I’m sure someone here can estimate the (high?) probability that the extreme sizes are left over.

    Thumb up 0 Thumb down 0
  120. eric says:

    It may be that there ARE more medium sizes at these outlets – at first. But, because there are more medium-sized buyers, they sell out faster than the extreme sizes. Say 60% of the inventory is midrange in size, and 20% is xsmall and 20% is xlarge. The midranges sell out quickly, leaving the impression that xs and xl were more numerous.


    Thumb up 0 Thumb down 0
  121. T-Bone says:

    Variance due to volume of sales seems to be it. Medium and large sizes are a more fluid, reliable market. XL and small probably sell out a lot for some items and go unsold for others.

    Imagine trying to perfectly estimate how much of each food type to serve for a buffet. During lunch hours, a lot of people show up. This is like medium and large clothing sizes (has a lot of customers). The best you can do is to estimate the popularity of items and put them out there in those amounts. Some items may sell out. Most of the less popular items will at least be eaten by someone. The proportion of leftovers to the total amount of food served will be pretty low.

    But during off-peak hours, you’ve got less customers. This is like XL and S clothes. You serve less food, but in the same proportions. But now you’re more likely to have some items go completely uneaten, and some items to be unexpectedly popular where you don’t even come close to meeting demand. A much higher proportion of leftovers will remain. (and leftovers represent what goes to clearance in the analogy to clothes)

    The margin of error increases as the sample size decreases.

    Thumb up 0 Thumb down 0
  122. MGB says:

    Two reasons:
    1. Manufacturers and distributers maximize their profits by selling clothing “runs” that include certain numbers of each size. Making a run that has too few of the sizes that will sell out quickest is a SMART business decision for them because it means stores will have to re-order more product than if they had more flexibility in choosing the quantities of sizes they wanted. This is also why the higher-end (and much more marked up) designer clothing doesn’t have this same problem– their profit doesn’t come from volume to as great an extent.

    2. Additionally, very large and very small clothing probably sells out at widely vacillating rates, depending on the style and cut– certain styles just don’t fit people at the extremes of the size scale very well, and this is hard to predict, because there are many different styles and fabrics and these factors interact in many different ways. Their popularity at the high and low end of the size spectrum also probably varies a lot by geography. So this is why stores don’t balk too much at the lack of choice they have in ordering specific sizes in most clothing ranges– I’m guessing that at least half the time, probably more, the clothes DO sell in numbers that are proportional to supply.

    Thumb up 0 Thumb down 0
  123. Dan says:

    My guess is it has to do with the distribution of shirt sizes in the stock that the discount store gets from the original store, not something about wealth or size. Here’s an example of how this works:

    A store’s goal is to have as few “left over” items at the end of the season (i.e. sell as many as possible at full price without leaving a huge inventory of unsold items). So if it does a good job ordering the correct numbers in each size, there will be a small number of each size of a particular item left.

    This means that while a store may order 1000 in size L and only 250 in XL of a particular item, if its done a good job it’ll be left with ~25 left in each size. The store’s goal is not for its stock to sell well proportionally (aka having 100 L’s and 25 XL’s left), it’s looking for them to sell well in an absolute sense (as few as possible in each category).

    This means that the discount store will generally receive a similar number of the item at each size. In our example, there will be 25 L’s and 25 XL’s for sale. Because L’s sell at a faster rate than XL’s, you’ll end up with only the larger size left on the rack.

    Thumb up 0 Thumb down 0
  124. Brad says:

    Another theory
    Shops need to have some stock at each size but as the number of stock required gets higher the % oversupply drops as you can’t have half a dress on a rack.
    These overstocked items are then transferred to discount stores but now the number of medium sized clothes is closer to the number of large and small sizes than the original stocking requirements.
    At this point the medium sizes are run out quickly and only the smaller and larger sizes remain

    — Posted by Richard”

    This explanation makes the most sense to me.

    Thumb up 0 Thumb down 0
  125. CJ says:

    All interesting observations. But none reflect the process of retailing. A full-price store acquires a batch of identical garments according to a normal distribution of sizes and offers them on the floor for full price, say $100 each. The maximum profit is squeezed out this way. Once about half are sold, the full-price store sells the remainder lot to downstream discounters like TJ Maxx, where they are marked down and resold. TJX keeps discounting them until a rock bottom like $2 is reached. As the price declines, along the way, normal size people are more likely to snap up the bargains. Large size people are fewer, so the residual unsold inventory listed at falling prices continuously shifts towards unsold large sizes. Ask anyone in this food chain of retailing and they will verify. What is happening is that the top-dollar buyers at full price (including the few large size buyers at full price) are subsidizing the bargain buyers at falling prices, with the large size skew in the unsold inventory subsidizing large size discount buyers the most.

    What if the original batch had fewer large size garments? That problem is same as the newsstand problem: too few = lost sales & lost profit, too many = unsold inventory & decreased profit.

    Thumb up 0 Thumb down 0
  126. Joel says:

    It’s the same reason we are having a financial crisis. Market liquidity. There are fewer XXL XXS items than M that go on sale, but the XXL and XXS sizes are an illiquid market, so they don’t respond to price decreases like you’d expect.

    Thumb up 0 Thumb down 0
  127. Edward says:

    If the shop stocks “a few extra” of each size to make sure they don’t run out during the regular sales period, that will leave a proportionally greater number left over when they clean up the inventory and discount what’s left.
    Say a shop expects to sell 15 L shirts and 3 XS shirts in the regular sales period. To make sure they don’t run out, the shop stocks 3 extra of each size i.e. 18 L shirts and 6 XS shirts. Now the shop carries 20% more L shirts than it expects to sell, but 100% more XS shirts. Relative to the size of the market, there will be more leftover XS shirts when the regular sales period and the remaining stock is discounted. The L shirts sell out early in the discount period, while some excess XS shirts remain until the end of the discount period.

    Thumb up 0 Thumb down 0
  128. alan mushnick says:

    Let’s assume that people of all sizes buy clothes at the same rate or frequency. Then the reason there are leftovers in the really small or really large sizes is due to the clothing sizes not matching the realities of the sizing models that the stores are stocking. Meaning, there are many more people in the middle size ranges than predicted.
    If in fact, the distribution of clothes sizes exactly matches the distribution of people sizes, other theories would come into play. I believe it is the former, and this should be relatively easy to collect the data.

    Thumb up 0 Thumb down 0
  129. Jay says:

    The answer, as other posters have indicated, must have to do with higher (unpredictable) demand variance in these outlier sizes.

    Assuming that retailers study how garments have sold in the past (a reasonable assumption), they will know that the average garment will sell 2x (or 3x or 4x) more in medium than, say, extra-small. What they don’t know is how well a particular garment will adhere to this average distribution; they only know that across all garments it is the average distribution of sizes sold.

    The article references the phenomenon of many XL and XS sizes in discount outlets. If these outlets had a policy of blasting a crater in the floor for every garment that quickly sold out in these outlier sizes, discount shoppers would have a different experience: they would notice the many garments that are oversupplied in XL and XS, and they would notice a lot of craters in the floor of the store!!!

    Thumb up 0 Thumb down 0
  130. Anthony says:

    I still don’t see how the proposed explanations really explain anything. If very small and very large people aren’t inclined to buy nice clothes at high prices, for example, (or any of the other explanations) the buyers should have figured that out by now. Could it be that it would be too much trouble and effort for sellers to worry about it? Could they already be making such a profit, that they don’t care about a fine point like stocking fewer of the extreme sizes? and what happens to those items that no one wants to buy after the sale period is over? a handsome income tax deduction for the seller? Why in these times does anyone still think that naive economic theories work?

    Thumb up 0 Thumb down 0
  131. John Feiner says:

    I looked through many of these comments, and may have missed it if someone in the last page came up with the answer. The answer is: it is largely not true. This is a case of recall bias. At 6’4″, 215 lb, size 14 feet and XXL gloves, there are not more XXL cloths/gloves/shoes left at sales. Nordstrom’s is the only store that actually stocks large sizes (bias created in using the example of Nordstroms to begin with, instead of considering whether this is true for most stores). People are only remembering the times when they did see leftover odd sizes, and not remembering the times where there were none (Recall bias forced by the very nature of this question). If I don’t get to an REI sale early, there is nothing left in in XL or XXL. I have to order size 14 shoes.

    Don’t start by assuming the assumptions of the question itself.

    Thumb up 0 Thumb down 0
  132. Jacob says:

    My best guess – for what it’s worth – is that when you go to the extremes of XS or XXL, the chances of finding a true fit are much lower. Most people, regardless of fit, will find a right-sized garment plus/minus one size. With the bookend sizes, you only have one way to go if the fit is not just right. And psychologically, if I buy a MED, I’ll feel more normal-sized, right?

    Thumb up 0 Thumb down 0
  133. Kevin says:

    As an XXL myself I would disagree with the statement altogether. I just think whenever you look on a discount rack, they will always have every size but the one you are looking for.

    Thumb up 0 Thumb down 0
  134. cv says:

    Another factor:

    I generally wear a small, but depending on the store, clothing line, or specific item I sometimes wear an XS or a M. If you assume that type of variance happens for everyone, then the people on the ends of the size range at a particular store will be less likely to find something in their size, since the XS people don’t have the option of a smaller size, and the XL people don’t have the option of a larger size. I would imagine that XS and XL people would then shop less at that store, preferring stores that more reliably have their size. The store still needs to stock those sizes, though, in order to reliably have the correct sizes for the S and L people who occasionally wear an XS or XL.

    In other words, people on the end of the size range offered by a store won’t shop there, and it’s an unavoidable problem. This reduces the demand for the already less popular sizes, amplifying the effects of some of the variance and other supply issues that others have mentioned.

    Thumb up 0 Thumb down 0
  135. Matt P says:

    Sizes are ordered categories, not independent categories like many of these arguments assume. If the price is right, a medium item might be purchased by a large or small person if the fit is close enough. This amount of overlap is not possible in the more extreme sizes XS and XXL. So if the purchasing population was perfectly modeled given the original price, at a discount price the middle size will be more quickly depleted.

    Thumb up 0 Thumb down 0
  136. Geronimo says:

    Let’s call this the “24 Per Store Theory.” Assume that no store wants to permanently alienate any customer by having a horribly reduced selection in his / her size. Say this season’s hot shirt comes in four colors. To give every size customer the same choice as the others (one of each color, with no duplication) you need 24 shirts, from XS to XXL.

    Now, in come the customers. The middle of the size range has a huge customer base (a census table showed 50% of US males between 5’7” and 5’11”, which are roughly the non-X sizes). The XS, XL and XXL sizes have much less numerous populations—all heights under 5’5” for males are less than 7% of the population; all over 6’1” are less than 6%. Stores are forced to take huge overstock hits, just by offering the same choices to one-tenth as many “X size” customers as “regular size” customers. It’s still too many “X size” customers to risk alienating, so you gotta do it. You can discount these assumptions to handle other variables in the process, but you still end up with one-tenth the customers for the X sizes, so only the absolute most popular stuff sells out to zero in all sizes.

    Thumb up 0 Thumb down 0
  137. Pierce Randall says:

    Fashion is difficult to predict, as well. I’m a medium. In high school, I might have worn large because baggier clothes were in style in the 90’s. That used to be a grungy thing; today, it’s still around, and it’s hip-hop. Different kids wear different things.

    Speaking in more broad market terms, fat people at least may be intimidated by buying clothes. It’s frustrating to convince yourself that it would look good on you, then find out that it’s not in your size. You might walk out and leave if you’re marginal for buying clothes that day anyway. I bet a $12 t-shirt costs $2 to make, ship, and distribute. That’s 6 shirts before you cut into the profit margin.

    Also, people of average size see something in very large or very small sizes. That generates interest, then it won’t fit them. There’s scarcity. Maybe the next time they get more of that shirt or those pants, they don’t have to put them on sale.

    Since you aren’t in the clothes distribution industry, or you would know already and not pose the question, do you know it’s always very large and very small sizes? If you’re writing for the Times, I bet you live in New York. Land use patterns dictate that per capita, New Yorkers aren’t that obese. Compared, at least, to people who live upstate, drive everywhere, and lead less active lives. Maybe in Binghampton (no knock on Binghampton!), weight distibution is more like the national average clothes stores count on.

    Thumb up 0 Thumb down 0
  138. prklypr says:

    I worked for a number of years in a high end store with a huge men’s dept, and I was always amazed that the stores in the Northeast always ran out of smaller men’s suit sizes, while the stores in the Southeast (esp FL, where there are a lot of retirees) always ran out of the larger sizes. So the depts would “trade” suits, sending larger sizes south and smaller sizes north. Maybe stores need to look at population demographics before they place orders.

    Thumb up 0 Thumb down 0
  139. htb says:

    For Ralph in #63, and others:

    If you’re looking solely at adults, then M-L-XL might be a reasonable guess at the size distribution. However, teenagers wear clothes, too, and they account for much of the XS-S size sales. Girls’ size 12 is not very different from a ladies’ size 5.

    Thumb up 0 Thumb down 0
  140. Margaret says:

    The theory doesn’t hold for smaller people. (Please define the “normal realm of beauty”).

    The problem with mainstream xs is that it simply isn’t cut for a fit, petite person. Being such, I would much rather pay over standard price for an article that actually fits than a discount for something that looks as if it were fitted in middle school home ec.

    The people with the proportions to actually fit into a standard xs would tend to be short and overwieght. This correlates with the tendency of low income consumers (who shop at discount stores) to be overweight.

    Thumb up 0 Thumb down 0
  141. Summer K. says:

    As an XL woman (who buys clothes full-price at Nordstrom when they fit well), I’ll tell you why: the clothes left over are ugly and ill-fitting on my frame. What usually happens is that the clothing is cut to flatter an idealized small or medium frame and then just sized up and down for everyone else. So I try on something in the store, and it is too loose or snug in the wrong places, or it just hangs wrong or creates the wrong silhouette for my body type. Same for the very small sizes that are far more commonly deeply discounted. My mother is an XS, and many shirts hang on her or have disproportionately small or large sleeves.

    The other major factor is that older people tend to occupy the very large and very small sizes (think matronly figures and “little old ladies”), but buyers often indiscriminately purchase trendy clothes in all sizes. Older people tend to be less concerned about trends (like jeans with butt art or Mrs. Roper-type psychedelic tunics), especially when they are meant to reflect a “retro” area well within the older person’s memory. Those trendy clothes are not snatched up by the older persons who tend to gravitate to the far ends of the size spectrum, so the clothes wind up discounted for the young very thin or fat folks.

    Thumb up 0 Thumb down 0
  142. mike97 says:

    Well, one problem with this analysis is that there is NO STNADARD XL or XXL

    Idfyou shop on the West Coast, our XL is far larger than the typical XL on the East Coast. When I lived in Orlando, I could never have worn a standard XL, it was too small. But here in the West, I can even sometimes where a L.

    I think there is a general phenomena of smaller sizes on the East Coast and larger on the Wast. I ask buyers at Eddie Bauer and Nordstrum’s about this, and they confirmed that there is a regional variation in the standard “XL-XS” sizing.

    Thumb up 0 Thumb down 0
  143. Cynthia says:

    At least for why there is an overabundance of smaller sizes, much more prevalent in high end brands, the sellers sell only in a package that carries small sizes disproportionate to the percent of people who would fit in them in order to project the idea that their brand is for skinny, attractive people.

    Thumb up 0 Thumb down 0
  144. C Sarles says:

    Well, every store needs to have at least one size of each garment they stock. Given that the demand for plus sizes in designer clothing is so much lower than that for medium or large, it is more likely that stores will be left with surplus plus sizes.

    Also, not everyone is equally likely to buy designer clothing. Many people who fit XL and beyond have extra baggage around the waist and hips, and designer clothing will not look right on them. Designer stuff is made for thin people whose bodies resemble coathangers. Fat people are more likely to wear non-designer clothing that is closer to a potato sack than an Armani suit.

    Thumb up 0 Thumb down 0
  145. Dave says:

    I have small feet (6.5-7) when I go to discount shoe stores and see very few shoes that fit I find maybe 5 or 10 pairs out of 1000s that fit me, in the whole store. I think the observation of lots of odd smaller sizes is highly inaccuarate observation, at least for footwear.

    Thumb up 0 Thumb down 0
  146. Brian Dunstan says:

    #8 seems most likely.

    Regardless of the effort of the store to adjust to their customers, low-volume items will always face more uncertainty and thus greater overstock; attempts to prevent backorder will increase the amount of overstock.

    I suspect that income reinforces this: on average in the USA, younger affluent people tend to in the moderate size ranges. The type of person who will walk into a department store and buy a lot of things at full price is most likely a S, M, or L.

    Also: stores do notice what size their customers are. For instance when I was in Bangkok, western-style department stores contained clothing, labeled XL, which would correspond to a size S in the USA. The only way I could find anything which fit me was to look in the discount rack.

    Thumb up 0 Thumb down 0
  147. Chaitanya Netkalappa says:

    So, how about the hypothesis that they just aren’t very efficient at ordering the correct amounts?

    Also, I’d like to see the data as to the disproportionally greater number of S, XL’s in discount stores for whenever ive gone there it always seems like ive found M’s in whatever I wanted. (I’m an M)

    Maybe its a variance issue..

    Maybe the store managers keep rotating and fresh store managers don’t understand historical data for their new store very well and over order some sizes.

    Maybe they pity XL’s and order it (tongue-in-cheek)

    Maybe there is a behavioral impulse to overorder the smallest quantity. (I actually think this is my best idea). I would think that if historical data suggests that you should by 200 M’s and 30 XL’s, you would irrationally assume that maybe the 200 is too large and the 30 is too small and try and compensate for it.

    p.s. If some behavioural economist at the Univ. of Chicago likes my last explanation I’d love to work for them, my email id being my first name at uchicago dot edu 😉

    Thumb up 0 Thumb down 0
  148. Sisi says:

    I’m XS and my husband is M/L. I can rarely find anything for him at an outlet or discount store, since he not only wears the most common sizes, he only likes the most popular colors (call him Average White Guy). If I’m hell-bent on having a specific item of clothing, I must buy it at full price at the beginning of the season, because Nordy’s et al. will only have one in my size and will not reorder it, because there are not many people in my size. If my motivation is so-so or non-specific (e.g., “wool skirt”) I can wait for it to go to clearance or even the outlet. This actually happened with a pair of $400 shoes that came to my local Rack for $100. Sometimes my size will come back as a return, probably because the original buyer wasn’t able to fit into it comfortably after all. Sorry, but it’s true that some people will buy smaller clothing as a motivator for getting fit, and then not make their goals after all.

    I do very well at outlets, etc. when I travel to other parts of the country, and even more so at vintage and resale stores. I am short, not BMI-impaired, so I find older clothing tends to fit better. Leaving the West Coast helps because I’m not competing with as many other Asians. Macy’s in Chicago has to stock my size as it does in San Jose, but the odds of there being anyone else my size over the age of 12 are low, so I score. And if I don’t buy it, it will probably go to a discounter on either coast. And since the wholesale price of garments is a lot less than non-industry people expect, this is not entirely without profit. When there really is a loss, as with better petites, production has shrunk.

    Thumb up 0 Thumb down 0
  149. Geoffrey Wiseman says:

    Long comment thread, I’m not going to read it all. If it were simply a matter of who was willing to pay for the clothes, then, again, the manufacturers should have figured out by now that they’re making clothes that people aren’t buying.

    I think it has more to do with retail distribution. There are fewer XS and XXL people out there, but in order to ensure that they have the opportunity to sell these clothes, all stores stock clothes in less common sizes, knowing that not all of them will be sold, but that some of them will — in essence, if you could ensure that XS and XXL people went to a specific subset of retail locations, you could stock only enough clothes for those customers, but the distribution of customers isn’t predictable enough.

    Thumb up 0 Thumb down 0
  150. Tkwon CMS says:

    My guess is that companies perhaps forgot to take a basic college statistics course? Maybe they made as many XXL/XS shirts as M or L shirts?
    My guess is that companies miscalculated the bell-curve and routinely oversupply XXL/XS clothes.

    Thumb up 0 Thumb down 0
  151. walter says:

    Implausable, becuase clothiers should know that small and large people don’t buy as much. But what if those sizes were the hardest to fit–while all mediums might be about the same size, and thus predictably easy to fit–small and large people on the distribution vary a lot more and thus it’s less likely that the sizes that have to deal with the extremes of the distribution fit as well. imagine a distriubtion from 0 – 100 as people’s true distribution of size, let’s say it’s normal. a person who is a fifty, wears a medium, there are more of them than anyone else. people who are 49 and 51 don’t fit as well in a medium, but it’s pretty good, and clearly not efficient for clothiers to divide the mediums into ‘small-medium’ and ‘large-medium’.
    let’s say Small and Large target 35 and 65, substantially different from 50, and well populated enough in the distribution that a lot of people will want this size to exist.
    But as you get toward the tail of the distribution, where XXS and XXL are people anywhere from 1 – 20 might be xs and 80-100 might be xl. no surprise if these don’t fit, it’s targeted to a much more diverse (if smaller) distribution.

    I think this is a much more reasonable explanation.

    Thumb up 0 Thumb down 0
  152. Schroeder says:

    I haven’t read every comment, but enough to see that everyone is pretty much missing the point: It is not hard for a retailer or manufacturer to run extremely accurate statistics for each individual store. You’re not giving these companies and corporations enough credit….

    Thumb up 0 Thumb down 0
  153. Joe Tillery says:

    I think that most clothing is bought in lots and the manuf. have weak or bad demographic departments. Buyers probably are never alerted to this mismatch. I doubt that buyers order by size, just lots. Since size variation in Ca. probably varies from Tx. a general, badly researched, size distribution results. Levi’s once stopped selling in lots smaller than 10,000 and “standard distribution” of size is probably not changeable on orders. I’ll bet they don’t vary what they ship to El Paso or Iowa. A big guy in one place would have a large selection and the other none. My son and I are long and tall, he is big also. We order his shirts, 18 1/2, 37/38!

    Thumb up 0 Thumb down 0
  154. Mitchforth says:

    I suspect that the reason stores frequently order more of XS and XXL than they can sell at full price is that it is not inefficient to do so.

    This probably has to do with the fact that there is often a very sizable mark-up from wholesale to retail price on designer clothing.

    As a result, of the high margins, the cost of losing a full-price sale is substantially higher than the cost of discounting and selling the excess stock at the outlet, so the retailer stocks plenty of those sizes, bearing the high risk of having excess stock to discount, in order to eliminate the risk of selling out while there is excess demand for the product.

    Thumb up 0 Thumb down 0
  155. Susanna K. says:

    If I know that my size is likely to go on sale soon, why would I pay full price for it?

    Thumb up 0 Thumb down 0
  156. Dale says:

    Perhaps its because demand for extra big and extra small sizes are harder to predict, therefore there’s more variability, and therefore more safety stock is required.

    Thumb up 0 Thumb down 0
  157. Steve says:

    I’ve never found this to be the case. Acres of sizes for average people and 3 ugly shirts that fit me!

    Thumb up 0 Thumb down 0
  158. Matt H says:

    It’s a problem of scale at the extreme ends of the curve. The malls are filled with clothes that fit the middle 98% of the population. I’m at the 99.9th percentile for height, and I hate shopping. It’s very rare that I go to the mall and walk out with what I wanted, in my size. Shopping is an exercise in frustration, so I do it only once or twice a year in a couple of concentrated big trips.

    For the smallest and largest people in the country, I’d wager their experience is similar. There is a smaller selection of the clothes we want, spread thinner and harder to find. Finding the clothes we like, that fit, at the store when we go there, is a low-probability outcome — so we do it less frequently.

    Thumb up 0 Thumb down 0
  159. fat guys in party pants says:

    Is there any actual evidence that there are more really small and really big garments left over for discounting? I’m really big and am (seemingly) always disappointed that there’s nothing left for little old me when it comes clearance time. When what I want is there, I buy it and don’t think about it. When there isn’t, I momentarily suspect an anti-fat guy conspiracy. Could all of you medium-sized people just be suffering the same kind of confirmation bias?

    Thumb up 0 Thumb down 0
  160. Alex says:

    Actually, I think the explanation is related to stocking and ordering. I was the merchandiser in the clothing department of a department store for a while, and I was only able to order most brands in sets of 12 or more. Even if we were only be projected to sell four in the XXXXL size of a given style, I still had to order 12 (corporate wouldn’t allow us to go without stocking every size.) If we needed 19, I either had to order 24, or lose 7 sales.

    Warehouses filling wholesale orders generally don’t just pick single items, they pick premade bundles. It saves them massive amounts of money on labor.

    Thumb up 0 Thumb down 0
  161. Laura says:

    … do we have any data to suggest we’re looking at a real phenomenon, or is it just someone’s impression that discount stores are full of XXLs and XSs?

    When I’m shopping, it always seems like stores NEVER have the clothes I want in my size. It’s easy to chalk that up to poor planning by store buyers, too many people in my size, vast right-wing conspiracy, etc. However, this has been a consistent experience, despite the fact that my size has fluctuated. The only logical conclusion? I notice when a store doesn’t have my size; I don’t notice when they’re missing another.

    There may be something similar at play here. Since S-L is the average size, most people shopping at discount stores will wear these sizes. So they will be more apt to notice when these sizes are missing, and not when XSs and XLs are missing. The perception that the “average” size is missing might simply reflect the fact that there are more average-sized people to notice when the sizes are not available.

    Thumb up 0 Thumb down 0
  162. Chris Whatley says:

    OK, probably nobody will read this since it is comment 158+, but my sense of the reason this happens is that when you are at the small or large ends of the scale, you are less likely to fit the model body used to make the garment patterns.

    Thumb up 0 Thumb down 0
  163. John Jay says:

    Years ago I did a consulting assignment for a shoe company that sold to KMart. KMart ordered a statistical distribution of shoe sizes that followed the demographic projection (men, women) as well as actual results that profiled a given bell curve for each geography.

    It was an inexpensive and effective way to eliminate the ‘order by size’ calculation from having to be estimated by each Buyer in each Department for each Geography. I wonder why this is not done in clothing? Or maybe it is?

    Thumb up 0 Thumb down 0
  164. rcs says:

    would be easier to have sizes that reflect real measures (small = X number of inches and so on) it’d be way easier to buy stuff online (like tees where you buy for the colors and graphic, instead of fit), i understand the japanese already do this

    if i ran an online “big and tall” i would put BMI as a serach option in the catalog, i used to be fat and after a while XXL or XXXL stop making sense, my BMI was way better in predicting how clothes would fit me

    anyone has a better measure system?

    Thumb up 0 Thumb down 0
  165. keith twombley says:

    People who are very large or very small may find the process of shopping frustrating and thus not do it as often as people who are middle-sized.

    As a large person it’s very frustrating to find something that I like and then not be able to find it in my size. My ex was a very small person and she found the prospect of shopping in the pre-teen section unappealing for obvious reasons.

    Thumb up 0 Thumb down 0
  166. Devin says:

    I don’t believe this explanation. If you could actually establish that people with physical dimensions outside the fat part of the bell curve spent less on clothing, then sure, you’ve got something to work with. But I don’t actually think that a woman who is smaller than the average is at a disadvantage in physical appearance, certainly not to the point that she might as well not bother spending money on clothes. The same goes for a particularly tall man. So you could also look, and if you found lots of XS and XXL menswear, and lots of XXL women’s clothing, but relatively little XS women’s, that might lend credence to the theory as well.

    A more likely explanation is that much off the rack clothing is sized to fit well in S, M, and L sizes. It’s likely that the XS and XL sizes are just scaled versions of whatever the base size was, and might not fit as well. So it may be that although stores order each size in proportion to the amount of clothing they sell in that size, XXL and XS shoppers must try on and reject more clothing than M or L shoppers before they find something that fits them well.

    Thumb up 0 Thumb down 0
  167. Tigger says:

    I don’t know if this is true in US. Here is a hypothesis for why there are lots of XS clothes available in discount stores in UK.

    The standard purchase tax (VAT) that is paid on most things in UK including adult clothes, is not levied on children’s clothes. So, identical items could have different prices, depending on whether they are sold as adult extra small or larger (older) child sizes.

    I once encountered a useful offer in UK. A shoe shop offering BOGOF

    Thumb up 0 Thumb down 0
  168. Phil says:

    Forgive me if this has already been posted — I haven’t read through all the responses.

    Suppose half the time the store doesn’t buy enough XS/XXL, and half the time it buys too many. When it doesn’t buy enough, you don’t notice, and the “normal” sizes don’t go to discount because they can still sell at regular price.

    But when they buy too many, they’re shipped off to the discount stores, and you notice.

    Thumb up 0 Thumb down 0
  169. a "normal" sized man says:

    Society accepts that Big & Tall stores are fine…that they fill a need. But if someone had a “Medium and Normal” clothing store, people would picket it saying it was prejudiced.

    Thumb up 0 Thumb down 0
  170. Stat says:

    Are you sure that this is actually the case?

    Thumb up 0 Thumb down 0
  171. BSK says:


    Great point. We notice the times when stores have excess XS and XL. But how many items do they not have ANY extra in? For every shirt there are extras of in certain sizes, there are probably 10 that did not have excess inventory. We just don’t notice it because they’re simply not there.

    I’m not even going to tackle all of the horrible assumptions people are making about people based on their body-type.

    Thumb up 0 Thumb down 0
  172. J Wynia says:

    I’m just trying to figure out where these amazing discount stores and sales full of XXL are. I’m 6’4″ and 250 and have needed XXL, XXLT or XLT my entire adult life (and will even when I finish losing weight.

    I have been in Nordstrom Rack numerous times here in Minneapolis and have NEVER seen an abundance of XXL. At this point, I actually just buy all of my clothes online because NONE of the stores except for the Big/Tall specialty stores has more than a handful of things in XXL.

    I’ve actually had this supposition presented to me numerous times and had people drag me to clothing sales because of “all of the XXL sizes” they saw at the last one. However, when we start actually looking, they discover that there really aren’t nearly as many options in that size as they thought.

    I wonder if the observation about lack of options in a given size is mostly a confirmation bias. Whatever size you’re looking for and don’t find triggers the observation that there are other options and you register it.

    For instance, if you go looking for a M polo shirt and *find one*, do you count the XXL and XXS on the same rack? Would you notice if there were *none* of either? However, when you go looking for a M polo shirt and come up empty, you are acutely aware of the excess of XXL and XXS in the same style.

    Given that I have heard the complaint from people of all sizes about their difficulty finding clothing in their size, I’ve got to wonder about how the actual distribution curve matches up with perception.

    Thumb up 0 Thumb down 0
  173. Conor - Ireland says:

    hmmmm, i wonder what it says about the Freakonomics readership that this post garnered 172, and counting, responses…

    It’s amazing how we can come up with so many theories rather than actually ask some retailers to review a few statistics. Or does the size issue touch a nerve with a lot or people???

    Thumb up 0 Thumb down 0
  174. Chris B says:

    There’s an automatic control group in large shoes versus large other clothes. No one has size 14 feet because he’s heavy (maybe one size up.) Among men, at least, I suspect most men shop because their previous shoes are worn out, not primarily for fashion, so the shopping-for-fun bias is also mostly out.

    At Nordstrom Rack, admittedly just one chain but one that concentrates on shoes and on sales, there are always far more size 10 men’s shoes than size 14, even for the number of shoppers. Maybe it’s my confirmation bias, but there always seems to be a much higher ratio of orange-and-pink-polka-dotted shoes in the 14s than the 10s, as if the ‘normal’ size 14 shoes have been picked over much more aggressively. I wear 13 quite wide, 14s often fit, and most of the time I can’t find a single pair of shoes in there that will fit me well, of any kind (walking, business, athletic.) It’s hard to imagine too many size 10s saying the same.

    In fact, it’s common for me to go into a regular non-discount shoe store and not find one shoe that fits at any price, especially here in my current part of Colorado where there is a large majority of Germanic-ancestry whites. An employee at a regular shoe store confirmed what I suspected – that the big shoes usually go within a few days of a new shipment coming in. There certainly aren’t many left to be on sale.

    That points to a bias in national chains. Areas with a lot of ethnically bigger people – blacks near a Rack in Chicago I’ve seen, whites in my current area of Colorado – seem to have fewer big clothes and shoes, whereas Racks in areas with more by-heritage Latinos, Asians, and southern Europeans (Italians etc.) seem to have more big clothes and shoes on sale, presumably because stores don’t adjust their ordering for such a difficult to quantify situation.

    Of course, given that Nordstrom proper is expensive, and there is a well documented income bias in favor of taller (and bigger-shoed) men, maybe it’s just that all those rich Germanic execs are buying my running shoes and hiking boots at full price before they can be demoted to the Rack.

    Thumb up 0 Thumb down 0
  175. cqfd says:

    Some people have stated it before but it is as simple as: medium sizes go quicker because there are more shopper of those sizes, which gives you the impression that XXL and XXS don’t sell. They do, just not as quick – because the population that buys them is smaller.

    Thumb up 0 Thumb down 0
  176. Jim Birch says:

    Sorry but there’s some pretty simple economics going on here, you don’t need to get all that complicated.

    You buy 2000 shirts for $3 each in China, spread them round a bunch of outlets where you retail them for $60. At the end of the run you get the stuff that’s left and flog it off at a discount outlet – or the discount rack – for $15 or something. Who cares if there’s a bunch of your less wanted sizes left at the end – it’s the 1900 shirts that you sold for $60 that’ll make your business work. The one you’ve got one on the rack to get the full price retail sale is worth about 10 leftovers.

    Thumb up 0 Thumb down 0
  177. CHill says:

    #13 seems to follow Occam’s razor (stated even more simply by #70). Mediums sell out faster. This hypothesis could be proved or disproved very quickly by getting data from discount stores on the distribution of sizes that they actually receive, as distinct from what the average rack looks like.

    I have a related question: Why are size S garments (my size) most frequently sold out in full retail stores? I rarely find things in my size, and some stores don’t even bother to stock size S, yet I would think a significant fraction of buyers are in this size.

    Thumb up 0 Thumb down 0
  178. Katie says:

    I think that the cut/make of clothes makes a difference. As someone who is larger, there are certain types of clothes that I always pass up. I’m never going to look good in skinny jeans – so I don’t bother to look at them/try them on. There are other types of clothing as well such as halter tops. Certain cuts look different on different types of people. Hence why there might be a excessive number of certain size in store (although this only gives an explanation for the larger sizes). I know that the clothes on clearance racks tend to not be the ones that look good on me.

    Thumb up 0 Thumb down 0
  179. David says:

    So I am an XL or XXL depending on the cut of most clothing, and I can find clothing that is not discounted because I want something that is well made etc etc. The thing that gets me is that if you go and but something that is XL/XXL they are made for someone who is 6 foot 6 inches and all legs. I went to EMS to get some hiking attire and no matter what I put on the crotch was at my knees and I would have had to cut a foot off the legs when having the garment hemmed. This is something I just don’t understand. Does XL stand for extra long all of a sudden?

    Thumb up 0 Thumb down 0
  180. Jonathan Ritz says:

    I have a simple explanation:

    They want it this way.

    If designers allowed the most popular sizes, S,M,L, to end up in the discount stores, the vast majority of people would simply shop there.

    They want to get full price for as much as they can.

    They leave the unwanted sizes XS, XXL for the discount stores as a kind of branding scheme, which is important.

    I agree with many of you that they use market studies to determine how many of each size to produce.

    What I am saying is that there is absolutely no incentive to allow the most popular sizes to land in the discount stores.

    Thumb up 0 Thumb down 0
  181. Keith says:

    Of course buyers know that most people fall around the average size. That’s why non-discounted stores always seem to have some of every size.

    But, in a discount store, people are more likely to buy larger quantities to take advantage of the deal. So, not only are there more medium sized people buying clothes for themselves as we would expect, but there are more people of all sizes snapping up discounted merchandise for the people they know (who, as it happens, are also mostly medium sized).

    Thumb up 0 Thumb down 0
  182. J Lewis says:

    The entire question is based on a false premise. That premise being that there actually are greater numbers of XXL and XS items in discount stores. I am an XXL and have shopped exclusively at Nordstrom for many years. When they have the annual Anniversary sale it is critical that I get in to see my salesperson during the preview and reserve my items before the sale starts. It’s a two week sale and anyone on either end of the bell curve that doesn’t get there in the first 48 hours, isn’t going to have much selection. At the stores closest to my home, 70-80% of all XXL and larger items are gone BEFORE the sale starts simply because the buyers purchase smaller quantities at either end of the size spectrum.

    The bottom line is your asking for an explanation for a phenomenon that doesn’t exist.

    Thumb up 0 Thumb down 0
  183. RJ says:

    sizes vary. A woman may be a 10 in one store, 12 in another and 14 in a third.

    When buying womens clothse, at least, you usually need to try on at least a couple of sizes.

    If there were an even number of people buying (3 per size), and one of them had to buy a size larger, and 1 a size smaller, then the person who is an 8 wouldn’t be able to go small enough, and the person who is an 18 wouldn’t be either – so only 2/3 could buy.

    Hence the overhang.

    Thumb up 0 Thumb down 0
  184. Louis says:

    I think this blog is full of people who know nothing about fashion. The Nordstrom “discount” store is not a dollar store – it is full of high-end fashion. Same thing with Off-Fifth (the Saks “discount” store).

    And yes, these stores are full of XS and XXL sizes. These are also the sizes that a person will find in high end stores. I have no empirical evidence but I imagine the reason is:

    The two types of people who spend a lot on clothing are the young and thin who want to appear rich and the old and fat who are. I know, I know that wealthier people are supposed to be thinner and this is true – up to a point. Once the wealthy reach the age of 50 a sort of inverse bell curve seems to play out. The thin and the heavy predominate and the “middle sized” become fewer and fewer.

    So to reiterate – the thin and young who want to appear rich (and so desire buying Armani for 80% off regular price) buy XS and the Wealthy and and fat buy XXL.

    These “discount” stores only have XS and XXL because these are the sizes that their parent stores carry most.

    Also, the few S, M, and L items get snatched up by the rest.

    And by the way, XS is not unnattractive. XS is a size 4 for a woman – as the pop gets bigger so do the sizes. What use to be size 8 is now size 6 because so many people are fatter.

    Thumb up 0 Thumb down 0
  185. SS says:

    An interesting question is raised by the graphic attached to this post. If we assume that size labeling conforms to size distributions in that target population (here, country of sale), should we expect that a woman sized Medium or Large in the US, finding herself in Europe, where her size–Large or XL–is discounted, will purchase the Large or XL shirts (let’s ignore exchange rates for the purpose of this example!)? Or, might she actually refuse to seek shirts larger than Large? If the shirts are equal we might expect arbitrage, but if the size label “changes” the shirt’s attractiveness, we might not.

    The example given on XXL shirts is presumably taken from the experience of male shopping, as many high-end women’s brands do not offer XL. On the other hand, I would guess the graphic is from an article of women’s clothing, where I would assume XXS is more popular than it is for men’s clothing.

    Thumb up 0 Thumb down 0
  186. steve says:

    I have not had time to read all the comments, so perhaps this was already stated.

    The manufacturers should base their volume decisions on their experience with previous items, not with the percentage of people within each sizing category. The store buyers should do the same. This would obviate most of the comments that I did see.

    Thumb up 0 Thumb down 0
  187. Terry says:

    183 comments, hey? So what I’ve learned from the comments on this blog– that Freakonomics blog commenters comprise a population that is significantly vocal on issues of size/obesity/image/clothes/etc and of, well, the science of insulting women.

    Thumb up 0 Thumb down 0
  188. CandyKay says:

    I’m usually a Size S, but on a few brands I’m an XS, and sometimes I opt for a M – usually for sweater-type garments when I don’t want male colleagues staring at my ta-tas.

    Anyway, there are three sizes that may potentially interest me, and I would assume that M, L, and XL garments would also have a wide pool of smaller or larger potential buyers who might “trade into” that size.

    XS and XXL, however, being on the end of the spectrum, would be of interest to only two size groups. Assuming that retailers feel obligated to carry a full spectrum of sizes – and you should hear how angry larger ladies get if a certain garment isn’t offered in their size! – that would leave them with extra merchandise in both very small and very large sizes.

    Thumb up 0 Thumb down 0
  189. Sara says:

    This might be more compelling for the larger sizes, but it could be that the stores in which the XXLs and XLs are originally sold are using them as framing items rather than expecting to sell them. Assuming many more people fall into the large (or XL) bracket than the XXL bracket, these people might be subtly influenced by seeing that there is still a size or two above them. If people are somehow reassured by being in the middle of the pack rather than, say, being reminded that they are the biggest size there is, then this would give stores incentives to stock sizes that they don’t actually expect to sell. Then when the season for that merchandise ends, it would make sense to send these sizes to the discount stores and see if they can make any profit on them.

    Thumb up 0 Thumb down 0
  190. Tammy says:

    Tell me these stores that have a surplus of XXL because I sure haven’t seen it. Usually it’s just the little sizes.

    Thumb up 0 Thumb down 0
  191. Daniel Reeves says:

    Brian at 20 hit the nail on the head.

    Sellers overbuy just enough for each size. But once the items go on clearance, the supply is not adjusted to the demand of different goods because clearance sales are just to get rid of excess inventory.

    Here’s a mathy example: if your size ratio, from XS to XXL, is 1:2:3:3:2:1, then you expect to sell roughly two or three times more the middle sizes than the outer ones. But the ratio for shirts leftover may look more like 1:1:1:1:1:1. But the ratio for quantity demanded by size may not change with a price drop.

    Thumb up 0 Thumb down 0
  192. MRB says:

    I suspect it’s a little less interesting than all this. Surely, clothing retailers have detailed information about how much clothes in each size they sell; if they know how many Medium’s to stock to run through their supply in the target amount of time, they probably know how many XS/XL’s to stock as well.

    However, they may not be able to purchase that clothing in those same ratios. They may be locked into contracts that require each size to be bought in tremendously large batches (economies of scale) that don’t match the desired ratio; they may need to order all apparel in the same ratio regardless of the specific product (ie, the default ratio of xs/m and xl/m is fine for plain tees, maybe not for ladies thong underwear).

    Thumb up 0 Thumb down 0
  193. Katie says:

    Seeing as the question came from a male, I would have to assume his personal experience is in men’s clothing. In that case it would make sense that xs clothing is unpopular amongst men… many men I know are likely to go up a size bigger than they have to avoid being labeled “extra small”

    Being an “extra small woman” I certainly haven’t observed that smaller sizes are in more abundance on the sales rack at all, much to my chagrin as I am definitely a bargain shopper. It always seems like size 12 has the biggest selection for numbered clothing, and XL for non-numbered sizes.

    Thumb up 1 Thumb down 0
  194. Clark Woolstenhulme says:

    Go to a Nordstrom’s Rack and look at the selection of women’s shoes in the size 10+ versus the size 5-7. The fundamental premise of this article does not seem to hold up. My wife (and her sister) love Nordstrom’s because it is one of the few stores that stock women’s shoes in size 10 or larger, but both women make a point never to walk down the rows of smaller shoes, because they’ll be frustrated that the “cuter shoes” they see there won’t be available in their sizes.

    Maybe the designers don’t spend as much time (and money and effort) on “sharp” clothes for the large and small because they constitute the tails of the distribution, so those customers typically shop at specialty shops or ONLY buy the clothes that do make them look best, leaving the dregs of XXS and XXL to rot on the racks of the outlet store?

    Thumb up 0 Thumb down 0
  195. @shwin says:

    Could it be plain old psychology? The idea of having “Options” on any product allows a customer to get a feel for what they want (even when a large percentage will end up buying the same thing!!). How many times have we selected a shirt we liked only to find that our size (S,M,L) is not available?

    What do we do? We stick around to a) Ask the store manager when the size we want become available b) Check out other shirts from the same label. Both options are very crucial in making sure the customer buys a shirt. The fact that there are XXL and XS shirts in a line-up just adds a bit of spectrum while keeping the customer focused on the product at hand.

    As human beings we always like to keep our options open and the idea that there are sizes of shirts on the rack which we may never buy only helps us in deciding. Therefore the number of such shirts may be greater than the actual requirement to act more as a blinder to get us to take a look at the product.

    And hey!! the fact that extreme sized shirts stay on longer on the rack (as not many people buy them) allows for a free showcase of the shirt even if it is not wearable.

    (P.S: read an article once about a similar explanation for the “I’m Feeling Lucky” link on the Google page. The conclusion was that if the link was removed it would lead to a fall in the number of Google users. Just letting the option exist solved the problem)

    Thumb up 0 Thumb down 0
  196. Kathy says:

    How do you quantify how many people who would have bought medium but could not because there are none left? They always run out of the medium to large sizes but always have too many of the other sizes. It drives me nuts when I go shopping.

    Thumb up 0 Thumb down 0
  197. Nell says:

    First, I do think that men’s and women’s are totally separate categories in the
    clothing busines, 2nd, Of course the stores stock more in the “middle” range;
    More average people-more average clothes.
    And, 3rd, what makes me truly cranky is that they charge MORE for Women’s sizes (meaning XL+) but not LESS for Petite.
    4 -AND that women are not allowed to be SHORT but just “petite”. Hey, I may
    be 5’1, but I come from solid peasant stock. Broad back and solid calves, etc.

    Thumb up 0 Thumb down 0
  198. Aditya Savara says:

    I have long legs so require unusually long jeans. It’s difficult for me to find jeans of these sizes.

    When I _do_ find jeans of these sizes, they are sometimes in a style I don’t like.

    Stores buy quantities to anticipate demand. There is a larger sample of “Medium” wearers, who will consume the medium clothing. The odds of an outlier encountering clothing of their size _and_ style during the _finite_ period of time which it is on the rack, do not approach 100% as quickly.

    If there are 2 of my size in the store, perhaps that year only 1 person like me will walk in. Perhaps another year 3 people like me will walk in. You can’t just get rid of our special sizes with a sale. Need the right customer.

    Note that unlike what some commentors state, in this model, mediums do not always sell out first. It’s just that on average, mediums are left over less. Poster named Kevin (#8) was well on the mark in my view.

    Thumb up 0 Thumb down 0
  199. Adun Ton says:

    here are my opionions.
    1. Most people have common figure, neither too small nor too big, so clothes of moderate sizes would sell quite quickly and easily. And even when clothes factories produce a big portion of clothes of common size and a small portion of clothes of excessive size, the clothes of excessive size will still probably pile up in the stores, taking into consideration the portion of different figures of people. The store would deal this in an efficient way, they generally don’t have the patience to waste time and space for those unsalable or not-so-salable clothes. Finilly, those clothes of excessive size go to discount stores.
    2. People who are fit for small and reall large clothes often have special figure, quite small or quite large. People like this are prone to wear clothes that make them seem more like normal people, which means a small person may buy clothes of sizes that are more or less bigger then his actually size, bigger but not too much bigger ,and a large person may buy clothes of relatively small or tight size, not too tight of course. This would also effect the sales of clothes of sizes in the middle or near the middle norms, and aggravate the situation in my point one.

    Thumb up 0 Thumb down 0
  200. echoclerk says:

    Really you would expect economists to recognise that it just comes down to retailers inability to accurately predict sales numbers on low volume items -(As Mark said above)

    Thumb up 0 Thumb down 0
  201. Scott says:

    The explanation in the post can not be right. Surely buyers would have figured out this behavior by now. In fact they probably only know how much of particular sizes sell, not how many people are of a particular size in the population. I think post #8 gets it right.

    Thumb up 0 Thumb down 0
  202. BG says:

    I asked my sister who is in retail industry and this is her answer:

    The way clothes are bought is by ratio for packaging. And the sum should add up to 6 (half dozen) for example if you order 2-2-2 that is ordering 2 small, 2 medium and 2 larges of the same shirt. We order 1-1-2-2 which means 1 small, 1 Medium, 2 larges, and 2 X-larges. The interesting part is that it used to be 1-2-2-1 and we changed it 2 years ago to accommodate for the over-weight epidemic that is taking place.

    Now when you look at it this way you really don’t have an option of ordering less than 1 small per half dozen. and it’s not a good idea to not order a small at all, because you are loosing that one customer who might need it! you can’t discriminate. so, if you order 144 of a shirt with the ratio of 1-1-2-2 you still end up with 24 smalls but you might only need 13 smalls so the rest ends up on the clearance rack.

    Buyers know this already and believe me the money that is generated from selling 30% of the order full price covers all the first cost. So, the price of loosing small customers is much higher than loosing a few dollars on clearance.

    Thumb up 0 Thumb down 0
  203. Chris says:

    Being an XXL myself (thankfully because I’m mostly tall, not mostly fat) it seems to me there is a much simpler explanation. However, I actually doubt the initial claim in the first place… show me some empirical evidence. Since I routinely look for clothes at discount stores and RARELY find my size. This especially applies to discount shoe stores. I wear a size 15, and I can tell you with certainty, that it is nearly impossible to find any kind of selection in this size.

    But, if the original assumption is indeed true, wouldn’t a simpler explanation be that there are far more people buying the “middle sizes”, so the likelyhood of finding “fringe sizes” would be much higher… the stock would sit longer and accumulate, making it more prevalent.

    Thumb up 0 Thumb down 0
  204. Alison says:

    I agree with #48 and #20 and #170- the whole premise is suspect – I’d like to see some facts or studies to back it up. I am a woman snd have never been is a discount store that had a glut of my size (XS). To the contrary, in my experience as a frequent discount shopper, I normally see very few XS sizes on the rack – but I do often see a high number of large and XL sizes.

    This just doesn’t jibe with reality to me.

    Thumb up 1 Thumb down 0
  205. Will says:

    In order to attract a customer, a retail store must display a reasonable selection of merchandise in that person’s size. What that means, in practice, is that stores must have a full selection of merchandise for each size. More shoppers are in the medium sizes, and so they make more purchases and buy a larger percent of the total selection of medium-sized clothes. Because there are less shoppers at the extremes (XXLs and XXSs, e.g.), there are fewer purchases of those sized clothes. However, since stores needed to purchase the full line of clothes in those sizes to provide adequate selection, a smaller percentage of extreme-sized clothes are purchased, relative to medium-sized clothes.

    My guess is that because the cost of production for an item of clothing is small, retailers are more willing to have excess stock on hand at the end of the season than to miss out on potential purchases during the season due to a lack of variety of certain sized clothes.

    Thumb up 0 Thumb down 0
  206. wisdom says:

    Ouch. I believe that people who do not look “naturally attractive” would probably spend More to compensate, than those who are seen as “naturally attractive”. Attractiveness is a form of social capital and people seek to optimize their “worth”… the same as any other kind of capital.
    People aren’t particularly rational. Or they’re rational enough to realize that clothing won’t do much to disguise excess weight and maybe a naturally unattractive face that first led to body neglect and that it’s not worth spending much on such clothing.

    Also, obesity among young people is growing, but still people tend to get heavy toward middle age. That’s also a period when interest in maximizing presentation declines, except maybe in white-collar jobs. (It’s no secret that many men’s suits are cut large in the body.)

    So yes, larger people probably are cheaper when it comes to clothes.

    Thumb up 0 Thumb down 0
  207. Nancy says:

    My son is too tall to wear boy’s pants and needs a men’s small or medium. We’ve had no luck finding these sizes, though. Men’s departments seem to be universally stocked with L-XXL pants, with nary a M or S in sight.

    At any given time, large numbers of boys will be in this size range. Why on earth can’t store buyers spring for a few more S and M pants?

    Thumb up 0 Thumb down 0
  208. Tom says:

    I doubt the observer finds clothing that they *want* in sizes too large or small…

    I am 6’9″, 300 pounds. Clothes don’t fit me well so I don’t go shopping very often. It’s a frustrating experience because most standard shops don’t carry my size.

    It probably only seems that very large clothing is available to the big or small but the reality is that those items are usually very ugly. The savvy large shoppers call ahead of a shipment to be able to buy the best items.

    The only shoes Nordstrom Rack has are very colorful basketball shoes (clearly leftovers from basketball team orders) and dress shoes (clearly the rest of us are successful business men). Casual clothing basically stops at shoe size 13 and shirt size XL (XLT if we are lucky). I am a 15 and have better luck on ebay.

    I am sure that average sized people are used to being able to buy whatever they like and so they more frequently go to discount stores. What they are experiencing is something similar to what very tall people and very short people experience perpetually… scarcity in their size…

    Thumb up 0 Thumb down 0
  209. Rony says:

    It is very simple – medium size cloths should fit larger people (OK a bit tight)or smaller people. So every size has an additional demand from “neighboring” sizes in both direction. Fringe sizes only from one direction. Provided that jobs order according to an equal distribution. The demand will allways be higher for the sizes in the middle than on the fringes

    Thumb up 0 Thumb down 0
  210. Marc F says:

    I suspect there is some truth to #180’s idea that they want it that way. I’m a S who hates to pay retail. Finding certain items (leather jackets are particularly hard) off price in S is tough. (XS is viryually nonexistent in most men’s departments) I once asked a salesperson at Lord and Taylor and he said if they ordered 1 small in a style it was a lot. Implicit in his reply was that he could sell a lot more if the buyers did not do this deliberately. I’ve often wondered if it was because many women could buy a mens small and get a good jacket at half the price it would fetch in the women’s department.

    Even sample sales I frequent have very few items in small sizes, and lots of small people willing to buy. Since retailers could probably dispose of their S’s at 20% off, if they bought more , why am I confronted by tons of L and XL at 50% off, despite the higher fabric cost?

    Thumb up 0 Thumb down 0
  211. Jemma says:

    Perhaps this means that people are more inclined to buy clothing that doesn’t fit them properly. In my daily observations of the students here at college, most college students either try to wear clothing that is too small for them or in contrast clothing that is far too big. People, who are more cautions about the way they look in public, even just going to and from class, tend to wear clothes according to the number. The smaller the number (size 1 and 2 for example) the better they think they look, and or consider themselves thin. On the other hand, you see people who are more obsessed with being comfortable and care little for their wardrobe. They can be seen around campus wearing clothing that looks to be two sizes too big. So although the average person is a medium that does not mean that is the size they would be most inclined to purchase. So perhaps stores anticipates this, having observed the tends in sizes over the years and stock more of the larger sizes and smaller sizes, which creates an excess of each and not so much of the middle sizes.

    Thumb up 0 Thumb down 0
  212. david says:

    I think that that’s about right. I don’t think it was a conscious decision, I think they just observed that it gets sold that way and continued.

    As another indicator of his argument’s validity, notice that discount stores rarely run massive discounts on clothing. They already have discount prices on other goods such as the last seasonal holiday’s candy or movie t-shirt, but Target, Old Navy, and other low-cost clothing providers typically cut prices only marginally, and when they do do blow outs, they’re typically to rid themselves of last season’s stock.

    Thumb up 0 Thumb down 0
  213. Frederick Michael says:

    #20 hit the kernel of the answer. The extreme sizes stay on the sale rack longer.

    This is classic Freakonomics stuff. For example, “Why are most of the prisoners in a jail on any given day serving longer sentences yet the jail says that most of their prisoners had short stays?”

    The answer is that the longer stays get “counted” over and over in the snap-shots of prison population. Similarly, a single XXL shirt that sits on the sale rack for 5 years can tilt the stats on what the rack looks like on a random day.

    Thumb up 0 Thumb down 0
  214. Eric says:

    The answer is very simple.

    Because each store needs to stock all the sizes.

    Most items are stocked in small numbers – a store wants to carry the minimum number of items that it can, both to reduce inventory costs and to reduce risks if something doesn’t sell. So, you look at the common distribution of sizes purchased at the store, and you find that it’s

    XS – 0.5
    S – 1
    M – 5
    L – 3
    XL – 1
    XXL – 0.3

    You don’t want your shoppers to be disappointed if they wear XS or XXL, so you buy 1 of each of those. Averaged across all your stores (and all clothing lines), that means you will have some of the outlier sizes left over. Off to the discount house they go.

    Thumb up 0 Thumb down 0
  215. Liz says:

    1) I don’t think they take the design/body type relationship into account when they plan retail inventory – and I think they should. I am a size 4 and when I’m looking at sale items the ones that would look good on me are typically only available in sizes 14+.

    2) A distribution of sizes for women’s tops at a 70%+ discount right now found at Shopstyle.com (includes over 200 retailers).

    XXS 38
    XS 291
    S 569
    M 572
    L 428
    XL 187
    Plus 2X 8

    Thumb up 0 Thumb down 0
  216. Pieter says:

    I might have missed someone else’s post, but I wonder how much psychology plays into this. Extreme aversion is a type of cognitive bias where a person will avoid the extreme options and choose a middle option instead. People will do this regardless of whether or not the middle “fits” or is the best option for them. This affects many people’s decision-making for fear of standing out from the crowd.

    This bias was recently documented in consumers’ choices of soda size at fast food restaurants . In this study, customers were often choosing the 16oz size of soda (the middle between the 12oz and 20oz). When fast food restaurants upped their drink sizes and the 16oz soda became the “small” and 21oz became the “medium,” most consumers chose the 21oz drink because it was the new “medium.”

    Given America’s increasing waistline, it would be interesting to see how different a size “M” or “L” from today compares with the same labeled size of 10 years ago. The point being: is the large from 10 years ago the medium of today? If so, we might have a case of extreme aversion in the clothing industry.

    Thumb up 1 Thumb down 0
  217. Jake says:

    Thought I might offer the following explanation to the question. It relies less on the psychology of XXL shopping, and puts a bit more faith in micro-optimization.

    A glut of XXL clothing usually ends up left on shelves during sales, but such an excess doesn’t usually exist at the regular prices. Why does a change in price, specifically, cause a surplus of XXL clothes?

    At regular prices, retailers can forecast and order enough XXL clothing given demand, considering the amount of data they collect. That XXL surpluses appear during sales suggests the following: that demand for XXL clothing is much less elastic compared to moderate sizes.

    The upshot of this inelasticity is that the careful forecasting which makes sure that, at regular prices, the Gap orders the right amount of blazers goes out the window when they put those blazers on sale. The ratio of small to medium to large sizes that is correct at regular prices no longer works at sale prices. The demand for moderate sizes rises to a much greater extent and excess medium sizes are snapped up quickly. The demand for XXL goods on the other hand, does not rise nearly as much, and thus, those XXL shirts are left on the sales rack despite the Gap’s best profit maximizing efforts.

    Thumb up 0 Thumb down 0
  218. Cindy says:

    That is an interesting question to ask! There is a similar situation with shoes. The shoes on sale feature disproportionate numbers of very small sizes and very large sizes. I am lucky to be a size 6. Because I know the past pricing behavior for shoes, I don’t buy shoes at regular price because I know they will be on sale soon. So if every customer thinks like I do, then the demand for regular priced shoes will go down and retailers will have to mark them down, resulting in the disproportion. For retailers, they don’t want to cut back on the supply either because making a lower profit is better than getting nothing. So the cycle just self-perpetuates.

    Thumb up 0 Thumb down 0
  219. thethorninyour says:

    The reason why there are more larger sizes left at the end of the season is because they are priced higher than their regular sized equivalents the moment they hit the store. A Nautica polo shirt that costs $57 S-XL, costs $67 2XL on up. Big and Tall Buyers wait for the discounts so they don’t have to pay the upcharge retailers such as Belk, Dillards and Macy’s automatically place on big and tall products. It has nothing to do with amount of fabric in the larger sizes. there is a bigger difference between small and XL than there is between XL and 2XL. If they want to price by size, why aren’t all the sizes a different price. This problem doesn’t exist in other parts of the store. Mens shoes size 6-13 are one price. Mens suits size 40S to 48L are all one price. Bras sized 32A to 38D are all one size. What makes big and tall Polo shirts so different. Pricing discriminataion. the same goes for Plus size women and Petite sizes. The customers who buy these items are a captive audience that the retailers feel perfectly fine overcharging. Kind of like a movie theater charging $8 bucks for a bucket of popcorn just because they can.

    Thumb up 0 Thumb down 0
  220. ginger says:

    i worked for many years in a clothing distribution. being fat skiny med. sample; all clothing is tagged by price first,walmart, kamrt etc.then its all counted by sizes and seperated. then its broke down to what size amounts and even in colors at times. for instance if thers ioo.oo small. 70 med and 50 large. it get broken down by a dozen 7 sm. 3 med 2 large. and that combination is done until what size is done and broke down again. all tagging is done bt distribution . so big lots, pennys mervys kmart tags are also switch using the same clothing but diffrent tags. your buying the name tag not the brand.

    Thumb up 0 Thumb down 0
  221. Sam says:

    I am a size xs. I do find a lot of clothes marked xs in discount stores but they are always clothes that are Medium or small, sometimes even large that are marked wrong. No wonder they end up in the discount section. People who are xs wont fit into them and people who are a size S or M wont obviously bother looking at the xs section

    Thumb up 0 Thumb down 0
  222. A Californian says:

    Having worked, many years ago, in the buying line at Macy’s, let me suggest this:

    You have to assort each store with a minimum number of pieces in each size (even if that number is 1 fr the largest and smallest), while you can buy the ‘right’ number of pieces for the middle sizes. The stores know the ratio of pieces in each size, say

    size 2 4 6 8 10 12 14 16
    # 1 2 4 8 8 6 4 2

    Let’s say you’re giving each store 12 pieces. How do you divide them?

    size 2 4 6 8 10 12 14 16
    # 1 1 2 3 2 1 1 1

    So maybe no one bought that one size 2 piece in 7 of the 20 stores you bought it for. Those 7 get aggregated and sent to the discount location.

    Thumb up 0 Thumb down 0
  223. Gene Craig says:

    Sizes have been changing, there is a psychology to it. When I was 25 lbs. lighter than my current 200lbs, I bought T-Shirts and Polo’s in XXL and they fit just a wee bit loose, but the way I liked them. I recently tried on a XXL Polo and there was room enough for my wife to fit in it with me. Now it is XL, and in some cases L fits fine. I think this size change helps (or not) people feel that they are smaller than they really are if they can fit into a smaller size; making them feel better about themselves.

    Thumb up 0 Thumb down 0
  224. Lyndy says:

    Well I would surely like to trade with some of these people as my XS and XXL sizes sell out quickly and I’m stuck with Mediums all the time. And responding to the following:

    ” My best guess is that for some reason small and large people are, in general, less willing to pay for clothes. Maybe because they are outside of the norms for physical beauty, they believe that sharp clothes won’t help them that much. ” Quoted from Erik

    I find this very unflattering as I myself wear a small, spend loads on clothing and have never been considered outside the norm for physical beauty!

    You might also note that clothing manufactuers don’t all use the same guidelines for measurements. I received a shipment of high quality dresses last week that had to be returned as the Large was a 32″ bust (hah to that) I carry sizes for almost all sizes size 0 to womens plus size 6X and they are all beautiful women and not outside the norm for physical beauty at all, so, I am terribly offended by Eriks comment

    Thumb up 0 Thumb down 0
  225. Jessica says:

    I don’t understand this scenario at all beings i am an extra small and cannot for the life of me find clothes that fit whether at full price or discount outlets the extra small to me seems to be extinct

    Thumb up 0 Thumb down 0
  226. Tim says:

    Seems nobody really knows the answer. There must be some unpublished economic forces at work here. Business people are not stupid. It seems almost every store I go to has S and XXL sizes in piles everywhere, and 1 M size in red.

    Thumb up 0 Thumb down 0
  227. Dave McGarvey says:

    My “best guess” is that XS and XXL are minus 3 and plus 3 standard deviations of clothing size sales and unless stores maintain careful records of what they have purchased in the past, it would be easy to overbuy these sizes, and underbuy the more commonly purchased sizes. This is supported by popular sizes usually being sold out. The idea that people with extra small or extra large requirements not caring about their appearances or not wanting to invest in clothing to look good in is speculative and biased at best.

    Thumb up 0 Thumb down 0
  228. Bill Ogorodny says:

    I believe that many people in the XXL and XL sizes are lower income people. It is expenmsive to eat foods with nutrients. It is much cheaper to fill up on McDonalds food and large amounts of soda.

    Thumb up 0 Thumb down 0
  229. Natalia says:

    Could it be that people just want to belive they are in the middle sizes and that’s why they are less prone to buy clothing that are labelled in the extremes? if you are and XXS you can probably also wear the XS. I notice that this does not happen so much with the kids clothing that are on sale, probably bacuase we parents buy right sizes for our kids…just a thought…

    Thumb up 0 Thumb down 0
  230. John Pilge says:

    In the discount stores, they get the items that OTHER STORES could not sell for the season. Usually the odd sizes make up the bulk of the shipment.

    When there is a sale, most people buy the most popular sizes. What is left are the least popular sizes. I have seen my size sell out in a day. (They have it in the evening the night before the sale, but they are all gone the day of the sale by the time I get off work and can visit the store.) Think about it. You are finding all those odd sizes, DURING a sale or right after a sale.

    Do large people buy fewer clothes? My experience says yes. They are less active and can make their clothes last longer.

    If you do your own shopping, notice how few extra large/small people actually shop at the store that has excess odd sizes.

    Thumb up 0 Thumb down 0
  231. jane says:

    There are fewer of smaller and larger people – store buyers know that, and adjust their orders accordingly.

    However, buyers’s are mostly trained in predicting desires and willingness to pay of average-size people. That makes sense, since there are more of these people, and being prepared to handle them causes more profit/loss overall for a retailer. So, buyers are more often wrong when ordering stuff for people of rare sizes, the extra-smalls and extra-larges.

    People often complain that it is impossible to find sizes 0 (zero) and 2, and there is nothing good to buy in sizes 22 and up. And, this is the stuff that ends up on the sales rack. Retailers seem to continuously stock the wrong things.

    Thumb up 0 Thumb down 0