Local Number Portability
and Number Conservation

by Linc Madison <Register recently ran an article about area code splits. One aspect discussed in the article is various proposals to allocate telephone numbers more efficiently, thus slowing the pace of area code splits and overlays. In particular, the Register outlined a three‑part plan to change the number allocation method: Local Number Portability, Number Pooling, and Rate‑Center Consolidation. This article discusses various Number Conservation proposals.

Numbering, the Old Way

Each telephone number in the North American Numbering Plan (NANP) consists of three parts. There is a three‑digit area code or NPA, followed by a three‑digit prefix or NXX, followed by a four‑digit line number or XXXX. Each prefix is uniquely assigned to a particular rate center (a town or part of a city), and also to a specific operating company. In most cases, both billing and routing use the NPA‑NXX to determine the cost of the call and route the call to the correct switch. Thus, all 10,000 numbers in a given prefix must be in the same rate center and also served by the same operating company. Until quite recently, there was only one operating company in any particular rate center; however, the advent of local telephone competition means that in some cities, each rate center may have dozens of operating companies. Each of those operating companies must have an entire prefix in that rate center, even before it can sign up its first customer. This method of allocating numbers was quite reasonable when local service was a monopoly, but it is alarmingly inefficient in a competitive environment. Indeed, it is this inefficiency — not cellular phones, faxes, modems, or additional lines for homes and small offices — that is the single largest factor driving the explosion of new area codes in the late 1990's. Clearly, the current allocation method must be changed.

Number Conservation

There are three major categories into which most number conservation proposals fall:

1. Rate-center consolidation (combine two or more adjacent rate centers)
2. Number Pooling (for instance, allocating blocks of 1,000 numbers instead of 10,000)
3. Local Number Portability, or LNP (allow customers to keep the same number when changing local carriers, at least within the same rate center)

Rate-center consolidation is self-explanatory: combine two or more adjacent rate centers into one. The number conservation aspect arises from the simple fact that each new competing carrier needs only a single block of numbers to serve the new rate center, instead of a block for each of the old rate centers. However, there is considerable controversy involved in the process. Small towns often bristle at the suggestion of losing their telephonic geographic identity if, for example, Springfield and Shelbyville are combined to form the SPFLDSHLBY rate center (or perhaps a more cryptic compromise name such as MONTYBURNS). In more ambitious rate‑center consolidations, there may be adjustments required to the local calling areas and nearby toll rates, possibly with an increase in the cost of local service to offset the loss of toll revenues. In any case, rate‑center consolidation has limited applicability. In practice, rate centers can be combined only to a certain point.

Number pooling promises greater number conservation benefits. The most common proposal is to allocate blocks of 1,000 numbers to each carrier in a rate center. The entire prefix is still associated with the rate center, in order to preserve billing by area code and prefix alone. However, routing uses an additional digit of the number. Thus, NXX‑1234 and NXX‑2345 would be numbers in the same rate center, but possibly served by different companies. The benefit is clear: a new entrant ties up only one‑tenth as many unused numbers. However, this level of number pooling is only an interim measure, compared to the best solution.

Local Number Portability, or LNP, at its most basic, allows a subscriber to change local carriers while keeping the same telephone number, so long as the subscriber remains in the same rate center. In areas with LNP, every single call requires a database lookup on the full number to determine the correct carrier to handle the call. Under the old system, this information could be determined with only the NPA‑NXX; with number pooling, the routing requires the first seven digits, NPA‑NXX‑Xxxx. However, some LNP schemes still require that each carrier receive at least one entire prefix in each rate center for assignment to new customers. Only existing customers can take their numbers with them; if a customer simply turns off service on a particular number, the carrier still holds that number for reassignment.

At least for the initial implementation, LNP will not affect wireless carriers (cellular, PCS, paging, and other uses), because a number of technical and policy issues need to be resolved. However, the frequency bands for wireless communications are limited, meaning that there can be only a limited number of carriers in a given area. Thus, the benefits of applying LNP to wireless services are less dramatic than for conventional wireline services.

LNP Done Right: New York City

The LNP model advanced by the New York Public Service Commission for the New York City area codes provides that all unused numbers, both never‑used and "recycled" numbers, will be placed in a common pool. Each carrier that is licensed to do business in New York City can draw numbers from that common pool to fill orders from new customers. No carrier needs to — nor is allowed to — sreserve a large block of unused numbers. Number pooling then becomes a step backwards, since LNP means that numbers are allocated to the carriers in blocks of 1 number, rather than 10,000 or even 1,000. Rate‑center consolidation becomes far less important, since there is no longer the gross inefficiency in number allocation caused by requiring large discrete blocks of numbers for each carrier in each rate center.

LNP Solves the Whole Problem, So Why Talk about the Others?

If LNP is implemented properly, it is the complete solution to the number conservation problem. LNP must apply not only to existing subscribers changing to different local carriers, but also to unused numbers being assigned to new subscribers.

Unfortunately, that model is not being followed in other areas where LNP is being implemented. The basic form of LNP allows only existing subscribers to change carriers. If Telco X has the 234 prefix allocated to it, then an existing customer can change to Telco Y and take her 234‑XXXX number with her, but a new customer who wants a number in the 234 prefix must get it from Telco X, although that customer can then turn around and switch to Telco Y or Telco Z. Further, each new carrier must still be allocated a block of 10,000 numbers (or 1,000 numbers, with number pooling) in each rate center. Worse yet, if an overlay is implemented, the incumbent carrier has an enormous advantage in being able to assign new numbers in the familiar area code, since all subscribers who simply turn off their service return the number to the incumbent carrier's private pool to be reassigned a few months later.

Part of the problem is that, under the current prevailing model of LNP, each carrier's switch is still assigned a unique NPA‑NXX combination that is used for routing purposes in the LNP database. In other words, the LNP database looks up the dialed number NPA‑NXX‑XXXX and determines which operating company serves that number and from which switch, giving the result in the form of an NPA‑NXX routing ID for the required switch. That means that each area code can accommodate a maximum of 792 rate centers, divided by the number of local carriers. If there are 24 local carriers (a modest assumption, in many large metropolitan areas), there can be no more than 33 rate centers per area code. As new carriers are added to the mix, the area code must split, even though there may be millions of unused numbers in the existing area code.

There is only one respect in which this form of LNP achieves any degree of number conservation. Without LNP, an existing customer who changes carriers can only keep the old telephone number by using some form of transparent call forwarding from the old number to the new, meaning that two telephone numbers are tied up even though only one is in active use. The savings from eliminating that cumbersome procedure are far smaller than the savings that could be realized by using the New York City LNP model.

The Recipe for LNP

In its best form, Local Number Portability should include the following elements:

1. Routing to each local number is performed with reference to a portability database that returns the data necessary to route to the correct switch of the appropriate operating company.
2. Any existing subscriber who changes carriers can retain the telephone number, so long as the subscriber does not physically move to a different rate center.
3. All unused numbers, including numbers never used and also numbers retired from service, in a given rate center will be assigned to a common pool, administered in concert with the portability database. Whenever a subscriber terminates service on a number, that number must be returned directly to the common pool at the conclusion of the redirection ("The number you have dialed is not in service/has been changed/etc. ...") recording period. The carrier who served that number will have no preference in assigning that number to a new customer.
4. Any carrier who needs a number to fill an order for new service can draw, without preference, from the common pool. However, carriers cannot draw numbers from the pool except to fill a specific order. Periodic audits should be performed to ensure that carriers are not drawing numbers in anticipation of future orders. As an additional impediment to hoarding, the carriers could be required to specify the details of the end user whenever they draw a number from the common pool. The number could not be transferred directly to a different end user without returning it to the common pool for a minimum time period, perhaps a week or a month. No carrier would be allowed to keep any numbers in reserve. Carriers would be fined for violations of these rules.
5. The switch routing ID should be a seven‑digit number, consisting of the three‑digit NPA and a four‑digit unique identifier for each carrier's switch in that NPA. (Note that a switch may serve more than one rate center, or conversely a company may have more than one switch in a single rate area.) The identifier need not bear any relation to any prefix in active service; it will be used only for internal routing in the network. It is clear that in some NPAs, a six‑digit switch identifier (NPA‑NXX) is inadequate, due to the number of competing carriers and rate centers.

   If there is a reasonable possibility of more than 7,920 switches in a particular NPA, then one or more of the following measures could be applied:

   • Utilize identifiers corresponding to non‑dialable prefixes, such as NPA‑0XXX and NPA‑1XXX.
   • Utilize identifiers containing the four unused DTMF pairs, such as npa‑234B or npa‑5B2D. (Note that combinations such as 0XXX and 1XXX, and any numbers with A/B/C/D embedded, could be used, since these are non‑customer‑dialable internal network routing codes only. However, there may be conflicts with existing internal routing or billing codes.)
   • Borrow identifiers from an adjacent NPA.
   • Extend the format to eight‑digit routing numbers, NPA‑NXXXX.

   Each of these measures should be used in preference to effecting an area code split or overlay, unless the overall demand for numbers dictates area code relief. In other words, don't split or overlay just because you're running out of switch identifiers, if you've got plenty of numbers still available.

6. Whenever demand in a rate center requires that new prefixes be assigned, those prefixes will be assigned to the common pool for that rate center. All carriers will be on an even footing for numbers in the new prefixes.
7. One or more prefixes in a rate center may be designated by the portability database administrator for assignments in blocks of 100 or 1,000, or other convenient blocks, to facilitate serving end users with small- to medium‑sized business telephone systems.
8. An entire prefix will be assigned to a carrier only if a single end‑user will be using the entire prefix. If that end‑user ceases using the entire prefix, the prefix must be turned over to the common pool.
9. An end user may reserve a block of numbers to allow for future expansion of a PBX, or similar application, but the carrier for that end user cannot take numbers from that reserved block to serve other end users. If the end user agrees to return a portion of a reserved block, those numbers must be returned to the common pool.

Cost/Benefit Analysis

The LNP scheme outlined above will cost more to implement than the most basic LNP. The portability database will be required not only to do real‑time lookups to determine the routing for call setup, but also to maintain a list of all unused numbers in each rate center and their assignments to end users. There will also be costs associated with auditing and enforcing compliance with the rules proposed here, as well as ongoing administrative costs for hardware and personnel. Those costs will be substantial. There will also be administrative costs in changing the switch identifier to seven digits (NPA‑NXXX) instead of six, and additional costs if that format requires further extension.

However, the benefits to be gained from this scheme are enormous. Area code 847 in northeastern Illinois is currently slated to overlay in April 1999, even though the Illinois Citizens Utility Board estimates that half of the possible numbers in 847 are not in use. Similar stories can be found in almost every state. Whether area code relief comes in the form of a traditional geographic split or an overlay, there are significant costs to individuals and businesses, as well as to carriers themselves. The formula for LNP shown here will ensure that area code relief will be imposed only when it is truly needed, when the area code is experiencing a genuine shortage of available numbers. The substantial savings from avoiding unnecessary area code splits and overlays will offset much of the added cost of the improved LNP scheme. Most especially, this improved LNP scheme may enable the NANP to prolong the life of the current ten‑digit telephone number format, currently projected to run out of numbers in about the year 2025.

Who Would Make the Decision?

The Federal Communications Commission has mandated LNP, with a schedule for implementation, beginning in the largest metropolitan areas. Presumably the FCC would also be within its authority to specify additional requirements for LNP. If the FCC fails to take such action, it would also be within the purview of the individual state public utilities commissions, since the FCC guidelines provide only a set of minimum requirements.

Return to LincMad's Telephone Area Code page.

LincMad.com received over a million visits from 1997 until the counter broke.

Check out my political blog at LincMad.blogspot.com.