Interim Report Concerning The Definition of Rural Areas

Prepared by the Subcommittees on Rural Health Care and Schools and Libraries

Pennsylvania Universal Telephone Service Task Force

Adopted July 14, 1997

Introduction and Background

In order to implement universal telephone service for health care providers, the Federal Communications Commission ("FCC") adopted a $400 million program comprised of three components: (1) all public and non-profit health care providers that are located in rural areas and meet the statutory eligibility criteria may obtain universal service support for telecommunications services up to and including a bandwidth of 1.544 Mbps by obtaining a price for service that is comparable to the price charged to urban health care providers; (2) rural health care providers may obtain a reduction to the distance charges incurred, compared to the distance charges incurred by urban health care providers; and (3) all health care providers--both urban and rural--may obtain support for toll-free access to an Internet service provider.

The principal aim of the federal program is focused on health care providers located in rural areas. The FCC adopted a definition of rural area to mean a nonmetropolitan county or county equivalent, as defined by OMB and identifiable from the most recent Metropolitan Statistical Area ("MSA") released by OMB or any census tract or block numbered area, or contiguous group of such tracts or areas, within an MSA-listed metropolitan county identified in the most recent Goldsmith Modification published by the Office of Rural Health Policy/Health and Human Services ("ORHP/HHS"). There are two main methods of defining rural and urban areas, according to the FCC: the Bureau of Census designation of rural and urban areas based on density, and metropolitan and nonmetropolitan areas based on the integration of counties with big cities. The FCC accepted the ORHP/HHS methodology because counties are units of identification more easily used and administered than the Bureau of the Census' density-based definition of rural and urban areas. The Goldsmith Modification identifies small town and open-country parts of large metropolitan counties by census tract or block-numbered area, as defined by the Bureau of the Census.

Pennsylvania is home to 3.7 million residents that live in rural areas according to the definition of rural used by the Bureau of the Census--the most in the nation. Although typically not thought as a state with a large rural constituency, the statistics reveal just the opposite. Consequently, the definition of rural area is of great importance to our state, so that we can be assured that the benefits of the universal service programs are made available to as many rural entities as possible. It should be noted that the FCC adopted the same definition of rural areas for purposes of administering the schools and libraries discount program. Thus, this issue relates to both the rural health care program and the schools and libraries program.

Under the FCC's approach, metropolitan counties are considered urban and non-metropolitan counties are considered rural. The Census Bureau defines a metropolitan area as one or more contiguous counties surrounding a central city of 50,000 or more. Outlying, contiguous counties are included in a metropolitan area based on their population density, growth rate, commuting patterns, and other factors. All counties not identified as part of a metropolitan area are considered non-metropolitan.





The Subcommittee on Rural Health Care was assigned the responsibility of analyzing the FCC's definition of rural areas to determine whether the definition was consistent with the Commonwealth's needs and objectives. The advice of numerous experts on rural issues was solicited: the Pennsylvania Rural Development Council (a sitting member on the PUC's Task Force); the Center for Rural Pennsylvania; the Pennsylvania Office of Rural Health; the Commonwealth's Department of Health, the Hospital and Health Systems of Pennsylvania; the American Association of Retired Persons. These rural specialists comprehensively analyzed the FCC's definition and concluded that it did not meet its intended objectives. An alternative method of classifying rural areas is proposed in order to assure that all of Pennsylvania's rural health care facilities may be eligible to benefit from the federal universal service program.

Application of the FCC's definition to Pennsylvania's 67 counties results in the exclusion of nine counties which are typically considered to be rural. This Interim Report sets forth a comprehensive explanation of the Task Force's concern that the FCC's definition of rural areas is too narrow to adequately meets our state's concerns. Accordingly, the Task Force recommends that the PUC petition the FCC for waiver or reconsideration of its rural definition so as to classify the nine affected counties as rural. This relief would apply not only for purposes of the rural health care program; also it would apply to the schools and libraries discount program.

It should be noted that these rural specialists considered several other logical and defensible methods for defining rural areas, all of which classified more than nine additional counties as rural. Instead, the group analyzing this issue chose to focus on the nine counties which are the most demonstrably rural in character.



Which nine counties are at issue?

Under the FCC's definition, 31 counties are classified as metropolitan, and therefore, urban; and the remaining 36 counties are classified as non-metropolitan, and therefore, rural. Under the alternative method suggested for Pennsylvania, there would be an additional nine counties classified as non-urbanized, and therefore, rural.

The nine counties at issue are: (1) Butler; (2) Carbon; (3) Columbia; (4) Fayette; (5) Lebanon; (6) Perry; (7) Pike; (8) Somerset; and (9) Wyoming. Each of these counties is classified as urban areas under the FCC's definition. Yet, according to the rural experts' consensus opinion, these nine counties share more in common with their non-metropolitan counterparts than with the other metropolitan counties, and have a rural rather than urban character.

Rationale for Classifying the Additional Nine Counties as Rural

In Pennsylvania, several of the counties which are classified as metropolitan share more in common with their non-metropolitan counterparts than with the other metropolitan counties. That is, many of Pennsylvania's metropolitan counties have a rural rather than an urban character. This is true not only of the distribution of health care providers and the telecommunications infrastructure within these counties but of their general culture and population composition.

The nine counties when compared to the other 24 metropolitan counties classified as urban under the FCC's definition have, for example:

A significantly lower primary care physician to population ratio;

A significantly higher proportion of residents living within designated areas of medical underservice;

Significantly fewer hospitals and hospital beds;

A significantly lower health care provider to population ratio for all types of providers;

A significantly lower per capita income;

A significantly higher population growth rate;

Lower per capita federal transfer payments.

In determining to use the metro/non-metro MSA classification system for differentiating between urban and rural areas, the FCC relied on the Joint Board's recommendation regarding this subject. The Joint Board expressly acknowledged that the comments of the Office of Rural Health Policy/Health and Human Services which stated that no method for defining "rural" is perfect; each method has deficiencies or problems. The Goldsmith modification was accepted by the Joint Board (and later by the FCC) as the means for classifying as rural large, nominally metropolitan counties particularly in western states which contain significant rural areas that are isolated and lack easy physical access to the central areas of metropolitan counties for health care services. The Goldsmith modification, however, only classifies a portion of one additional county in our state--Lycoming County-- as rural when it would otherwise be classified as urban. The Joint Board's rationale for accepting the metro/non-metro MSA approach with the Goldsmith Modification is instructive in setting forth inherent limitations on the accuracy of this methodology:

For the task of determining the size and boundaries of the rural areas in a state, we believe it is appropriate to use a method that seeks to include as many of the truly rural areas as possible. We agree with OHRP/HHS that no currently-used method of designating rural areas is perfect. We conclude, however, that the OMB MSA method is, by itself, under-inclusive of many rural areas and therefore does not meet the standards set by the Commission in the NPRM. The Goldsmith Modification, by identifying by census tract or block more densely-populated areas in large, otherwise rural counties somewhat ameliorates this problem. This method meets the "ease of administration" criterion as well. Lists of MSA counties and Goldsmith-identified census blocks and tracts already exist, updated to 1995. Through the use of these lists, any health care provider can easily determine if it is located in a rural area and therefore whether it meets that test of eligibility for support.

Universal Service Joint Board Recommended Decision, ¶694 (November 9, 1996) (emphasis added). The FCC accepted this recommendation of the Joint Board. Universal Service Report and Order, ¶649.

It is clear the MSA metro/non-metro classification with the Goldsmith Modification does not sufficiently ameliorate the concern of accurate classification of rural counties for our state.

The additional analysis which led the Pennsylvania rural specialists to identify these nine additional counties as rural was based on a statistical review of the counties as well as an examination of what characteristics appropriately measure urban and rural areas. The recommendation is to consider urbanization as the primary means of defining urban areas:

A county is considered urban if 50 percent or more of its population resides within an urbanized area. In addition, any central county of a metropolitan area is also considered urban. All counties not defined as urban by this definition are considered rural.

The Census Bureau defines an urbanized area as the central city of a metropolitan area and all contiguous areas which have a population density of 1,000 more persons per square mile or are highly connected to the area by vehicular roads. In a few instances, the central county (the county in which the metropolitan city is located) is less than 50 percent urbanized. In these counties the strong urban and local presence of the metropolitan center results in a county with an urban rather than a rural character. The concept of "urbanized" is more highly consistent with urban culture and more closely corresponds to the service infrastructure which characterizes urban areas.

These nine additional counties should be classified as rural, either on the basis of a waiver of the FCC's definition to allow for this outcome, or alternatively via the FCC's reconsideration of the measure of rural/urban to encompass the additional urbanization criterion set forth above. Because we are not familiar with the circumstances of other states, we do not advocate that the rural/urban definition be modified for purposes of the entire federal program. If, however, the FCC believes such an approach is preferable to a waiver, we would have no objection to such an outcome.

What is the fiscal impact on the Federal Health Care Program of including the nine additional counties within the definition of rural?

For the rural health care program, the fiscal impact is estimated to be less than 2/10 of 1% of the overall cost of $400 million for the federal program. A "priceout" of this recommendation was undertaken to determine whether it was financially feasible. Recognizing that concerns over fiscal management led the FCC to impose a $400 million cap on the rural health care program, the Task Force was very concerned that endorsement of this recommendation could not be even considered unless a "fiscal impact" analysis was conducted and presented to the FCC. The Task Force is confident that its recommendation can easily be accommodated within the existing parameters of the $400 million cap.

The Subcommittee on Rural Health Care identified the components of the federal program which would be financially impacted by this recommendation: the rate averaging provision which provides for rural health care providers to receive a rate that is comparable in price, including an allotment of mileage charges, to the price charged to urban health care providers for commercially available telecommunications service up to a 1.544 Mbps (T-1) capacity. Note that the toll free Internet access provision has no financial implications for this proposal since all eligible health care providers, both urban and rural, are able to receive such benefit.

The following methodology was employed to calculate the fiscal impact of this proposal.

1. All eligible health care providers in the nine counties were identified from various public sources:

o Rural health clinics, community health centers, and migrant health centers were obtained from Community Health Centers and Other Federally Affiliated Clinical Sites in Pennsylvania, Schwartz, Mike (April, 1997). University Park; Pennsylvania Office of Rural Health. The source list includes all grantees under the PHS Act, members of the Pennsylvania Forum for Primary Care and all Medicare certified providers.
o Non-profit hospitals were obtained from the Pennsylvania Department of Health. Note that there are no county or municipal health departments in these counties.
o Post-secondary educational institutions were obtained from an Internet search. A listing of all 2 and 4 year colleges was located; institutions in the nine counties were identified and a web site for each institution was identified. The web site was scanned for a listing of programs offered. Each institution which offered a health care provider program was included in the list (these include nursing, nurse practitioners, others).
o This data was configured for use by a geocoding program.
o A total of 46 providers for all nine counties were identified.


2. All locations within the state with populations of 50,000 or more were identified: Pittsburgh city; Penn Hills Township; Reading city; Altoona city; Bristol Township; Bensalem Township; Harrisburg city; Upper Darby Township; Erie city; Scranton city; Lancaster city; Allentown city; Lower Merion Township; Abington Township; Bethlehem city; Philadelphia city.

3. GIS Methods

o Addresses from the above data file identified under step no. 1 above were geocoded using the Streetbase geocoder. This geocoder operates within an ATLAS GIS shell. The status of the address matching resulting from the geocoding is archived in the variable "code" in the final database. A "I" indicates a ZIP Code only match, other codes indicate various levels of address matching. Latitude and longitude coordinates were stored in the database for each eligible provider.
o The points were then mapped using ATLAS GIS and an algorithm constructed to calculate the distance to the closest population center with 50,000 or more residents. These distances measured the distance from the health care provider to the centroid of each of the population centers of 50,000 or more.
o The calculation of the minimum distances was computed.


4. The maximum diameter of each of the locations with populations of 50,000 or more was computed by the Pennsylvania Department of Transportation, Cartographic Information Division, using Integraph Microstation Design files; and digitizing boundary lines from existing general highway series maps (approximate scale: 1 inch = 1 mile).

5. For each of the 46 eligible health care providers identified in step no. 1 above, the incumbent local exchange telecommunications company (ILEC) was identified, and the ILEC was identified for each location with a population of 50,000 or more.

6.

A. The maximum distance between the health care provider and the farthest point on the boundary of the closest location with 50,000 people was computed. Because the distance in step no. 3 computed the distance between the health care provider's location and the centroid of the location with 50,000 people, the radius of the maximum diameter of each location with 50,000 persons or more was added to the distance in step no. 3 to arrive at the maximum distance.
B. The maximum diameter of each location with 50,000 persons or more was deducted from the maximum distance between the health care provider and farthest point on the boundary of the closest location with a population of 50,000 person or more, to arrive at the distance which would be subsidized from the federal universal service program via the mileage charge provision.


7. The T-1 rates for each ILEC was compared to the T-1 rates applicable to the locations with populations of 50,000 or more, and all differences were identified and quantified. For example, the local channel charge for T-1 is higher in rural areas than in urban areas.

8. The T-1 rates were then computed for each eligible health care provider based on the above steps.

The total additional cost for including the eligible health care providers of the nine additional counties within the rural definition is $475,087, or less than 2/10 of 1% of the $400 million cap for the program. We believe that this recommendation, therefore, can be accommodated within the existing program and will not necessitate any additional financial resources to be committed by the FCC. We clarify that it is imperative for the FCC to resolve this concern because the nationwide rural health care program is being funded from assessments on both interstate and intrastate revenues of providers of interstate telecommunications services. Consequently, it is critical that the rural health care providers located in these nine counties be placed on the same footing as the rural health care providers located in the counties that already are classified as rural.

What is the fiscal impact on the Schools and Libraries Discount Program of including the nine additional counties within the definition of rural?

The Subcommittee on Schools and Libraries submits the following assessment:

By reclassifying these counties, there assuredly will be a financial impact to the annual Universal Service Fund of $2.25 billion. In order to estimate this cost of the alternative definition, the following rationale was used:

Schools:

The FCC Order estimates that schools will spend $3.0 billion annually to purchase the technology services eligible for discounts. The weighted national average of discounts is 60%, thus discounts on those services will cost $1.8 billion. If $1.8 billion is divided by the total number of schools, 113,000, the approx. discount for each school is $15,929. Because we know that the most a school's discount can increase by reclassifying its county is 10%, we can then determine that $1,592.92 is the average amount that each of those districts will benefit under the new definition. We then multiply $1,592.92 by the number of schools in those nine counties (317) to calculate the approx. cost = $504,955.



Libraries:

The calculation is the same, assuming that $180 million is the estimated amount that libraries will spend annually to purchase technology services eligible for discounts. The weighted national average of discounts is 60%, thus discounts on those services will cost $108 million. If $108 million is divided by the total number of libraries, 15,000, the approximate discount for each library is $7,200. Because we know that the most a library's discount can increase by reclassifying its county is 10%, we can then determine that $720 is the average amount that each of those libraries will benefit under the new definition. We then multiply $720 by the number of libraries in those nine counties (55) to calculate the approx. cost = $39,600.

Therefore the approximate impact of the alternative definition of rural is $544,555, which has a relatively smaller impact on the E-Rate schools and libraries discount program than the impact felt on the health care fund. The fiscal impact is less than 3/100 of 1% of the $2.25 billion E-Rate program.

Because these calculations were done using a weighted average, the cost is only a good estimate. These calculations are likely to be higher than the actual cost impact on the program because the methodology assumes that all schools and libraries will receive a 10% increase in discounts. However, we know that schools and libraries that fall within the two most economically disadvantaged categories will not receive an increase in discount, since there is no difference between the rural and urban discount for those two levels. A more detailed analysis of the financial impact is being prepared by the Center for Rural PA and should be available by the end of the week. If available, the report will be issued at the July 14 Universal Service Task Force meeting.

Task Force Recommendation

The Universal Telephone Service Task Force recommends that the Pennsylvania PUC submit a petition for waiver, or in the alternative, reconsideration of the rural definition to permit the additional nine counties to be classified as rural. The Task Force recommends that this Report be attached to the Petition submitted to the FCC, and that the Petition be filed by no later than July 17, 1997. This Task Force did not examine the desirability or need for intrastate support for this or similar programs. No party to the Task Force waives its right to develop and support its own position if the Pennsylvania PUC determines that it wishes to examine this issue in the future.