Please provide comments to the issue below as part of the 2012 WCB cost model virtual workshop for inclusion in the record. Comments are moderated for conformity to the workshop’s guidelines.

Background

Hybrid Cost Proxy Model: When adopting the HCPM, the Commission concluded that it should use geocode data to the extent available for determining customer locations. The Commission later affirmed this conclusion, but found that "no source of actual geocode data has yet been made adequately accessible for public review." The Commission expected that a source of accurate and verifiable actual geocode data would be identified in the future for use in calculating federal support.

In the absence of such data, the Commission used road network information as the basis for locating within a census block boundary customers whose precise locations were unknown. This approach entails associating customer locations with roads and excluding some road types or road segments (e.g., interstate highways, bridges, and on- and off-ramps) that are unlikely to be associated with customer locations. The Commission used an algorithm proposed by PNR Associates to distribute locations uniformly on road segments (based on Topologically Integrated Geographic Encoding and Referencing (TIGER) data). The PNR Associates' National Access Line Model is used to estimate the number of customer locations within census blocks and wire centers. The PNR Associates model's information sources include: survey information; the Local Exchange Routing Guide (LERG) database; Business Location Research (BLR) wire center boundaries; Dun & Bradstreet's business database; Metromail's residential database; Claritas's demographic database; and U.S. Census Bureau estimates.

CQBAT: The CQBAT model estimates the number of customer locations using multiple sources. These include obtaining the number of residential locations from commercial data (GeoLytics) and obtaining the number and distribution of business locations from commercial data (GeoResults). CQBAT also determines the type (SIC code) and size of businesses (used to estimate the bandwidth required at each location) from the US BLS Economic Census at the ZIP-code level, using the commercial data to distribute ZIP-level counts to census blocks. The model distributes locations randomly along roads, excluding some road types and road segments (e.g., interstate highways, bridges, and on- and off-ramps). CQBAT's approach to placing customer locations randomly along roads appears to be consistent with current best practice and represents an improvement relative to the HCPM. This distribution has the effect of putting more customer locations in areas with higher road density—i.e., more homes near a town grid and fewer along open rural roads. This approach is similar to the approach used in the Commission's Broadband Assessment Model, used during the National Broadband Plan, and to the approach used in the National Broadband Map (to estimate coverage in blocks greater than two square miles).

Questions for Comment

1. Is there any reason to deviate from CQBAT's approach? Are there any improvements to this approach that would be beneficial to incorporate?
2. An alternative approach is to determine customer locations by only using commercial data from GeoResults that provides actual and estimated customer locations. Does this approach offer any significant advantage over CQBAT's method?