Zip, indeed.
July 3, 2009 12:39 PM Subscribe
Since we all have nine-digit zip codes now, why is it still neccesary to include the city and state in a postal address?
Zip codes, as far as I know, are unique. If so, why should we still have to include the city and state in a mailing address? Isn't that redundant?
Zip codes, as far as I know, are unique. If so, why should we still have to include the city and state in a mailing address? Isn't that redundant?
Yes and redundancy is good in this context. If you transpose two digits or incorrectly enter one or more, your letter is still far more likely to get with redundant information. It comes at a low cost too.
posted by allen.spaulding at 12:41 PM on July 3, 2009
posted by allen.spaulding at 12:41 PM on July 3, 2009
Zip codes, as far as I know, are unique.
Well, unique, but not necessarily to the granularity of a single mailbox.
A ZIP+4 Code consists of the original 5-digit ZIP Code number plus a 4-digit add-on code. The 4-digit add-on number identifies a geographic segment within the 5-digit delivery area, such as a city block, office building, individual high-volume receiver of mail, or any other unit that would aid efficient mail sorting and delivery....
ZIP+4 codes are intended for use primarily by business mailers who prepare their mail with typewritten, machine-printed, or computerized formats that can be read by the business's automated scanners during processing.
posted by dhartung at 1:06 PM on July 3, 2009
Well, unique, but not necessarily to the granularity of a single mailbox.
A ZIP+4 Code consists of the original 5-digit ZIP Code number plus a 4-digit add-on code. The 4-digit add-on number identifies a geographic segment within the 5-digit delivery area, such as a city block, office building, individual high-volume receiver of mail, or any other unit that would aid efficient mail sorting and delivery....
ZIP+4 codes are intended for use primarily by business mailers who prepare their mail with typewritten, machine-printed, or computerized formats that can be read by the business's automated scanners during processing.
posted by dhartung at 1:06 PM on July 3, 2009
Technically you don't; it'll usually get there anyway. Sometime's it'll take longer, though. There's a student group at MIT that often sends itself international mail to "[group acronym], MIT, USA"; the fun is in seeing how long (or how short) it takes for the mail to arrive. I think it usually does.
This is not actually the main activity of that group ... just a weird tradition.
posted by spaceman_spiff at 1:08 PM on July 3, 2009 [2 favorites]
This is not actually the main activity of that group ... just a weird tradition.
posted by spaceman_spiff at 1:08 PM on July 3, 2009 [2 favorites]
Response by poster: dhartung, I don't think you read the question completely. I didn't suggest the street address was superfluous, only the city and state.
posted by dinger at 1:35 PM on July 3, 2009
posted by dinger at 1:35 PM on July 3, 2009
Your zip+4 code isn't unique. Well, it can be but it probably isn't. My Zip+4 has 374 people in it. Hope your postman knows your name and where you live. It would be even worse than when I lived on "Rural Route 1", and address that I shared with about 50 other people around the rural town I grew up in. When we got a new mail carrier no one on the whole route got the right mail.
Besides, who wants to live at a number? I want to live on a street! "Hey Bob, come over for beer and BBQ, I live at 94108-1945!" isn't going to get anyone coming over any time soon.
posted by Ookseer at 1:36 PM on July 3, 2009
Besides, who wants to live at a number? I want to live on a street! "Hey Bob, come over for beer and BBQ, I live at 94108-1945!" isn't going to get anyone coming over any time soon.
posted by Ookseer at 1:36 PM on July 3, 2009
Ah, I guess I missed that part too.
City and state? Redundancy and human readability. If your postal carrier in LA sees can read that the city or state is wrong he can put it back in the system. They're much less likely to notice a couple transposed numbers. Also if an envelope gets smudged or torn it's a lot easier to reconstruct the missing parts of the address from the city and state than from the zip+4.
posted by Ookseer at 1:40 PM on July 3, 2009
City and state? Redundancy and human readability. If your postal carrier in LA sees can read that the city or state is wrong he can put it back in the system. They're much less likely to notice a couple transposed numbers. Also if an envelope gets smudged or torn it's a lot easier to reconstruct the missing parts of the address from the city and state than from the zip+4.
posted by Ookseer at 1:40 PM on July 3, 2009
I have sent mail without the city and state and it has arrived at its destination. I don't think the Post Office encourages it, because if you misprint the Zip Code your mail is SOL, but it will arrive if you're careful.
I think it's basically the redundancy that keeps the city/state around. Even in other countries where they have more granular postcodes (e.g. the UK, which has alphanumeric postcodes that get you ZIP+4 accuracy without adding anything extra), they still put the traditional human-readable address on there.
Although people typically don't introduce themselves by saying "hi, I'm from 04241, what about you?" I have noticed an uptick in recent years of people using their Zip code to identify their location in things like CraigsList postings. Presumably this is because you want to give your location with more accuracy than a city/state allows, but you also don't want to give your complete street address. So people frequently just give the Zip, and it's enough to let other people familiar with the area know where they approximately are.
posted by Kadin2048 at 1:47 PM on July 3, 2009
I think it's basically the redundancy that keeps the city/state around. Even in other countries where they have more granular postcodes (e.g. the UK, which has alphanumeric postcodes that get you ZIP+4 accuracy without adding anything extra), they still put the traditional human-readable address on there.
Although people typically don't introduce themselves by saying "hi, I'm from 04241, what about you?" I have noticed an uptick in recent years of people using their Zip code to identify their location in things like CraigsList postings. Presumably this is because you want to give your location with more accuracy than a city/state allows, but you also don't want to give your complete street address. So people frequently just give the Zip, and it's enough to let other people familiar with the area know where they approximately are.
posted by Kadin2048 at 1:47 PM on July 3, 2009
As someone who ships a lot of things online, you would be surprised how many people don't even know their own zip codes. At least once a day, I have to look up a zip code for a city because the shipping program I use says that the city and zip do not match. If it relied on the zip code only without cross-referencing it with the city and state, I would have a dozen or more lost packages a month.
posted by Ugh at 1:47 PM on July 3, 2009 [1 favorite]
posted by Ugh at 1:47 PM on July 3, 2009 [1 favorite]
Response by poster: All good points so far. My understanding is that most addresses are machine-read these days, and the numbers are far easier for the scanners to read than the alpha characters, especially if people use cursive. I understand the value of redundancy, but really, if the numbers are accurate and legible, won't the city/state be ignored anyway? Anyone from the Post Office care to weigh in?
posted by dinger at 1:56 PM on July 3, 2009
posted by dinger at 1:56 PM on July 3, 2009
Do we have to include city and state?
posted by apetpsychic at 2:21 PM on July 3, 2009
posted by apetpsychic at 2:21 PM on July 3, 2009
Best answer: There is a pretty good video on the mail sorting system here. This is an issue I did some research on a few years back (work-related). My former employer had several big contracts with the USPS.
Basically, when a mailpiece comes into a sorting center, it will go through a machine that barcodes it (if it's not already barcoded) to the ZIP+4 level. It does this by scanning the mailpiece and attempting to perform OCR on the address. Until fairly recently, their system was only capable of resolving machine-printed ZIPs and addresses, but they currently claim a fairly high degree of success even on handprinted items. (If an item can't be machine-read, the image is sent to a human for analysis, and only if the scan is unreadable does the actual mailpiece get pulled out for physical inspection.)
So if you only put a 5-digit ZIP on your letter, they're going to look at the whole address in order to figure out the ZIP+4, because they want that to use in sorting. I don't know whether their system looks at the City and State in order to "sanity check" the provided ZIP code, but it might. (I think it does — I think it checks against the same database that many retailers now validate against.)
The result of all this is a barcode, printed in UV-reactive light orange ink typically along the bottom edge of the mailpiece. This is what the actual high speed sorting machines use.
The sorters will sort down to different levels of granularity depending on the destination address. It obviously doesn't make sense for a sorting facility in Denver to sort letters bound for Miami down to the carrier-route level; they might only get sorted to the first three digits of the ZIP or so. (I remember reading about this and thinking that it was very similar to the idea of CIDR or route aggregation in Internet routing.) But for local mail they'll sort with much higher granularity.
If you were to address a package and apply a ZIP+4 barcode that was completely different from the human-readable address, I'm not exactly certain where the error would be caught. It's possible that it would get sorted, based on the barcode, all the way down to the carrier-route level, and it's only when the carrier looked at it while preparing for delivery that it would get kicked back and re-barcoded. (You could test this trivially; send something from an East Coast address with a human-readable local address, but a West Coast ZIP+4 barcode, and see how long it takes to arrive. If it goes to the West Coast first, the delay would be obvious.) However, they may run even pre-barcoded non-bulk mail through the automatic barcode machines, just to prevent this sort of error. (Such a condition wouldn't be uncommon if someone reuses an envelope, or writes "Return to Sender" and drops a mailpiece back in a mailbox — it would still have the old barcode on it. Lots of people don't even notice them.)
I think the Post Office would lurve it if everyone started using ZIP+4s accurately, all the time, so they could just do numeric OCR and know exactly where a mailpiece was headed. There's literally millions (probably billions, if you include all the development costs) of dollars worth of infrastructure dedicated to figuring out people's chickenscratch handwriting and turning it into ZIP+4s for sorting purposes. But since people don't do that, and because they frequently get ZIP codes wrong or they get smeared or whatever, they're not likely to drop the City/State component of the address anytime soon.
posted by Kadin2048 at 2:30 PM on July 3, 2009 [3 favorites]
Basically, when a mailpiece comes into a sorting center, it will go through a machine that barcodes it (if it's not already barcoded) to the ZIP+4 level. It does this by scanning the mailpiece and attempting to perform OCR on the address. Until fairly recently, their system was only capable of resolving machine-printed ZIPs and addresses, but they currently claim a fairly high degree of success even on handprinted items. (If an item can't be machine-read, the image is sent to a human for analysis, and only if the scan is unreadable does the actual mailpiece get pulled out for physical inspection.)
So if you only put a 5-digit ZIP on your letter, they're going to look at the whole address in order to figure out the ZIP+4, because they want that to use in sorting. I don't know whether their system looks at the City and State in order to "sanity check" the provided ZIP code, but it might. (I think it does — I think it checks against the same database that many retailers now validate against.)
The result of all this is a barcode, printed in UV-reactive light orange ink typically along the bottom edge of the mailpiece. This is what the actual high speed sorting machines use.
The sorters will sort down to different levels of granularity depending on the destination address. It obviously doesn't make sense for a sorting facility in Denver to sort letters bound for Miami down to the carrier-route level; they might only get sorted to the first three digits of the ZIP or so. (I remember reading about this and thinking that it was very similar to the idea of CIDR or route aggregation in Internet routing.) But for local mail they'll sort with much higher granularity.
If you were to address a package and apply a ZIP+4 barcode that was completely different from the human-readable address, I'm not exactly certain where the error would be caught. It's possible that it would get sorted, based on the barcode, all the way down to the carrier-route level, and it's only when the carrier looked at it while preparing for delivery that it would get kicked back and re-barcoded. (You could test this trivially; send something from an East Coast address with a human-readable local address, but a West Coast ZIP+4 barcode, and see how long it takes to arrive. If it goes to the West Coast first, the delay would be obvious.) However, they may run even pre-barcoded non-bulk mail through the automatic barcode machines, just to prevent this sort of error. (Such a condition wouldn't be uncommon if someone reuses an envelope, or writes "Return to Sender" and drops a mailpiece back in a mailbox — it would still have the old barcode on it. Lots of people don't even notice them.)
I think the Post Office would lurve it if everyone started using ZIP+4s accurately, all the time, so they could just do numeric OCR and know exactly where a mailpiece was headed. There's literally millions (probably billions, if you include all the development costs) of dollars worth of infrastructure dedicated to figuring out people's chickenscratch handwriting and turning it into ZIP+4s for sorting purposes. But since people don't do that, and because they frequently get ZIP codes wrong or they get smeared or whatever, they're not likely to drop the City/State component of the address anytime soon.
posted by Kadin2048 at 2:30 PM on July 3, 2009 [3 favorites]
I would say not as many people know their Zip+4 as know the 6-digit version of their Zipcode. And the 6-digit version in a few rare instances can span adjacent cities.
posted by Jeff Howard at 2:35 PM on July 3, 2009
posted by Jeff Howard at 2:35 PM on July 3, 2009
Five digit.
posted by Jeff Howard at 2:48 PM on July 3, 2009
posted by Jeff Howard at 2:48 PM on July 3, 2009
I would say not as many people know their Zip+4 as know the 6-digit version of their Zipcode. And the 6-digit version in a few rare instances can span adjacent cities.
Yes. Monte Sereno, CA and (parts of) Los Gatos, CA are 95030.
posted by dogmom at 2:56 PM on July 3, 2009
Yes. Monte Sereno, CA and (parts of) Los Gatos, CA are 95030.
posted by dogmom at 2:56 PM on July 3, 2009
Seconding Ugh above on how frequently people get their zip codes wrong. I used to work for a company that shipped lots of packages on a daily basis, and we had a zip code error rate of almost 5%, and this was in an online form that would automatically fill in your state once you filled in your zip, which helped some people catch their errors, but not all. I did once see a package that someone had tried to ship to Miami, Ohio, as a result of that.
We shipped all our packages via FedEx, and what happened with them was that the package would arrive at the destination sort facility, somewhere in Ohio, before it was sorted down to the level where someone would notice that the zip didn't match the address. Then they'd call us to try to get a correct address, we'd update it, and it would be rerouted for delivery the next day. So it caused a delay of at least a day (assuming that process went smoothly and we could provide an address) in an overnight package. I imagine it'd be considerably more in a slower package.
And if the package didn't show city and state, it'd be even worse, because it might actually go out for delivery, especially if it was a common address, like "17 Oak St" or something.
posted by dizziest at 3:36 PM on July 3, 2009
We shipped all our packages via FedEx, and what happened with them was that the package would arrive at the destination sort facility, somewhere in Ohio, before it was sorted down to the level where someone would notice that the zip didn't match the address. Then they'd call us to try to get a correct address, we'd update it, and it would be rerouted for delivery the next day. So it caused a delay of at least a day (assuming that process went smoothly and we could provide an address) in an overnight package. I imagine it'd be considerably more in a slower package.
And if the package didn't show city and state, it'd be even worse, because it might actually go out for delivery, especially if it was a common address, like "17 Oak St" or something.
posted by dizziest at 3:36 PM on July 3, 2009
Best answer: Kadin2048 is accurate in their breakdown of the sorting systems.
The sorters will sort down to different levels of granularity depending on the destination address. It obviously doesn't make sense for a sorting facility in Denver to sort letters bound for Miami down to the carrier-route level; they might only get sorted to the first three digits of the ZIP or so.
3-digit sort is the first sort, yes. Then at the more local facility, it will sort letters down to the carrier route, putting them in order and everything. Formerly -- and this is still true of parcels and flats (ie. everything that doesn't meet the specifications of a letter) -- stuff would get sorted down to the carrier, and then the carrier would have to prep all their mail and put it in order. They've got a system they're working on (I don't know if I'd quite call it beta stage, or what, but they've got working prototypes last I heard) that will do the carrier-route sorting for flats, as well (Flats Sequencing System, FSS).
I think the Post Office would lurve it if everyone started using ZIP+4s accurately, all the time, so they could just do numeric OCR and know exactly where a mailpiece was headed. There's literally millions (probably billions, if you include all the development costs) of dollars worth of infrastructure dedicated to figuring out people's chickenscratch handwriting and turning it into ZIP+4s for sorting purposes.
I would second the lurve, and second the billions. They've managed to cut back a ton on the physical labor involved in sorting, at least, so that's good. But there are still buildings (or just one, not sure anymore) filled with people who see images on their screen, and key in the address info for stuff that couldn't be read via OCR.
/former contractor with USPS
posted by inigo2 at 4:21 PM on July 3, 2009
The sorters will sort down to different levels of granularity depending on the destination address. It obviously doesn't make sense for a sorting facility in Denver to sort letters bound for Miami down to the carrier-route level; they might only get sorted to the first three digits of the ZIP or so.
3-digit sort is the first sort, yes. Then at the more local facility, it will sort letters down to the carrier route, putting them in order and everything. Formerly -- and this is still true of parcels and flats (ie. everything that doesn't meet the specifications of a letter) -- stuff would get sorted down to the carrier, and then the carrier would have to prep all their mail and put it in order. They've got a system they're working on (I don't know if I'd quite call it beta stage, or what, but they've got working prototypes last I heard) that will do the carrier-route sorting for flats, as well (Flats Sequencing System, FSS).
I think the Post Office would lurve it if everyone started using ZIP+4s accurately, all the time, so they could just do numeric OCR and know exactly where a mailpiece was headed. There's literally millions (probably billions, if you include all the development costs) of dollars worth of infrastructure dedicated to figuring out people's chickenscratch handwriting and turning it into ZIP+4s for sorting purposes.
I would second the lurve, and second the billions. They've managed to cut back a ton on the physical labor involved in sorting, at least, so that's good. But there are still buildings (or just one, not sure anymore) filled with people who see images on their screen, and key in the address info for stuff that couldn't be read via OCR.
/former contractor with USPS
posted by inigo2 at 4:21 PM on July 3, 2009
> Do we have to include city and state?
No. The PO must attempt to deliver the letter, given enough information for address identification, and the 5-digit zip code correlates to city and state.
posted by Rash at 7:38 PM on July 3, 2009
No. The PO must attempt to deliver the letter, given enough information for address identification, and the 5-digit zip code correlates to city and state.
posted by Rash at 7:38 PM on July 3, 2009
@Rash: that's not completely correct. In rural areas, the same 5-digit ZIP code may serve more than one town if there is only one main post office responsible for delivery. For instance, Amherst and Pelham, Massachusetts, share the same 5-digit ZIP code (01002).
posted by brianogilvie at 8:19 PM on July 3, 2009
posted by brianogilvie at 8:19 PM on July 3, 2009
brianogilvie: Rash is pretty much completely correct. the ZIP code 01002 ties to the USPS city of Amherst (which, apparently, overlaps both Amherst and Pelham) with Pelham being an acceptable alternative name. Look it up at Find All Cities in a ZIP Code.
Anyway, (which also partly answers the original question) in the context of addresses, the USPS would like there to be a one-to-one mapping of ZIP codes to Cities/States. So ZIP codes are "unique" in the sense there's only one City, State for a particular ZIP code that you'd put on an address label if you do address correction.
But, they want the printed City, State name for, as others have said, redundancy and the ability to do more error checking if needed. Even if the City, State and ZIP code don't match, at least one of them has a decent probability of being correct. So you can see if the written street address is in either one to find which is correct; the odds of the particular street address being in both the mismatched City, State and ZIP is pretty small (unless the written city is some huge area like New York, NY, I suppose).
posted by skynxnex at 9:58 PM on July 3, 2009
Anyway, (which also partly answers the original question) in the context of addresses, the USPS would like there to be a one-to-one mapping of ZIP codes to Cities/States. So ZIP codes are "unique" in the sense there's only one City, State for a particular ZIP code that you'd put on an address label if you do address correction.
But, they want the printed City, State name for, as others have said, redundancy and the ability to do more error checking if needed. Even if the City, State and ZIP code don't match, at least one of them has a decent probability of being correct. So you can see if the written street address is in either one to find which is correct; the odds of the particular street address being in both the mismatched City, State and ZIP is pretty small (unless the written city is some huge area like New York, NY, I suppose).
posted by skynxnex at 9:58 PM on July 3, 2009
Back in the day, one could simply write "city" under the name and street (as long as the destination city was the same as the origination city) and, miraculously, the mail would get there. Can one still do that?
posted by Thorzdad at 4:49 AM on July 4, 2009
posted by Thorzdad at 4:49 AM on July 4, 2009
Seconding the thing about redundancy, and the thing about people not knowing their own zip codes (or their postal state abbreviations, etc.) I live in Arkansas, abbreviation AR, though people seem to write AK (that's Alaska) on a regular basis. Between automated sorting and sketchy handwriting and human error, the redundancy seems very worthwhile.
All that said, when I'm writing my return address on the upper-left-hand corner of something, it's usually just:
B. Ox
123 Fake St.
90210
posted by box at 8:38 AM on July 4, 2009
All that said, when I'm writing my return address on the upper-left-hand corner of something, it's usually just:
B. Ox
123 Fake St.
90210
posted by box at 8:38 AM on July 4, 2009
This thread is closed to new comments.
posted by sciencegeek at 12:40 PM on July 3, 2009 [1 favorite]