posted 1999.03.01 rewritten 2002.02.10 last edited 2007.07.17
Since 1984, area codes throughout North America have been running out of phone numbers and exchanges. To relieve the shortage, changes were made to AT&T's North American Numbering Plan. Existing area codes were split and new area codes were added, until the existing rules allowed no more area codes. Then the rules were changed to allow more area codes. The first changes seemed reasonable at the time, and still seem reasonable. However, a very costly and disruptive kind of change got started and continues to this day, 18 years later, with no expectation of stopping.
One of the changes (1995) increased the number of possible area codes from 160 to 640. The new area codes seemed like an easy resource, a patch that would allow North America to retain 7-digit local phone numbers for a long time, rather than reorganizing in some other way such as adding one more digit. For years now, area codes have been added constantly, one by one, month by month, unpredictably, all over the country. Typically the new area codes are added by "splitting" existing area codes, which means that half of each existing area code changes to a new area code. Some users have had their area code changed three times since the splitting began. Almost every existing area code, whether it has already been split or not, is a candidate for another split or overlay.
I believe that the present approach of splitting and overlaying area codes is in fact the worst of all possible approaches. The time and energy and money wasted by this haphazard approach is astronomical.
I would like to compare 3 approaches to dealing with the shortage of phone numbers and exchanges:
1) Adding new area codes (splits and overlays)
2) Forced rule-based overlay
3) 8-digit phone numbers.
Existing area codes that are close to filling up are declared to be in "jeopardy" and become candidates for "relief". Hearings are held and a lot of peoples' time is wasted (and paid for) and plans are decided upon that involve either a "split" or an "overlay", or some combination that is even more confusing.
The new area codes generally have no numerical or other predictable relationship to the old area code.
In a "split", the existing area is divided into two or more new areas, one of which usually keeps the old area code, while the other area takes on the new area code. In the area that changes, MILLIONS of individuals and businesses have to re-print stationery and signs, write letters and send e-mails announcing the change of their phone number, or risk having their incoming calls misdirected. And the MILLIONS of people who call the affected numbers typically have to call twice and then update their paper and computer records. The area code change is usually phased in. As soon as the new area code exists, unused numbers and exchanges COULD be assigned in either area code. First there is a "permissive period" during which the new area code works for the changed area, but calls to the changed area using the former area code go through automatically. Then there is a less-permissive period during which calls to the changed area using the old area code get a recorded announcement. (Sometimes the less-permissive period is phased in, due to sheer volume of calls involved. For example, first 25% of calls get the recording, then 50%, then 100%.) Finally, numbers and exchanges are given out in the new area code that would have been duplicates in the old area code. Calls to numbers or exchanges in the old area code that were not re-issued usually continue to mention the area code change, just like during the less-permissive period. But if the number in the old area code has been re-issued, a wrong number will be reached.
When an area code splits, there is no easy way to update a database of phone numbers. For a data center that needs to maintain a list of national phone numbers, the only way to stay updated would be to get a map of each split, when it happens, and apply the lookups just after the new area code becomes legal, but before numbers get re-used in the old area code. For example, area code 516 splits. Exchange 735 stays in area code 516, but exchange 589 moves to area code 631. Eventually a new exchange 589 is created in area code 516. This is an unending process. The alternative is, to simply not even try to update the database, and let the wrong numbers be called and corrected on a case-by-case basis.
Sometimes splits can exploit natural boundaries and are not seen as a bad thing. For example, Manhattan remained 212 while Brooklyn and Queens became 718. However, splits often require arbitrary divisions of existing areas. Often downtown keeps its area code while a ring around it gets a new area code - unattractive and unfair. Then afterward, downtown splits anyway. One plan called for splitting Manhattan, using 42nd Street as the dividing line. Obviously that will never happen; the new area code(s) for Manhattan are overlaid. Splits have been applied until we have arbitrary regions that make no sense at all. Area code loses its meaning beacuse the areas are too small to be defined. There are too many areas and too many codes. Arbitrary splits should probably never have been done, because overlay is inevitable.
One last problem with splits: If a rate zone is split geographically, every service provider in the area code before the split immediately needs an exchange in the new area code.
In an "overlay", the existing area keeps the same boundaries, but after the overlay it has two or more area codes. No phones change their numbers, so it is unnecessary to change one's signs and stationery that carry the 10-digit phone number. It is less necessary, but still advisable, to change stationery that has the 7-digit number -- implicitly the older area code for the area is the one that will apply, but people might not be sure what that is. Also, after an overlay, it is typical that local calls will forever afterward require 10-digit dialing (AreaCode+7digit phone number), in the interest of not punishing people who have the "new" area code. Of course, dialing 10 digits increases the odds of a mistake, especially the possibility of reaching a long-distance number when attempting to dial a local number.
Overlay also removes the meaning of "area code" in two ways. There is no one code for each area if each area has multiple codes. Neighbors can have different "area codes". When local calls are specified and dialed by a 10-digit phone number, there is no area code any more, just a 10-digit local phone number. (The transition is to dialing 10 digits, not 11 digits. When the area code is always dialed, the "1" is no longer needed to indicate that an area code follows.)
Even the original nationwide toll-free area code 800 has been overlaid. Now Toll-Free is area codes 800, 888, 877 and 866 (with reservations on 855, 844, 833 and 822).
An overlay plan can also overlay one existing code with two new codes that don't overlap. This is called a split overlay. I suppose this arrangement leaves room to split the original area code later on.
This is a variation of the patchwork plan; this approach just would have made things more predictable. (It is of course too late to consider this approach.)
Every region as soon as it needs relief would be forced to accept overlay. This way, no one wastes time arguing about split and the new boundary lines or split versus overlay (saving a billion dollars) and no one changes phone numbers (saving billions more dollars). The drawback is, every overlaid area goes immediately to 10-digit dialing.
The new area codes should have had a numerical relation to the original area codes. For example, the area for 213 would have had 223, 233, 243, and 253 reserved for it, while the area for 203 would have had 293, 283, 273, and 263 reserved for it. Every area instantly has a 5-fold increase in available exchanges. (If the required increase is more than 5-fold, exceptions could be made.) Yet no areas get split up. Knowing where 212 and 213 are, everyone would automatically know where 222 and 223 are.
This approach leaves room for some splits, but only splits that fit into the old numbering scheme. For example, 212 into 212+718 would have still been reasonable (because 718 can overlay with 728 738 748 758), but 213 might or might not have split into 213+818.
Anyone who has traveled will surely have noticed that other countries in the world use other systems of phone numbers. Europe seems to like variable-length phone numbers; something America would probably not want to try. German telephone numbering has one interesting property, though. Small villages have longer area codes and shorter local phone numbers.
Increasingly common around the world is the use of 8-digit phone numbers. Shanghai, Malaysia, Singapore (2001-2002).
Japan has one-digit City Codes and, since 1991, 8-digit phone numbers, at least in Tokyo. It makes good sense; they have a small number of cities or regions, but lots of people in each of them. Having 100 2-digit city codes and 7-digit phone numbers would not make much sense in Japan.
Australia recently (1994-1997) converted to 8-digit phone numbers nationwide. They have one-digit City Codes and 8-digit phone numbers. Some rural areas converted from odd 5-digit phone numbers and the like, so they had some significant adjustments to make. For the smaller number of people calling long-distance to the small towns, tables are included in the phone directories. The big cities are another story. All the big cities already had mostly 7-digit phone numbers. After completion, the conversion there seems instant and painless, though it was phased in over several years to avoid any disruption at all, while dealing with some special cases (6-digit numbers that began with "9"). For example, in the Sydney area (City Code 2), every existing 7-digit phone number gets a "9" added before it. That's all, just one rule for one city. The other cities had similar rules. Even though everyone's phone number changed, nothing had to be reprinted, because everyone knew the rule. Existing stocks of stationery could be used up instead of discarded, because everyone knew the rule. Even as a foreigner in the country, I knew the rule. I wanted to call my airline, but the folder showed a 7-digit phone number. I simply added a "9" first and dialed the other 7 digits. No wrong numbers, no recordings. 5-year-old phonebooks are still perfectly useable (for reaching people who have not moved).
The most "drastic" solution is thus the easiest and least traumatic!
America could have, and still could, apply the same transformation. Simply change to 8-digit numbers. For every existing 7-digit number, a simple rule, prepend (for example) the first digit of the area code. Instantly the number of exchanges and phone numbers available in every area code would increase by a factor of ten. Every existing split and overlay could optionally be un-done, as they were never necessary. Not one split or overlay will be needed for a very long time, and if one is ever needed, all those unused area codes will still be waiting.
Switching to 8-digit phone numbers could have preserved every original North American area code and boundary without need of a single changed area code. People would still be able to remember more area codes and make more calls without having to look up every phone number individually.
Computer databases of phone numbers can be updated with this change, in just one pass, using such a simple rule. (Or not converted at all -- since the rule is so simple, everyone calling the numbers will know what to do.) No one would have to hurry to re-publish old phone numbers, since they the old numbers will remain understood for several years.
Area codes would remain geographically large, with fewer arbitrary boundaries. People would be able to drive 2 miles without having to wonder if they have changed area codes. People would be able to dial within a larger area without dialing an area code. No one would have to dial an area code to reach their neighbor. Everyone would dial 8 digits for a local call and "1" + 11 digits for a domestic long-distance call.
Best of all: changing to 8 digits would affect everyone equally, and it would be over all at once.
There are costs for this approach, but they are different. One cost is adjusting every telephone exchange in the country to handle and route 11-digit/8-digit calls, and then cutting over. Another cost is, some computer databases cannot be converted if they don't allow enough room for the added digit, forcing the owners to convert to new software.
I don't know what it is about America. Other countries seem to find and apply the best solutions, yet we are incapable of figuring out the same solution or admitting that someone else got it right and then copying their economical, intelligent solution. Instead we seem determined to squander the maximum amount of effort on essentially nothing.
NANPA goes on using a scheme that disrupts everyone, expensively, repeatedly, and endlessly. They split and overlay area codes in the name of retaining 7-digit local phone numbers, even though it really means 10-digit local phone numbers in every area that is overlaid. They split and overlay area codes in the name of 10-digit national phone numbers, even though their own escape plan calls for 4-digit area codes (causing a phase of mixed 10-digit and 11-digit phone numbers) once they run out of 3-digit area codes. They split and overlay area codes supposedly as a temporary measure until number portability shows up. (Number portability means portability within the same area.) Yet the splits continue, which makes the areas smaller and smaller, greatly reducing the value of portability -- users might keep the same number if they change phone companies, but they are more likely to change numbers if they move, because my area is so small. And number portability still has not shown up in a big way.
The longer-term plan calls for 4-digit area codes - when the 3-digit scheme overflows around 2025. First: phase in area codes X9XX for all existing area codes (more changes), then phase out all the 3-digit area codes, then continue expanding with 4-digit area codes (numerically related to the initial X9XX codes I think).
A lot of groups are busy managing the current numbering plan:
Someone's essay, says 8-digits can't work now, but looks forward to a future with 4-digit area codes and (of course) 8-digit phone numbers: Why Not Use 8-digit Local Numbers? [ http://www.lincmad.com/whynot8.html ]
Hits to this site since 2001.07.29:
Dan Veselitza index
posted 1999.03.01 rewritten 2002.02.10 last edited 2007.07.17