iassia quarterly - 19

Public Data Use:
A View From The

Telecommunications Industry

In The United States

by A. Dianne Schmidley,

Staff Manager, Bell Atlantic

Bell Atlantic is one of eight U.S.

telecommunications firms resulting from the

breakup of the AT&T owned Bell System on

January 1, 1984, the largest divestitiu-e and

reorganization in corporate history. Bell Atlantic

owns an assortment of companies engaged in

various aspects of providing telecommunications

services and products. These companies can be

divided into two categories which we call the

"Enterprises Group" and the "Network Services

Group."

The Enterprises Group provides services and

products to a variety of geographic locations in

the United States and Canada and it is a

relatively unregulated entity. Activities of the

Network Services Group are concentrated in the

seven politica] jurisdictions of Washington D.C.,

Delaware, Maryland, New Jersey, Pennsylvania,

Virginia, and West Virginia. Management
Services, Incorporated (MSI) and the operating

telephone companies are part of the Network

Services Group.

The Economic Analysis District (EAD), my
organization, is located in the Business Planning

and Financial Management Department of the

MSI. The primary occupation of staff members
in the EAD is the provision of internal

consulting support to the Network Services

Group, which is dominated by the concerns of

the operating telephone companies: the

Chesapeake and Potomac Telephone Companies

of Washington, D.C., Maryland, Virginia, and

West Virginia, New Jersey Bell Telephone

Company, Bell of Pennsylvania, and Diamond
State Telephone Company in Delaware. These

concerns can be divided into four functional

areas: Regulatory, Personnel, Facilities Plaiming.

and Marketing.

Because the network services group is almost

wholly comprised of regulated telephone

companies, "regulatory issues" are the most

important concern of the EAD. State regulatory

agencies, often called Public Utility

Commissions, determine, through pricing

decisions, who will bear the burden of the rates

the companies charge to recoup operating

expenses and guarantee the investors in Bell

Atlantic stock a competitive rate of return on

their investment dollar. Demographic and

economic analyses of the size, distribution and

composition of each company's market provide

the basis for determining the effects of various

pricing configurations.

Summer 1986



20 - iassist quarterly

Analytical work undertaken by Bell Atlantic

economists and demographers has been made
more complicated by the divestiture, since the

breakup of the Bell System hterally led to the

breakup of telephone-served geography. The

seven jurisdictions we serve contain 19 "local

access and transport areas" or LATAs, which do

not correspond to any other political or

statistical entity, although they are associated

with Metropolitan Statistical Areas (MSAs) in

many cases. These LATAs, or large exchange

areas, obtain their external communication links

from the Interexchange Carriers (lECs),

telecommunications companies engaged in long

distance calling services which the local

telephone companies (such as those owned by

Bell Atiantic) are constrained from offering,

owing to federal regulatory restrictions.

Relations between the lECs and the local

telephone companies are regulated by Federal

Communications Commission (FCC) rulings,

legislative requirements established by the U.S.

Congress, and Executive Branch decisions made
through entities such as the Justice Department

and the Federal Coiut System. In order to

comply with the various rulings and legislative

mandates, and sometimes question their logic.

Bell Atiantic must have knowledge of the size,

distribution and composition of the market

within and between the LATAs.

In addition to the Regulatory function, another

important activity of the companies the HAD
supports is the Personnel function. Whether we
are addressing Equal Employment Opportunities

issues, employment site location studies, force

planning or how to strategically locate our work

crews relative to population growth and

migration chum, we turn to data available in

the public domain to answer questions about the

size, distribution and composition of the local

labor force and the telephone-served population.

Our third major area of support for the

operating telephone companies involves the

"Facilities Planning" function. The telephone

companies operate networks, which consist of

central offices containing switching systems

(large computers), miles of cable, and microwave

towers. We are constantiy concerned with plant

capacity and demand for our services which

translates into changes in demand for central

office switching and transmission capabihty.

Population and economic forecasts based on

public data make the forecasting of demand
possible and enhance oiu ability to plan

efficientiy and effectively.

Although we are heavily regulated by the

government, we do have many marketing

concerns, and the market we serve constitutes

the fourth area of functional responsibility for

the EAD. Some of the more familiar marketing

efforts the telephone companies engage in

include the distribution of the white and yellow

pages directories, and the provision of operator

services such as call completion and information

retrieval. In addition, we offer products such as

business to business directories, and services

such as cable television access. We serve the

government at the national, state and local level.

We serve large industries, such as the steel mills

in West Virginia and Pennsylvania, and small

enterprises such as a savings and loan company
in Maryland. We serve the elderly, the

handicapped, homeowners, travellers, and, a new
customer since the divestiture, Interexchange

Carries [sic]. Knowledge of the consitutents of

this market is derived from public data coupled

with our own internal surveys.

What are the kinds of pubUc data used by Bell

Atiantic? Generally, we use as much of the

demographic and economic data as we can

obtain from the federal, state and local

governments, whether it comes from censuses,

surveys or administrative records, but our most

important source of demographic or

socio-economic information is the 1980 U.S.

Census of Population and Housing. These data

are available in many forms: published, on

microfiche, and on magnetic tapes. The

problem is that there is more data than we can

handle, so we have implemented an online

Summer 1986



iassist qvuirterly - 21

demographic data retrieval system to assist us.

As I mentioned earlier. Bell Atlantic has a

imique problem. The divestiture left the

company with odd service areas called LATAs.

In order to provide information to our

companies, the EAD modifies public data from

the economic and demographic censuses and

surveys to make it conform to the geographic

area Bell Atlantic serves.

Prior to the divestiture, the operating telephone

companies were concerned with the same

geographical areas they serve today. The imit

of concern, however, was the wire center area

or central office district (COD) as it is

sometimes called. To complicate matters

further, local exchange areas (smaller and

different from the LATAs described above)

were also a concern. Fifty years ago, all three

entities were represented by the same

geographic area, corresponding to a community

or settlement Technological change, which

allowed the newer central offices to serve more
than one of the old wire center areas,

population change, and concessions to consumers

with regard to their calling access led to an

erosion of this one-to-one conespondence. As

a result, the telephone companies not only have

served and continue to serve areas unlike any

other known pohticjil or statistical geographical

areas, they serve a number of entities that do

not correspond to one another.

Because of the continuous need to determine

the demographic/economic characteristics of

telephone service areas, in order to address the

functional areas described above, the

requirement for tailored public data arose long

before the divestiture. The key to tailoring the

demographic and economic data used to develop

construction plans, engage in force planning and

answer the questions of the regulators is Census

geography.

Census tract and block group information,

aggregated to user described areas, is the

Rosetta stone of managers engaged in economic

and demographic analysis. With the advent, in

1970, of the first fully automated census, the

laborious task of aggregating census tract and

block group information by hand became,

mercifully, obsolete.

Today, there are three major methodological

approaches underlying the automated

demographic data retrieval systems which

provide information for user defined geography.

1. Federal Information Processing Codes (FTPS)

are assigned to every pohtical and statistical

entity in the United States. This means that all

political and statistical geographic units, such as

states, counties, MSAs, and census tracts, have

unique identification codes. In the automated

system, the user can retrieve information

associated with these codes. This approach is

efficient if the user is seeking information for a

list of states, coimties, or municipalities. The

first attempts to aggregate data for user

described areas, such as wire center areas, were

based on combinations of block groups/census

tracts, and relied on this mechanism. When
thousands of geographic units were involved,

however, (the old Bell System had 10,000 wire

center areas) this particular approach proved to

be extremely time consuming, even after

automation.

2. The assignment of geo-coordinates

(latitudinal/longitudinal coordinate points) to

census data provided the basis for a major

breakthrough in the automation of demographic

data retrieval. Every census block in the United

States received, in 1970, a centroid assignment

of a unique set of coordinate points. The

centroid is the geographic or population center

of a block area; there are variations in the way

these assignments are made, but discussion of

this topic is beyond the scope of this paper. In

1970, point assignments developed by the U.S.

Bureau of the Census, were listed in the Master

Enumeration Districts List (MEDS) and in 1980,

Census Bureau point assignments were listed in

Summer 1986



22 -
iassist quarterly

the Master Area File Reference List (MARF).
In 1990, they will probably be found in the

Topographically Integrated Geographic

Referencing and Encoding System (TIGER).

In themselves, the centroid assignments are

useless for solving the problem of

demograhically describing user defined areas.

Software linking the coordinate assignments and

user described boundaries of study areas, which

have been transcribed into binary code, are

needed to complete the demographic data

retrieval operation. To date, most of the

software for this type of application is owned
by non-governmental sources, and licensing

arrangements must be purchased in order to

make use of the private sector product Before

the divestiture, the operating telephone

companies in the jimsdictions now served by

Bell Atlantic had transcribed their wire center

area boundaries into binary coded polygon files.

Census data based on the MEDS and MARF
assignments could be aggregated to produce

demographic profiles of the user described

areas. Since LATAs are aggregations of wire

center areas, all that had to be done after the

creation of the LATAs was to aggregate the

wire center polygons into LATA polygons. As
was mentioned earlier, LATAs axe also

aggregations of the smaller exchange areas,

and/or COD areas. At the LATA level,

however, the difference between the three

telephone entities (wire center areas, central

office districts, and exchanges) disappears. Thus,

aggregating the wire center areas to equate to

LATAs does not cause discrepancies. The final

result is census data tailored to our LATA
areas.

3. The third type of geographic linking system

available for tailoring govenmicnt produced

demographic data to meet user defined needs is

the Geo-Based Files/Dual Independent Map
Encoding or GBF/DIME process. Briefly, this

process makes possible the matching of Census
address records for urbanized areas with user

records. In the case of the telephone company.

these are customer records. Customer records

processed through the GBF/DIME program can

be linked, at the census tract level, with specific

socio-economic characteristics. This process was
used to provide the Washington, D.C. Public

Utility Commission with information concerning

hnks between telephone availability and

characteristics of the inhabitants of areas under

study. This process is more limited than the

other two, however, since the GBF/DIME files

are only available for urbanized areas.

At Bell Atlantic, economic data available from
the government for political or statistical areas

are disaggregated into user defined telephone

service areas through the use of population

weights derived from the centroid point

assignment process described above, or the FEPS
code process. Since economic data are available

from the government for the whole counties

contained in the LATAs, disaggregation only

occurs in the case of spht counties. The census

tract components of counties are assigned to

their respective LATAs using the procedures

outlined in 1. and 2. above.

The census profiles developed through the use

of the geographic Unking systems provide the

basis for developing time series data and

forecasts of population through iterative

proportional fitting schemes, when linked to

historic and forecast information for the

aggregates of counties which correspond to the

LATAs. Economic data, in turn can be derived

through the use of the time series and forecast

versions of the population weights.

Thanks to the geographic linking processes

developed jointly by the government and private

sector firms with software capabilities. Bell

Atlantic is able to address problems in the

major corporate functional areas outlined earlier

in this paper, utilizing public data as they relate

to our odd geographic areas, n

Summer 1986