Infrastructure, Deregulation, Universal Service, Universal Access

Consumers in all regions of the Nation, including low-income consumers and those in rural, insular, and high-cost areas, should have access to telecommunications and information services, including interexchange services and advanced telecommunications services that are reasonably comparable to those services provided in urban areas and that are available at rates reasonably comparable to rates charged for similar services in urban areas. . . . [1996 TCA, section 254]

The term advanced communications capability is defined, without regard to any transmission media or technology, as high-speed, switched, broadband telecommunications capability that enables users to originate and receive high-quality voice, data, and graphics, and video telecommunications using any technology. [1996 TCA, section 706]

With those words, the U.S. Congress enshrined into law a commitment establishing nationwide access to a broadband hypercommunication network that would serve all areas of the country, including rural areas. The new legislation represented an important departure from the previous state and federal approach to communications because of sweeping deregulatory provisions combined with an attempt to reduce regulatory asymmetries. Under the 1996 Telecommunications Act (TCA), rural interests (including agribusinesses, their customers, and suppliers) have the right to access a greater range of hypercommunication services than the traditional telephony services they had been entitled to under the 1934 Communications Act. Additionally, the new, high-speed services had to be offered at prices that were "roughly comparable" to those offered to customers in urban areas.

Chapter 5 discusses the how and where of hypercommunications infrastructure and regulation. How refers to how deregulation, re-regulation, taxation, and policy goals such as universal access and universal service will be used as mechanisms to help determine where infrastructures are deployed and access is available. The economics of policy responses influence how and where infrastructures will be developed, and what kinds of hypercommunication services will be available to Florida's agribusinesses.

Details of specific infrastructures were found throughout Chapter 4, especially 4.3.2 (telephone), 4.3.3 (Cable TV), 4.3.4 (fiber backbones), and 4.4 (wireless). Service availabilities and the unique hypercommunication needs of Florida's agribusinesses are illustrated in detail in Chapter 6. However, the availability of access to high-speed networking services may be determined by hypercommunications policy in some parts of Florida. A related question to be covered in Chapter 5 is how much competition there will be in the provision of hypercommunications to agribusiness and rural areas. Part of this answer depends on regulation, while part depends on the hypercommunications market structure, which is covered in Chapter 7.

As hypercommunications become more of a requirement for profitability and growth in a world economy linked by better communications, rural communities and producers without adequate access to hypercommunications could become marginalized. Furthermore, agribusinesses that must rely on infrastructures in such under-served areas are a disadvantage compared with domestic or international competitors who enjoy better access.

The Chapter begins with a brief overview of rural Florida's hypercommunications environment in section 5.1. Section 5.2 covers the meaning of other central concepts of the regulatory setting: a discussion of what the term rural means, along with definitions of infrastructure, universal service, and universal access. Section 5.3 traces the regulatory history of wireline and wireless communication infrastructures, pointing out how the converging hypercommunication sub-markets (mentioned in the latter half of Chapter 4) diverge from the market definitions of the regulators. Section 5.4 considers several rationales for government involvement in hypercommunication sub-markets. Brief economic analyses of specific mechanisms that are being used currently to extend service to rural areas and regulate hypercommunications in general are given in 5.5. Section 5.6 presents a conceptual model that is then used to analyze three general topics in the economics of utility regulation given that convergence and technical change are outpacing regulation. Finally, Section 5.7 discusses several minor rural development measures some that can directly help agribusinesses with hypercommunication planning and needs. Since regulation plays such an important role in geographical market definition, Chapter 6 will use results from Chapter 5 to look at how market boundaries make the local hypercommunication market facing Florida agribusinesses difficult to regulate and define.

5.1 Overview of Florida's Rural Hypercommunications Environment

For many agribusinesses, all or part of their ability to hypercommunicate depends on whether carriers extend inexpensive high-speed access to services in rural areas. However, the cost per subscriber of the rural hypercommunications infrastructure is high compared with urban areas, due to low population densities and other factors. This section gives a brief overview of how high-speed network access technologies affect the rural hypercommunications environment.

One way to understand the magnitude of the problem of extending hypercommunications to rural areas is by examining available data. An excellent source to start with is the FCC's HPCM (Hybrid Proxy Cost Model) data on all telephone, T-1, ISDN, and DSL lines (local loops) in Florida [FCC, CCB, 2000]. HPCM is a model used to estimate the forward-looking cost of providing "advanced telecommunications services" to high-cost areas [Bush et al., 1998]. The FCC and state utility commissions gathered information about the locations of all local loops throughout every state. The HPCM data are summarized at the exchange or wire center levels on down to small GIS (Geographic Information System) clusters.

Some of the extremes shown in Table 5-1 help to frame the issues involved in rural areas. The heart of the wireline access policy problem is summarized through five variables: average cost per loop of providing, average copper loop length, maximum copper loop length, total area of each exchange or cluster, and line density per square mile.

Source: FCC Common Carrier Bureau, 2000.

First, average costs per loop in rural exchanges tend to be high when compared with urban ones. It costs over ten times more per local loop to provide POTS (Plain Old Telephone Service) to Kenansville than it does to parts of Jacksonville. When it comes to DS-0 and higher dedicated lines and circuit-switched circuits the differences in costs are even larger. The investment cost of such dedicated and direct circuits ranges from $0.40 per line to over $600 per line per month (in addition to loop costs), depending on the area's infrastructure and average connection length [FCC, CCB, 2000]. Such dedicated and direct loops provide businesses with access to enhanced telecommunications services, private data networking, and the Internet. With these dramatic differences in costs, it is hardly surprising that the full menu of enhanced telecommunications and high-speed data and Internet access are unavailable via telco lines in many areas of Florida.

The area covered by a CO or wire center is another important variable. For example, as Table 5-1 shows, the Lake City CO serves over 400 square miles, while certain COs in both Miami and Jacksonville serve areas smaller than two square miles. Large serving areas are correlated with longer loop lengths. Loop lengths stretch up to 468,000 feet (over eighty miles) in Sprint's Everglades exchange with the average loop length there at almost 138,000 feet. Perrine (south of Miami) has one cluster with an average loop length of over 343,000 feet, while clusters in the downtown Miami Grande exchange are only 465 feet on average from CO equipment.

The longest loop in any cluster in Miami's Grande exchange is 999 feet, while the longest loops in dozens of clusters in other exchanges exceed 100,000 feet (almost nineteen miles). Loop lengths determine two things. First, many circuits such as DSL and ISDN cannot be offered on loops of 18,000 to 35,000 feet or more. Other circuits such as dedicated T-1 lines can be offered to points at greater distances, but only at enormous expense. The added expense results because longer distances may require older transmission technologies that need repeaters every two to six thousand feet, specially conditioned lines, and other expensive qualifying factors. Even 56 kbps modems will not transmit at full speed over long loops. As many as twenty percent of all loops in Florida are 24,000 feet or further from CO equipment.

Line density per square mile is another statistic used to determine where high-speed access technologies are likely to be profitable (and hence, deployed). There are enormous teledensity variances within Florida. For example, the Miami Grande exchange has over 120,000 lines per square mile, while Belle Glade has 0.84 lines per square mile. The sparsest area of all is a cluster within the Eglin AFB (in the Panhandle) exchange served by Sprint that has 0.66 loops per mile. ILECs are reluctant to invest in new infrastructure for sparsely settled areas with low densities per mile. Table 5-2 shows the number of total, business, special, home, and single-line business lines per teledensity category in BellSouth's service area during 1999. Teledensities are frequently a favored way to analyze communications and geography because they are correlated with service levels and costs [Kellerman, 1997].

Source: FCC CCB (2000), HMWKFL2151919999.xls, Excel spreadsheet.

According to Table 5-2, under ten percent of the 1.8 million multiple business lines and trunks were in areas with densities below 850 lines per square mile, and almost one-third lay in areas with over 5,000 per square mile densities. Almost twenty percent of households and single-line businesses were in areas with densities below 850 lines per square mile, with a smaller percentage located in areas with over 5,000 lines per square mile.

Figure 5-1 shows the distribution of loops among teledensity categories. Roughly, about sixty percent of all lines are located in areas with a teledensity of 850 to 5,000 lines per square mile. Approximately twenty-five percent are in areas with higher densities than that and fifteen percent in areas with lower teledensities than 850.

Low subscriber densities and long local loops are not new problems. However, both result from achievement of two former policy objectives of ILECs: deploying POTS lines in high-cost rural areas at prices comparable to other areas and extending POTS to previously unserved places. The REA (Rural Electrification Administration) and RTB (Rural Telephone Bank) were two federal programs that were used to build previous generations of POTS infrastructure in high-cost, low-density rural areas. However, these programs were targeted towards the service territories of small co-operatives and independent ILECs that were certified as rural carriers.

Most of Florida (rural and urban) is served by three large ILECs: Sprint (serving territories of the former Central and United Telephone Companies), GTE (now merged with BellAtlantic-Verizon), and BellSouth, an RBOC (Regional Bell Operating Company). The three large ILECs are classified as non-rural carriers though they serve over ninety percent of Florida's area. Generally, the specific design of rural loop plants in non-rural-certified carrier areas has been left up to the carrier. Often, plans for dedicated circuits, circuit-switched ISDN, and data transmission were left out since the law mandated POTS alone.

In addition to special programs for rural carriers, traditional services and technologies have been subsidized, regulated, and taxed to achieve price equity between urban and rural areas. The methods that have been (and are being) used typically include rate averaging (5.5.1, 5.6.2) and rate of return regulation (5.6.3) by the FPSC. Price cap regulation (5.6.3) is used for non rural-certified carriers. More detail about specific mechanisms is presented in section 5.5, but essentially three transfers (along with overall utility regulation) are used to accomplish the extension of service.

First, business telephone rates are used to subsidize residential service. Prices for business service average two to three times that of residential customers in part so that businesses (which tend to be located in higher teledensity areas) will subsidize residential service (which tends to be found in lower teledensity areas). Second, urban and suburban rates (of businesses and residences) subsidize wireline service to rural areas. Hence, customers in low-cost areas (such as Jacksonville or Miami) pay almost the same rate for telephone (and some other tariffed services) that rural customers in Kenansville or Okeechobee do. These equal costs are charged although it may cost the ILEC ten times or more to provide rural service. The third transfer is through taxation and direct subsidy so that telephone companies that serve high-cost areas are compensated directly for the costs that cannot be met through either of the other two methods.

One bit of evidence supporting the success of these ILEC programs in improving the rural telephony infrastructure can be seen in Figure 5-2. DLC-RAV (Digital Loop Carrier-Remote Access Vehicle) access loop hardware is how the overwhelming majority of telephone subscribers in areas with densities below 850 lines per square mile obtain service. Interestingly, the 850-2,499 teledensity category has a much lower deployment of DLCs and fiber in the loop than more urban teledensities shown on the right.

Cynics argue that Figure 5-2 is evidence of two separate developments rather than one positive one for high-speed access. BellSouth has extended DLC-RAV solutions to many rural areas to save itself money, since many deployments in those locations cannot support high-speed services because of the DLC-RAV equipment used there. BellSouth is now extending newer DLC-RAV technologies (capable of supporting high-speed services) mainly in denser areas (at the right of Figure 5-2) that are financially able to support DSL and other high-speed services.

Before the 1934 Communications Act established a federal mandate to improve rural communications, even the telephone was still far from universal in the United States, especially in poorer, more rural areas. Rural America was considerably behind the rest of the nation in communications infrastructure and Southern rural America even more so. In 1920, for example, fewer than ten percent of Florida farms reported telephones, compared to a national average of 50 percent [Statistical Abstract of the United States, 1924, p. 305].

Now rural areas of Florida have basic traditional telephony services and some enhanced telecommunications services. However, the capacity of rural networks to handle enhanced services varies dramatically. Depending on location, high-speed wireline services such as DSL or ISDN-PRI may be unavailable or prohibitively expensive to all but the largest agribusinesses. Although rural Florida has a reasonably high cable TV pass rate, broadband networks built by Cable TV companies do not extend outside city or town limits to low density areas in most cases. Electric utilities including rural cooperatives have begun to provide retail high-speed access to their "dark fiber" capacities. However, most rural electric co-ops either partner with telephone co-ops to deploy high-speed wireline access or remain focused on electricity alone.

Wireless is often seen as the solution for many rural customers. However, in addition to the technical barriers mentioned in 4.4, access varies. Satellite TV services are widely available, but satellite-based telephone and Internet offerings can require expensive CPE and offer asymmetric data rates. DirecPC, the most popular satellite Internet service requires a telephone modem for the upstream path and offers only 400 kbps downstream. Fixed wireless may depend on unobstructed horizons or thunderstorm-free conditions. Many of the most promising fixed wireless technologies such as MMDS and LMDS are not yet deployed and will be targeted towards office building users when they are. Cellular and PCS carriers are beginning to offer third generation services where no long-distance charges prevail; yet, these may lack full functionality in rural regions. In addition to other regulation, wireless hypercommunications is subject to FCC spectrum allocation rules (covered in 5.5.4).

5.2 Rural, Infrastructure, and Universal Service

Now that a brief overview has been given, it is time to provide a better idea of three essential terms used in hypercommunication policy as it applies to agribusiness. This section covers dimensions of infrastructure (5.2.2) and the meaning of universal service and its high tech twin, universal access in 5.2.3. Before these topics can be covered, a few words must be said about the meaning of the word rural. The definition of what constitutes a "rural" area is much more difficult than it at first appears.

5.2.1 Definitions of Rural

The Department of Census, USDA, U.S. Congress, Department of Commerce, FCC, postal service, zoning, and public safety agencies each have different definitions of how populated a certain geographically configured area must be to be called rural. However, when using the term rural with respect to hypercommunication policy, mere recitation of legal, zoning, Census, or USDA definitions is insufficient. The definition of the term rural is also used to divide Florida ILECs (and other carriers) into rural-certified and non-rural-certified to allocate 1996 TCA subsidies that support the extension of service to rural areas. Precise definitions of rural can be obtained with GIS data that pinpoints population densities for any location.

In Figure 5-3 [U.S. Census Bureau, 1998], the 1990 population density of each Florida County is shown. By 1998, the density for the entire state stood at 274 persons per square mile. The same recent Census estimates showed that Pinellas County (Saint Petersburg) is still the most densely populated County with over 3,185 persons per square mile. Pinellas is followed by Broward County (Fort Lauderdale) with 1,185, Seminole County (Sanford) with 1,146, Miami-Dade County (Miami MSA) with 1,065, Duval (Jacksonville) Hillsborough County (Tampa) with 893, and Orange County (Orlando) with 889 persons per square mile.

Then, a host of less populated cities and counties fall in the 333 to 485 persons per square mile range. Included in that category are Sarasota county (Sarasota), Lee county (Fort Myers), Palm Beach County (West Palm Beach), Escambia (Pensacola), Volusia County (DeLand and Daytona Beach), Manatee County (Bradenton), Brevard County (Melbourne-Cocoa), and Pasco County. In 1998, the least densely populated counties were Liberty (9.3 persons per square mile), Glades (12.7), and Lafayette (13).

These county-level data show that some areas are more sparely populated than others are. However, communication networks rely on the distribution of population around access nodes such as COs and other POPs. In many counties, the density figure may be artificially low because of unusable wetlands (as it is in Dade, Broward, and Palm Beach counties), and the presence of federal or state land. The tendency to define an entire county as rural or urban based on census characteristics of the whole county's land area is therefore flawed. However, this is not the only problem with defining rural.

There are two additional problems with using county densities to define rural for hypercommunications before several alternative definitions of rural are introduced. First, densities in the fastest-growing counties are creeping higher as agriculture loses land to urbanization. Figure 5-4 [US Census Bureau, 1995]shows U.S. Census estimates of growth by county from 1990-2000, as estimated by the U.S. Census bureau in 1998.

The top three 1990-2000 county population increases in percentage terms were projected to occur in Osceola, Hernando, and Flagler Counties. However, recent estimates suggest that Flagler, Liberty, and Sumter County will each achieve population growth of over 40% from 1990 to 2000. In total numbers, five counties are expected to see increases of over 100,000 people: Broward (180,000), Palm Beach (151,000), Miami-Dade (146,000), Orange (136,000), and Hillsborough (102,000).

Second, average densities of rural population do not tell the whole story concerning the market for hypercommunications services. If rural residents are widely scattered, one technology (wireline or wireless) may be better than another depending on how density is defined. Widely scattered could mean 900 people in a single cluster or 900 people in 900 equally spaced-apart dwellings, depending on how density occurs in reality.

Table 5-3 summarizes some of the more frequently used definitions of rural, giving the percentage of Florida's population that falls into each category along with some problems with that definition. The county-based definitions of rural and urban used by the Census Bureau have just been mentioned. Under the census metro versus non-metro county designation, 914 thousand of Florida's population (7.1%) was classified as rural in the 1990 census. An alternate Census definition of rural as persons living outside incorporated and unincorporated CDP places with fewer than 2,500 persons finds 1.97 million rural residents of Florida, making up 15.2% of the total population. The USDA economics agencies use several methods to define rural. For the most rural definitions, a mere 28 thousand inhabitants of three counties made up a small fraction of one percent of the total population.

Sources: ERS, 1993, US Census Bureau, 1995; RTF, 2000, p. 21.

FCC rural-designated carriers served a population of 193,000, 1.5% of the state's population on 14% of the land area. The 1996 TCA defines rural telephone companies as those that do not serve either any incorporated location larger than 10,000 inhabitants or to any territory incorporated or unincorporated that is within an area designated as urban by the Census Bureau [Egan, 1996, p. 264]. Regardless of what other definition of rural is used, the FCC rural-designated carrier definition clearly understates the number of rural Floridians. Using almost any other definition of rural, most residents of rural Florida are served by BellSouth, GTE, or Sprint. None of these three large multi-state ILECs meets the 1996 TCA litmus test for rural-designated carriers.

Unincorporated places had over 6.5 million inhabitants, just over 50% of the state's inhabitants in 1990. While most of the population in these locations could not be classified as rural, almost all rural land falls inside unincorporated areas. Each definition has its own strengths and weaknesses as shown in Table 5-3.

Even with a particular definition, the actual physical task of locating, marking, and defining fast-changing parts of the state are difficult. In the FCC HPCM database, a GIS (Geographic Information System) is employed to map communication network nodes, telco exchanges and wire center, flows, directions, and network topologies [FCC, ACC, 2000]. An enormous body of data for nodes, exchanges, ILECs, and regions is available such as that summarized in Table 5-1 and Figures 5-1 and 5-2.

Several final points help characterize the meaning of rural in the agribusiness hypercommunications context. First, rural is not necessarily agricultural and agricultural is not necessarily rural. The proportion of total economic activity in small communities resulting from agriculture varies dramatically, but in all but a few cases is less than half. Estimates of the amount of agribusiness value added in rural or suburban versus urban facilities are hard to find. Typically, the farm gate price (received by farmers or growers themselves) is a small fraction of the full retail price. Depending on the product in question, as much as ten or fifty times of value is added off the farm at processing plants, warehouses, or packaging facilities in urban areas or small towns. Agribusinesses in these areas have a better chance of getting high-speed hypercommunication services than most farms do.

Second, rural and remote are not necessarily synonyms. Compared to states in the western US, teledensities in Florida are not low. Where they are, in many cases long loop lengths do not lead to isolated farms or sugar plants, but instead to vacation hideaways, hunting and fishing facilities, or private islands [Egan, 1996].

Another argument is often heard that if residents of rural areas want high technology, they should move to areas where it is available rather than vice versa. Characteristics of non-metropolitan residents are often generalizations, but many people gain considerable utility from living in rural areas.

Roper organization surveying, published in 1992, shows that 83 percent of non-metropolitan residents say they are family oriented; 69 percent identify themselves as having a strong commitment to their community (only 5 percent of metropolitan residents do). Rural residents are much more likely to give their communities high marks for personal values, friendliness, cost of living, police protection, recreational facilities, low pollution, and quality of life for themselves and their children. Sixty percent of rural residents think their community is moving in the right direction, only 36 percent of urban residents do.

That Roper survey shows 33 1/3 percent of Americans believe that rural America is the ideal place to live--and 35 percent of all Americans say they would like to be living in a small town in 10 years. [Doll, 1996, pp. 237-238]

This argument may be particularly true of the increasing number farms and ranches where most income comes from off the farm. Over 13,000 of Florida's 34,000 farm operators spent over 200 days off the farm in 1997, while over 19,000 reported that their principal occupation was not farming [USDA NASS, 1997 Census of Agriculture, Table 1].

Furthermore, not all rural residents can move at will, even if they wanted to. This is particularly true of certain agribusinesses because assets are fixed. Hypercommunications infrastructure must come to the rural customers if they are to have access to high-speed networks.

5.2.2 Infrastructure Dimensions

Just as rural has more than one conceptualization, so too does infrastructure. Rural Florida's communications infrastructure is often erroneously equated with the PSTN (Public Switched Telephone Network) operated by the monopoly ILEC (Incumbent Local Exchange Carrier). That same wireline infrastructure may be unbundled and resold or used to interconnected to networks controlled by IXCs, ALECs, ISPs, or wireless carriers. Infrastructure is the physical plant through which a service provider offers access to the hypercommunication network and value added services to subscribers. Next, the dimensions of infrastructure are considered on several levels.

Some authors (Hubert, 1996) use the term infostructure to dramatize the difference between "information communications infrastructure from more traditional national systems such as: transportation, power generation and resources, natural resources, education, water, sewer, health care, housing, and the postal service" [Hubert, 1996, p. 21]. Hubert defines infostructure (what this study calls hypercommunications infrastructure) as:

The combination of all the elements of a country's information communications infrastructure, which would include the following capabilities related to the creation, capture, storage, processing, transmission, and reception of all forms of information: transmission facilities, information services (cable, fiber, wireless) gateways/utilities, switching, terminal equipment/receivers, network management software/systems, data storage devices, broadcasting equipment, computers and data processing equipment, software operating systems and utilities, financial settlement systems, data capture devices, security/validation systems, applications software, and data/content libraries. [Hubert, 1996, pp. 91-92]

While part of a modern business' infrastructure consists of local level devices it owns (CPE), an agribusiness connects to the outside world using a carrier's access infrastructure. Indeed, the CPE the agribusiness buys depends on the access infrastructure. There is no point in purchasing fiber optic transceivers if analog POTS lines are the only wireline connection in an area.

Chapter 4 discussed (in section 4.3) the differences among copper, coaxial, and fiber optic wireline conduit as well as the frequency spectra over which wireless communications paths are beamed. Chapter 4 also explained the differences between the QOS reference model (access level, local level, and transport level) with the OSI layers of data communication. It becomes important to capsulize that material so that the most important incentives and barriers to infrastructure development can be discussed free from technical detail.

Table 5-4 shows some of the main aspects of outside infrastructures that could serve agribusiness. Rows of the table feature the four general hypercommunication sub-markets from Chapter 4 along with essential high-speed technologies used to provide wireline and wireless access to the appropriate communication network. Since regulatory markets are defined somewhat differently (and these affect infrastructure directly), their influence is shown in the first column.

A full discussion of the technical details of most of the entries in Table 5-4 is given through reference to the appropriate section of the text. Each transmission technology (wireline and wireless) and each transport and service type has its own infrastructure dimension. The extension of a hypercommunications infrastructure to rural areas is often far more complicated than the narrowly definition of Table 5-4 allows.

Alternative network structures such as CATV, ISP-ALEC, or local electric utility grids represent attractive possibilities for rural areas, but have been less well-researched by firm R&D efforts when compared to ILEC and IXC R&D histories. Research by AT&T, Lucent, Bellcore (Telcordia), and the REA has examined almost every aspect of hypercommunications. So far, the sub-industry with the best long-run R&D record of accomplishment (telephony) appears poised to gain the largest market share of a converged hypercommunications industry except that regulatory burdens on telephony are the highest as well.

Importantly, regulatory constraints on services have shaped telephone industry R&D portfolios. During the years that AT&T was prohibited from the computer business, for example, R&D in that area worldwide suffered, so great was AT&T's position. AT&T (MediaOne), BellSouth, GTE (Bell Atlantic), MCI Worldcom, and Sprint (soon to be MCI-Sprint) are five of the top Florida hypercommunications providers with roots in the telephone industry. For many years, BellSouth as an RBOC was prohibited from hardware sales in the computer communications and telephony-related CPE markets while GTE, MCI, and Sprint were not. While asymmetric regulation prevent ILECs from offering long-distance and other services, the past monopoly status of AT&T and the RBOCs enabled them to spend generously on forward-looking R&D that still yield distinct market advantages.

Table 5-5 concerns an especially important dimension of infrastructure, how economists analyze costs of infrastructure development. Currently, most economists and regulatory authorities believe that forward-looking costs are the most applicable way of analyzing infrastructure costs [FCC, 97-160, 1997]. Egan defines forward looking as:

A descriptive term normally associated with incremental costing methodologies and techniques utilizing estimates of quantities, costs, and conditions that will prevail during a future period. It also implies costs yet to be incurred as opposed to costs previously committed. [Egan, 1996, p. 347]

Sources: Egan (1996, pp. 318-319); FCC, 99-5, February 1999; FCC, CCB, 2000; Zaatari, 1999; USDA-NTIA, 2000.

Table 5-5 shows Egan's summary of some of these AIC (Average Incremental Costs) for deploying various infrastructures on a per subscriber basis. Alternative sources (when available) have also been posted in Table 5-5 to obtain a range of estimates, both for 1996 and for 1998-2000. All AICs shown are broad estimates of nationwide averages. In some cases, especially wireline and cable, costs in parts of Florida would be lower than the national average.

It is important to note that most of the forward-looking costs for the services listed in Table 5-5 have dropped dramatically from Egan's 1996 survey to the present according to several sources. The cost decreases in some cases have come from lower conduit prices or improved efficiency as the installed base grows. In other cases, the decrease in costs is a result of technological advances such as new multiplexing or spread spectrum technologies that allow more traffic to be carried over a single wire or radio channel.

5.2.3 Universal Service and Universal Access

Having served as the main method of business communications for decades, telephone interests have created much of the current debate about Universal Service. Importantly, the term universal service was first employed by AT&T's Theodore Vail, who has been called the father of communications marketing. That early definition of universal service meant that service would be universally provided by AT&T and was akin to interconnection. Early telephone customers could not even talk with each other because of differences in equipment and lines. AT&T created de facto and later de jure standards that would eventually apply to the entire PSTN.

Universal service (and its high-tech cousin universal access) are the next essential terms of hypercommunications policy. Indeed, the specific policy mechanisms of universal service are covered in 5.5.1 and the topic occurs throughout Chapter 5. Therefore, it needs to be defined. The 1996 TCA defines universal service as "an evolving level of telecommunications services that the FCC shall establish periodically, taking into account advances in telecommunications and information technologies and services" [1996 TCA, Title 1, Sec. A]. More specifically, the statute conceived of seven policy bases of universal service as given in Table 5-6.

Sources: Bases from 1996 TCA. FCC, 2000; FPSC, 1999.

Local-state joint boards and the FCC were directed to "base policies for the preservation and advancement of universal service on" seven factors or bases [1996 TCA]. For each basis, the purpose and status are given. Some of the purposes behind universal service reoccur as rationales for overall government involvement in hypercommunications in 5.4 as well.

The first policy base of universal service is availability of quality services at just, reasonable, and affordable rates. The purpose is based on equity rather than directly on economics. However, if network externalities are positive, then spending on mechanisms to support the policy may bring more benefits to society than costs. There is some evidence that the policy has not been as successful as it might have been at promoting universal availability of telephone service. While 92.6 percent of Florida households have telephones, fewer than 80 percent of certain income and educational categories do [FCC, 2000; FPSC, 1999].

The second policy base is to ensure access to advanced telecommunications and information services to all regions of the nation. This is also known as universal access to underscore the fact that the goal is not restricted to analog POTS. Advanced telecommunications has been defined as 200 kbps or greater on a federal level [FCC CCB, FCC 00-290, 2000]. However, the FPSC states their goal is universal 56 kbps, but it will be as many as three years (until ILECs upgrade plants) to achieve even this base universally throughout Florida. Until 20002, 56 kbps is defined as broadband accessible in Florida [FPSC, 1999, p. 3]. Extending the definition to more advanced services is still under study by the FPSC. In September 1999, fewer than 40% of Florida COs could support DSL. However, enhanced telecommunications services are increasingly available in urban and suburban parts of Florida.

A third basis of universal service is that access and costs in rural and high-cost areas be reasonably comparable to those in urban areas. Rural rates are still balanced by FPSC regulations to ensure rough comparable. This subject is considered later as a mechanism of government involvement in 5.5 and in 5.6.2 in the context of universal access.

The fourth policy basis, equitable and nondiscriminatory contribution has probably not been met. ILECs and IXCs still bear the burden of paying support mechanisms. The incidence of the support fee on customer bills for extending universal service is such, the telcos argue, that all of it must be passed on to customers. While bills of large businesses have fallen drastically, bills for small businesses and residential accounts (especially for extremely low use customers) have risen up to thirty percent. The increase is partly attributable to the universal service fund and other charges and taxes resulting from the 1996 TCA.

Specific and predictable support mechanisms for universal service are the fifth policy base for universal service. Collected funds are subject to a variety of administrative formulas that regulators and telcos are still sorting out. Indeed, two new mechanisms have just been enacted by the FCC in the year 2000 [FCC, May 31, 2000]. However, the sixth policy base for universal service, access to advanced telecommunications services for schools, health care, and libraries; has been quite successful. An overwhelming majority of Florida schools (even in rural areas) has obtained advanced services through the Schools and Libraries program.

Finally, the 1996 TCA allows for dynamic adjustment mechanisms as needed. While the FCC has concluded that Internet traffic is jurisdictionally mixed (part local and part interstate), it has taken a hands off policy to regulation. Given the increase in low cost and even free IP telephony services, there have been calls by telcos for regulation of voice over Internet. While the FCC has ruled that Internet traffic is jurisdictionally mixed (and thus potentially subject to interstate-intrastate separations rules), the Commission has taken a hands-off approach to regulating the Internet. The FPSC and FCC have been considering numerous cases regarding reciprocal compensation, or how one carrier is paid by another for terminating calls, especially to ISPs.

According to some views (such as those of the rural communications lobby OPASTCO), deployment of infrastructure depends on the federal commitment to universal service. A recent OPASTCO study suggests that even if carriers plan to upgrade rural infrastructure, their ability to realize those plans "is largely dependent upon the universal service mechanisms enacted and the extent to which continued infrastructural upgrade is economically feasible." [Hobbs and Hobbs, 1998, p. 28].

Importantly, the term universal access is used instead of universal service to signify that universal POTS telephony or even universal enhanced telecommunications services leave out some important ingredients of the infrastructure of the future. As the NTIA-RUS report in June 2000 put it:

We urge the Federal Communications Commission to consider a definition of universal service and new funding mechanisms to ensure residents in rural areas have access to telecommunications and information services comparable to those available to residents of urban areas. [NTIA-RUS, June 2000, p. iv]

The idea here is to replace the current static definition of universal service as "both a voice grade bandwidth of 300 to 3400 Hz and a data rate of at least 28.8 kbps" [NTIA-RUS, June 2000, p. 42]. The new definition, universal access would be an evolving standard that would fulfill the requirements of the 1996 TCA to support advanced services that is more in line with the policy bases given in Table 5-6. The regulatory differences between universal service and universal access (the result of convergence) are discussed further in 5.6.

Before considering the economics of how specific universal service (and other regulatory mechanisms) operate, it is important to cover the history of regulatory and infrastructure to create the base scenario for the economic analysis of the chapter.

5.3 Regulatory and Infrastructure History

Economics provides an excellent context for studying the history of infrastructure and government actions in communications. To quote Louis Schmidt in 1916:

The agricultural economist needs to be familiar with the economic life of man in the past in order to realize and appreciate the organic nature of society. He should be historically minded if he would deal most effectively with the problems of the present?. The great problems of rural communities are human rather than merely materialistic. That is to say, they are economic, social, and educational, and cannot be understood except in light of their historical evolution. [Schmidt, 1916, p. 45]

Infrastructure development is important to agriculture because the rate at which it occurs is markedly different between rural and urban areas, as well as varying among regions of Florida. Re-regulation, taxation, and other policies are meant to speed up the development of infrastructure in markets that have been missed by the uneven pace of competition. Federal, state, and local governments have a variety of policy mechanisms designed to ensure universal service or rough parity among geographic locations in Florida. However, the regulations themselves can often lead to barriers to entry, limitations on service, and decreased competition. Public policy will influence the choices and costs many agribusinesses will face in the hypercommunications marketplace.

The history of how traditional telecommunications has converged with other sub-industries to create hypercommunications and the rural-urban infrastructure dichotomy depend on the four sub-markets presented in the last half of Chapter 4 but with a different organizational framework. However, the traditional telephony (POTS) and enhanced telecommunications infrastructure is considered first (5.3.1) because it is still Florida's widest in geographical terms, largest in dollar terms, and an important and measurable source of efficiency and coverage. Wireline access to enhanced telecommunications, private data networking, and the Internet are part of the same story. Agribusinesses still rely mainly on wireline local loops to access hypercommunication networks other than the PSTN.

The second regulatory history is of broadcast and wireless technologies (5.3.2). This segment covers a wide regulatory and historical territory, ranging from rural AM "daytimer" radio stations to international satellite networks capable of video, voice, Internet, data networking and broadcast transmission. Mobile services in this sub-market duplicate those of traditional and enhanced wireline services and interconnect to the PSTN through most of Florida. New fixed and mobile services are being introduced that can compete with wireline cable or telephone companies alike.

Before the 1996 TCA, regulation was enacted separately for wireline and wireless transmission technologies and services. Even after the TCA, ILECs are regulated differently from other telcos (ALECs and IXCs) who are regulated differently than cablecos and mobile wireless providers. The sources of these regulatory asymmetries are based on history, technology, and recent regulatory reform. Hence, after covering the wireline and wireless infrastructure and regulatory history, 5.3.3 looks at the 1996 TCA which was designed to update regulations to include the emerging reality of hypercommunications convergence.

5.3.1 Wireline Infrastructure History

No discussion of hypercommunications would be possible without a thumbnail sketch of the telephone system, as it has evolved over the last 122 years in Florida. Florida's hypercommunications infrastructure is a product of her Reconstruction days as an impoverished and remote Southern state, her unique right-of-way laws (stemming from the 1920's land boom), and her rapid growth after World War II fueled by Northern immigrants. However, the state's wireline infrastructure also includes enhanced telecommunications, private data networks, cableco networks, and Internet backbones and access methods. The result is an amazing diversity, with Yahoo! naming Miami the tenth most wired city in America in 1998 [Yahoo!, 1998].

Within the wireline category is a short history of new entrants including cablecos and so-called "dark fiber" providers (such as electric utilities) with existing fiber optic capacity, sometimes in rural areas. These providers have their own infrastructure economics that often allows them to supply FTTH (Fiber to the Home) or FTTN (Fiber to the Neighborhood) at less cost than telcos can. However, their existing plant (described in detail in 4.3) does not reach many rural areas and depends on each cable or electric company as to how (or whether) it has been developed. While access to the Internet and private data networks can be obtained through cableco connections, the main wireline source is still via telephony local loop.

The story of Florida's modern communications infrastructure begins after Samuel Morse's famous "What hath God wrought?" telegraph message sent to Baltimore from Washington, D.C., in 1844. With the first transatlantic telegram transmission in 1858, Cyrus W. Field was able to communicate to London from America. That inaugural transatlantic message would predate the first transcontinental one by three years, so important was the Northeast during the introductory years of the telegraph. Only in 1861, as the Union entered the Civil War was east west communication within the United States considered vital.

One reason the South suffered during and after the Civil War was her inferior telegraph system. After the War, when Western Union and later, the new Bell Telephone Co. began to expand the telecommunications infrastructure, some argue that the Reconstruction South was neglected. However, telephones were in use as early as 1878 in Jacksonville. By 1880, Florida's first telephone exchange opened there, a mere 28 months after the first CO started in New Haven, CT and four short years after the inaugural Bell-Watson conversation.

Instead of neglect, others say that the slow spread of the telephone southward was based on a marketing reality: there were too many rural, poor, and illiterate people in the Reconstruction South (especially Florida) for telephone service to be profitable [Oslin, 1992]. The few early telephone customers in Florida in 1900 had to contend with severe static and interference. At the turn of the century, the sunshine state had just over 520,000 inhabitants and but two large towns, Jacksonville with 28,000 people and Tampa with 16,000. Miami was a small town with 1,700 inhabitants and only 3 telephones. The essential role of transportation in the development of Florida's communications infrastructure as well as the agricultural settlement of the state as the 20^th Century began cannot be overemphasized.

However, post-Civil War efforts at developing Florida's transportation infrastructure were stymied by many factors resulting directly and indirectly out of Reconstruction, as well as by Florida's rural, underdeveloped, backwater frontier image. One example of the difficulty the state had in developing an infrastructure can be seen by examining the Internal Improvement Fund. This fund enabled the state to finance (through bond measures) the construction of the railroad system during the 1850's. Postwar Florida's state government could not pay the bearers of these bonds in cash, and the bondholders refused land. In 1881, after federal courts were implored to force the state to sell land to satisfy bondholders, the state sold 4,000,000 acres of "swamped and overflowed" land for twenty-five cents an acre to Hamilton Disston [Patrick, 1945, p. 480]. It was then free to offer land grants to individuals desiring to make transportation improvements [Patrick, 1945, p. 87]. Therefore, it was not until the 1880's and 1890's that serious improvements to Florida's meager transportation infrastructure of the 1860's were begun.

However, even with some new railroads and roads, by 1920, Florida's rural communications infrastructure was still behind every other part of the country except for Louisiana, Georgia, and South Carolina as Figure 5-5 shows [US Census Bureau, 1924, p. 305].

During the 1920's, parts of urban Florida had become a vacation mecca for wealthy northern tourists who demanded better communications. Henry Flagler launched a modern railroad to deliver tourists by train, Collins built the Dixie Highway to deliver tourists by auto, and Florida's agribusinesses required better roads and railroads to deliver their goods to market. Rather than bring a universally better communications infrastructure, the new transport structure brought about a tradition of what USWest Chairman Solomon Trujillo calls "electronic redlining". The most modern telephone plants were installed only in particular areas, with minority, rural, and poor areas left with below average services and aging equipment.

It only took a generation for the telephone to become a luxury status symbol in urban America. It would take longer to reach rural America to become a status symbol there while becoming commonplace on the urban scene. By the later years of the Great Depression, in 1938, fifty years after the telephone arrived in Florida, telephone service was either non-affordable or not available to many Floridians. While firmly established in the city, it would take programs such as the REA (Rural Electrification Administration) telephone bank, a depression-era federal agency, to establish co-operatives to serve certain rural areas.

Federal regulatory history began with the 1910 Mann-Elkins Act, the first legislation to regulate telephone rates. The Act was enforced by the ICC (Interstate Commerce Commission). By 1921, Congress passed the Willis-Graham Act allowing the ICC to administer acquisitions and mergers in the telephone business. The 1934 Communications Act turned authority over to a new agency, the FCC.

By 1948, wartime wireless breakthroughs caused the FCC to force AT&T to carry television signals at regulated rates and AT&T was powerless to stop TV network microwave networks that were separate from the PSTN. Wireless technologies resulted in the eventual competition in long-distance. Additional background to the history of wireless regulation is found in 5.3.2.

In 1949, the federal government brought the action United States vs. Western Electric based on the Truman Administration's public philosophy. The government contended that AT&T should divest itself of Western Electric (its wholly owned subsidiary) that made all business and residential telephones and virtually all telephone infrastructure equipment. The settlement of the case did not come until 1956 and was seen at the time as a victory for AT&T because AT&T's structure was left almost unchanged. By this time, Western Electric had withdrawn from making TV, radio, and other non-telephone equipment. In the 1956 settlement, AT&T agreed to stay out of equipment markets outside of the telephone industry. However, valuable patents developed by the Bell System had to be licensed at reasonable rates to other companies according to the settlement.

In spite of anti-trust pressure from the government regarding AT&T's Western Electric subsidiary monopoly on telephony CPE, AT&T staved off all attempts to interconnect non-AT&T equipment into the telephone network. One of the earliest interconnection technologies that AT&T had to defend itself from was a plastic device called the Hush-a-Phone that snapped onto telephone sets (earpiece and mouthpiece) so that conversations would be private. Hush-a-Phones were not electronic and could not cause negative externalities to ripple through AT&T's network. Nonetheless, AT&T defended (unsuccessfully) against the use of Hush-a-Phones in FCC hearings and in court from 1948-1956.

Competition in long distance began as MCI gradually chipped away at AT&T's monopoly of the telephone transport network once the FCC mandated new microwave interconnection rules. In 1969, MCI was allowed to offer inter-city microwave transmission for subscribers, beginning a new industry of SCCs (Specialized Common Carriers). Initially SCCs offered point-to-point circuits so that calls could be placed only among subscribers to the SCC's network and not to all telephones in the world as in the PSTN.

In 1974, MCI filed suit against AT&T, charging anti-trust violations such as monopolization and failure to interconnect. A final verdict in the case was not received until 1985, when AT&T was ordered to pay MCI $120 million dollars. In addition, in 1974, the Justice Department brought a suit that sought to force AT&T to spin off local service from the Bell System.

MCI also established its own long-distance service, Execunet, in 1975. To place a long-distance call, a local number (essentially a POP) was dialed and the long-distance caller entered a series of access codes to Execunet. Then, a call could be completed to PSTN telephones in any area code Execunet served, establishing in essence a long distance call through MCI's microwave relay network instead of AT&T's transport network. AT&T brought legal action arguing that such a service was outside the point-to-point nature of SCCs and was actually a long-distance call. In 1978, after protracted appeal, AT&T and the FCC (which took AT&T's side) lost the case to MCI. Competition in long distance was established. MCI also was allowed to provide FX (Foreign Exchange) service.

The Justice Department's 1974 suit reached settlement in 1982 with the breakup of AT&T. Since over 871,000 pages of testimony, exhibits, and pleadings were made in the case, only a bare summary of the main issues can be given. The essence of the government's case was in four areas. The first area concerned the right of SCCs to interconnect to AT&T's local or transport network. A second issue was the fact that Western Electric still provided equipment for the entire Bell System so that competition in switching and other telephone network equipment was stifled. A third issue was AT&T's refusal to allow CPE (owned by customers but made by non-AT&T manufacturers) to interconnect with the PSTN. AT&T insisted that customers with such non-approved CPE use only AT&T manufactured devices to permit interconnection. Fourth, and finally, was the issue of how the SCCs and bypass carriers should compensate AT&T for completing calls that were originated and transported over non-AT&T networks.

By 1982, parts of the Justice Department's objectives in the case had been obtained in the Execunet case or were being re-tried after a 1.8 billion dollar settlement in favor of MCI in MCI's own anti-trust case. Additionally, FCC and state regulatory authorities were liberalizing CPE interconnection rules. Greene (1997) argues that neither the government nor AT&T had anything to gain by waiting years for the trial and appeals process to grind on. Therefore, the case was postponed in early 1981 to allow settlement negotiations to be made.

According to Stone (1997, p. 81) the breakup settlement consisted of AT&T agreeing to:

divest its twenty-two local operating companies and withdraw from the local-loop business. In exchange, the Justice Department agreed to let AT&T keep its manufacturing arm and retain Bell Labs, the research facility, although AT&T agreed to help the seven newly formed RBOCs set up Bellcore, their own research facility. Most importantly, from AT&T's perspective, many of the restrictions on AT&T contained in the 1956 consent decree were lifted. [Stone, 1997, p. 81]

The settlement would be known as the MFJ (Modified Final Judgement) and was issued at the beginning of 1984. AT&T was now allowed into the computer business and permitted to remain in the nascent wireless (cellular) telephone business. AT&T had to spin off local telephone and other local services into the RBOCs who themselves were generally prohibited from offering long distance or non-regulated services. In addition to long-distance, AT&T was allowed to operate primarily in competitive markets such as data communications, network management, etc. The RBOCs had to allow equal access to long distance and directory assistance services provided by other companies, but were permitted to hold on to the lucrative Yellow Pages.

BellSouth Corporation, the RBOC that covered much of Florida (see Figure 5-6 later in this section) was formed through the merger of two AT&T spin offs, the Southern Bell Telephone Co. and South Central Bell Telephone Co. The MFJ allowed the seven RBOCs to each remain larger than GTE, the nation's second largest telephone company before and after the breakup. RBOCs could sell or lease CPE, but not manufacture it. The future of BellSouth, GTE, Sprint, and other ILECs (Incumbent Local Exchange Carriers) was assured within Florida as well. The MFJ changed the infrastructure and regulatory structure enough that ILECs could eventually be joined by new wireline carriers, such as ALECs (Alternative Local Exchange Carriers). IXCs (Interexchange Carriers) or long distance carriers were established by the MFJ.

However, POTS and enhanced telecommunications services provided by telcos are not the only story of wireline access and regulation to be recounted. Cable and electric companies bypass the local access level of the ILECs (using their own local loop) to provide POTS, enhanced telephony, Internet, and private data networking access. It would take new legislation, the 1996 TCA (covered in 5.3.3) to allow wireline regulation and infrastructure to admit new services and new carriers into the burgeoning hypercommunications market.

The technical side of cable TV infrastructure is covered in 4.3.3. Models suggest that to compete with the cablecos, there are two steps telcos must follow to provide cable broadband services via the telco plant. First, there is infrastructure construction as covered in Table 5-5. Second, there are service deployment and price structure. Nationwide, a revenue requirement per customer per month of $42 is needed just for construction, and $51 per month for service deployment. In rural areas, the figures are $64 and $72 per month per subscriber. However, to provide rural business infrastructure alone the construction cost per subscriber per month is estimated at $41 and $73 per customer per month for service deployment [Weinhaus et al., 1995].

Discussion of specific Internet and private data networking access technologies was given in 4.8 and 4.9. Importantly, wireline access to both depends on telco, cableco, or dark fiber infrastructures. The recent infrastructure and regulatory histories regarding these newer developments will have to wait until after the wireless story is told and the 1996 TCA is used to level the playing field and stimulate competition.

5.3.2 History of Wireless Infrastructure and Regulation

The regulatory history of the wireless sub-market is based on the original broadcast services of radio and television, which are non-interactive mass communications media. It was once necessary to have a radio receiver or TV tuner to receive live or taped broadcasts. Now, audio and video communications can be transmitted over the Internet, through private data networks, through the CATV infrastructure, or be heard and seen via CD, DVD, or videodisc. However, the FCC still regulates the airwaves with some surprising new twists such as spectrum auctions.

During World War I, AT&T was put to work on projects for the Army Signal Corps. It was out of Western Electric's development of air-to-ground, ground-to-air, and air-to-air communications that interpersonal wireless communications developed in the United States. In 1920, a Catalina Island radiotelephone link was established by AT&T due to postwar copper shortages. Another wireless operator, the American Marconi Company, was nationalized and privatized during World War I, but Marconi's business plan was focused on navigation, telegraph transmission, ship-to-shore, and radio broadcasting rather than person-to-person wireless.

The regulatory history of wireless began in 1912 when federal courts ruled that the Secretary of Commerce and Labor could not refuse radio licenses to any qualified applicant. The ruling came in spite of a developing "tragedy of the common airwaves" where shared public frequencies were creating massive interference. The problem became so serious that by 1929, there would be 50 fewer radio stations on the air than in September 1927 when there had been 757) [Doll, 1996]. Not until 1938 would the 1927 total be reached again.

Herbert Hoover's intervention as Secretary of Commerce changed the radio market into a larger-scaled, closely regulated, and more profitable business than it had been in pre-regulatory days. In 1927, Hoover was able to form the FRC (Federal Radio Commission) to phase in the allocation of frequency licenses. The telephone ICC and radio FRC merged into the FCC in 1934. The FCC became in charge of wireless spectra using its newly granted licensing powers.

It was not until after World War II that it became clear that wireless (specifically microwave) transmission could compete with AT&T's transport network. Spectrum allocations made by the FCC permitted mining companies, railroads, the oil and gas industry, and other firms to create their own microwave networks. AT&T opposed such moves, arguing that it had sole right to transport voice communication over microwave links. However in 1959, the FCC agreed to permit private microwave networks as long as they were not interconnected with the PSTN, opening the door to limited competition with AT&T in two-way voice communications.

By the late 1960's, digital microwave transmission became technically possible, bringing new pressure on AT&T's monopoly over the telephone transport system through which long-distance calls traveled. Changes in telecommunications transport cost economics and large customer bypassing were two policy issues that directly flowed from the rapid expansion in digital microwave technologies.

First, the availability of microwave changed the economics of telephone transport from its traditional natural monopoly footing. Support for this idea came as early as 1975 when

economist Leonard Waverman argued the advent of microwave as a long-distance transmission medium dramatically altered the (telephone) industry's economics, so that the natural-monopoly justification no longer existed as it had when open wire pairs were the principal medium for intercity wire traffic, which was the case until the 1940's. Accordingly, he urged in an influential paper [Waverman, 1975], increased competition was in the national interest. [Stone, 1997, p. 139]

Second, the availability of microwave transport in support of bypass carriers threatened to fragment the telephone market and wreak serious consequences on small local customers and local independent telcos.

At the time of the (AT&T) breakup the fear . . . that microwave would allow large business customers to bypass the local loop was widespread. . . . Lower local rates for large customers and higher access charges from long-distance companies to local companies were conceived as the most effective response to the threat, which would inexorably take its toll on the local operating companies. [Stone, 1997, p. 141]

There were two kinds of bypass involved: facility and service. With facility bypass, users built a microwave link or leased a line from a bypass vendor to connect the user directly to a long-distance location. Service bypass involves users purchasing private lines from the local operating company to reach a bypass carrier's POP [Stone, 1997, p. 142].

As was mentioned in 5.3.1, AT&T's breakup came because bypass carriers (such as MCI and Sprint) argued successfully that the economics of wireless transport had changed enough that AT&T's natural monopoly was worth protection. The first case of a company other than AT&T being allowed to interconnect with the PSTN came with the 1968 Carterfone case. Courts held that wireless Carterfones (predecessors of SMR) could be interconnected directly to the PSTN. Carterfone operated radiophone equipment mainly for use as a two-way dispatch system in the oil and gas industry, but operators could connect remote callers to the PSTN as well. AT&T bitterly fought Carterfone, but was forced to allow direct PSTN connection of other kinds of CPE (including computers and customer-owned microwave systems) due the precedent set by the monopoly's loss of the Carterfone case. AT&T retained the exclusive right to supply control devices to allow customers' computers access to the telephone network for data calls and sessions.

Another wireless area where AT&T faced challenges was in satellite communications, which it lost control of with the formation of Comsat (the Communications Satellite Corporation) in the 1962 Communications Satellite Act [Stone, 1997, pp. 53-57]. The FCC had jurisdiction over new private satellite firms who were allowed to provide long-lease service but not connect to PSTN due to the international Intelsat treaty. Much later, fiber optics (as a substitute technology to satellite transport) changed the equation again. By 1997, FCC endorsed elimination of Intelsat interconnection restrictions [Stone, 1997, p. 125].

Recent developments in wireless infrastructure are noteworthy for two reasons. First, the FCC has used economic theory to auction previously unallocated radio and microwave frequencies. This action has increased available spectra, making a variety of new services such as broadband PCS, digital cellular, and a variety of upperband services such as MMDS, LMDS, and DEMS possible. Spectral auctioning was proposed first in the early days of radio but was rejected. Professor Coase's 1959 article about the FCC was the first of many arguments used to show that auctions would be more efficient solutions than licensing alone [Coase, 1959]. The advice began to be taken with digital PCS auctions in the 1990's.

However, many of the newly auctioned frequencies are silent as new technologies and the wireless infrastructure needed to transmit at specialized wavelengths are deployed. Carriers are investing billions of dollars in tower construction to support cellular, PCS, GSM, and fixed wireless technologies that were covered in 4.4.2. While wireless access and transport offer great promise for rural communications needs, an infrastructure is still required, though in many cases wireless infrastructures may be built more cheaply than in the wireline case. Satellite communications, because it requires expensive launches of possibly large constellations of satellites, is currently more expensive than terrestrial wireless or wireline infrastructures as evidenced by the recent bankruptcy of the massive Iridium project.

Finally, the FCC also regulates broadcast stations including local rural radio stations. Local radio stations contribute substantially to their communities in financial and non-financial ways:

Much of what they do goes unnoticed outside the community and may not always be appreciated within the community--until the station goes silent or is moved out.

In many small towns, it is only the local radio station that carries the fact that a friend has passed on, or that a neighbor or relative has had a new baby, or the neighbor's barn burned, or that the county or city board is planning an increase in property taxes. These stations are the vehicles that carry urgent pleas for a blood donation to help save a local youngster's life, that the high school band needs a donation to march in the President's inaugural parade, that the school system is deficient in a vital but non-budgeted area. [Doll, 1996, p. 232]

However, a criticism of the FCC regarding the non-metro radio market is that the FCC has caused it to be overbuilt [Doll, 1996, p. 237].

It is unclear how the convergence of web broadcasting and new low power FM, HDTV, and spread spectrum technologies will change broadcasting. Just as ILECs, IXCs, RBOCs, and cableco offerings caused an uneven playing field to develop in wireline regulation, so too have mobile, fixed terrestrial, and satellite systems created regulatory inefficiencies in the wireless area. By the mid-1990's, it was plain that regulations could not keep up with the new services and technologies in the multi-faceted communications markets. As the future of convergence became more likely, there was enormous political pressure to replace the 1934 Communications Act with a regulatory framework capable of dealing with the communications infrastructure of the information economy.

5.3.3 Regulatory Pseudo-Convergence: The 1996 TCA

The need for unified legislation to replace the hodgepodge of regulations that applied to different services (from POTS to the Internet) and to different providers (from wireline RBOCs to wireless cablecos) became apparent by the early to mid 1990's. Regulation or the lack thereof is especially important in any analysis of sub-markets in the pre-converged hypercommunications sector. In studying the history of hypercommunications, it becomes clear that the regulatory environment depends closely on how markets define technologies and services. Yet, market definitions themselves depend on regulation. This dual relationship makes it hard for agribusinesses (used to a century's worth of thinking about communications as being provided by their monopoly ILEC) to get used to two important cost-cutting ideas about convergence that will simultaneously increase communication.

The first important idea about convergence (that the office telephone and computer are converging into one device) may be hard for some agribusinesses to imagine, since device-device convergence is in the early adoption stage. Second, perhaps easier to understand, and already available (thanks to competitive convergence and regulatory convergence) is the ability for an agribusiness to connect to two or more communication networks using one connection and one carrier. Adoption of the second idea requires accepting the idea that the hypercommunications marketplace will not be monopoly-controlled, even if the telecommunications marketplace was. However because of technological change, this idea was a theoretical possibility before 1996, but it only became possible with the passage of state and federal de-regulation legislation.

Recall from section 4.1 that there are more than two dimensions to convergence. However, device-device convergence, market convergence, and regulatory convergence are infrastructure-related. While device-device convergence stems from the local infrastructure owned by the agribusiness including CPE (Customer Premise Equipment), DCE (Data Communications Equipment), and DTE (Data Terminal Equipment), the other two convergences depend on outside infrastructure. Regulation is designed to bring infrastructure to rural areas so that high-speed hypercommunications will be both accessible and inexpensive. A fringe benefit of deregulation is new competition from firms previously in separate sub-markets now that multiple infrastructures (wireline telecommunications, cableco, fixed wireless, mobile wireless, and dark fiber) are available. However, since nothing prevented AT&T from buying the cable companies MediaOne and TCI, while continuing to offering long-distance and both fixed and mobile wireless services, market convergence might mean fewer providers.

The 1996 TCA (Telecommunications Act, Public Law 104-104) was signed by President Bill Clinton on February 8, 1996 [Thomas, 1998]. The exact date is important because time has sped up in the Information Age. Compressed time makes it more important strategically for regulators and industry alike to keep up with competitors, new technologies, and new services. Florida passed its own state deregulation bill in 1995, before the federal 1996 TCA went into effect.

The U.S. Senate had passed its version of the TCA (S. 652) by a vote of 81-18 on June 15, 1995, while the House passed its telecommunications measure (H.R. 1555) on October 15, 1995 [Thomas, 1998]. Though enaction had been expected during 1995, powerful interests successfully lobbied the House-Senate Conference Committee and the White House to delay the bills, knowing the Clinton administration and Congress both were expecting a tough election battle in 1996. W.T. Stanbury's 1996 chronology of US press accounts quotes the October 16, 1995 Business Week as saying:

Delay gives more time for regrouping by the White House--which has threatened a veto over cable deregulation, media concentration, and the terms under which the Bells can get into long distance. Other bill opponents include several major consumer organizations. The longer passage takes, the more leverage critics will have to extract concessions. [Stanbury, 1996, p. 23]

Some critics alleged that the legislative playing field had been tainted by campaign contributions and while the field was not completely level, contributions from each special interest category may have tended to offset one another. Perhaps no solution could balance the needs of a diverse set of local telephone companies (ILECs, ALECs, and RBOCs), long-distance telephone companies (IXCs), cable TV providers, software makers like Microsoft, and data and enhanced telecommunication CPE and DCE manufacturers. A host of (at the time) smaller players rounded out the list of affected parties. These included satellite carriers, cellular telephone companies, newspapers, radio and TV stations, educational institutions, ISPs, microwave carriers, and other firms who had begun to gobble up new wireless frequencies at FCC auctions.

The historical local telephone monopolies (ILECs in general and RBOCs in particular) had the most to gain and lose. The MFJ had prohibited them from the long-distance business and imposed a cumbersome patchwork of LATAs that affected dedicated point-to-point and circuit-switched connections sold for private data network use as well. Furthermore, ILEC responsibilities as carrier of last resort and the decisions they made concerning rural and urban infrastructure were at stake. Also endangered was the enormous capital investment on their own access and transport networks that ILECs would now be expected to allow new competitors called CLECs (Competitive Local Exchange Carriers) to interconnect with. Under the TCA, these former monopolies would still have to provide universal service to any business or residential customer.

Florida ILECs (shown in Figure 5-6) included the giant RBOC BellSouth along the East Coast and certain inland locations through the state, GTE in the Tampa Bay area, and Sprint. Sprint's Florida territories are products of the Centel and United telco acquisitions mainly in Southeast Florida and the Tallahassee area.

Certain rents, regulatory obligations and credits, special taxes, and assessments flowed to the ILEC in each part of Florida before the TCA because of existing state and federal legislation and FPSC actions. After the TCA, it was left to a combination of the market, and federal and state regulators as to what each ILEC would be required to do to furnish "universal service" to customers. In 1996, BellSouth, GTE, and Sprint had 59, 21, and 18 percent respectively of local access lines in Florida [Florida Statistical Abstract, 1997, Table 14.60, p. 435].

The lightest area in the map shows (denoted as other in the key to Figure 5-6) shows the smaller independent ILECs of Florida. Independent ILECs are small companies that had competed against the former Bell System in an earlier age of communications deregulation (before the 1934). When the present regulatory era began with the 1934 Communications Act (the organic statute of the 1996 TCA), independents were permitted to continue serving their markets. Many of those such as Alltel's North Central Florida territory were cobbled together when an existing telecommunications company purchased an existing ILEC. Others were rural telephone co-operatives.

According to RBOC U.S. West (now Qwest) Communications president Solomon Trujillo, ILECs have been trapped into an unfair and discriminatory bind by the 1996 TCA. Specifically the TCA did three things to limit RBOC operations. First, the TCA has deregulated the lucrative local line business while allowing competitors (ALECs) full right of interconnection to ILEC access and transport networks. Secondly, the TCA also mandates that ILECs (especially RBOCs) not do business in certain hypercommunications segments such as interstate long-distance.

Third, ILECs are carriers of last resort (covered further in 5.5.3), meaning they must provide service to high-cost hard-to-serve areas although infrastructure competitors do not have the same obligations. For example, cable TV rates were deregulated, but cablecos are subject to state regulation as ALECs if they provide local telephone service. Data communication, traditionally divided into basic and enhanced categories, saw the continuation of certain basic tariffs, while online services such as ISPs and OSPs were generally free from regulation. Other aspects of the TCA were discussed (along with universal service) in 5.2.3 and will be revisited elsewhere in Chapter 5.

5.4 Rationales for Government Involvement

The next three sections consider rationales (5.4), mechanisms (5.5), and the evolving economics of government involvement (5.6) in converged hypercommunications. Although seven policy bases for universal service (discussed when defining the term in 5.2.3) are important rationales for government involvement, there are broader rationales. As Alan Stone writes, the rationale for government involvement in hypercommunications has moved into an era of privatization and de-regulation. However, government will still need to play a role:

Rejecting public ownership, however, does not mean that government should play no role in shaping market structure or behavior. . . . A historical view of (telecommunications) shows that the relationships between government intervention and market structure are complex and that the appropriate relationship has varied over time. At times when technology is advancing slowly, the service is homogenous (POTS), and there are clear scale economies, monopoly subject to regulation has led to excellent performance. [Stone, 1997, pp. ix-x]

Indeed, there are some areas where new communications media and deregulated services have led to the need for new kinds of government oversight. The FBI has dedicated resources to Internet security after recent viruses shut down millions of computers worldwide that used Microsoft's Outlook e-mail program. Slamming (the involuntary changing of local and long-distance telephone companies) and cramming (the appearance of unordered services on telephone bills) are the subject of new efforts by the FPSC.

There are four main rationales (and a host of minor ones) used to justify government involvement in hypercommunications. Social and economic goals (or a combination of both) support each rationale. Some of the rationales for government involvement in hypercommunications are simply extensions of the regulatory philosophies that historically governed the telephone, broadcasting, and cable TV industries. Other rationales are tailored to the new policy issues created by changes due to convergence and cyber communications.

5.4.1 The Natural Monopoly Rationale

The first rationale may be traced directly to the argument that since the telephone network is a natural monopoly, it is socially and economically desirable for government to regulate it. Similar reasoning is sometimes applied to hypercommunication sub-markets or providers of access to hypercommunication networks. However, the focus on the traditional telephone network as a natural monopoly often misses important economic changes that have come because of new technologies, new services, and new kinds of networks.

Monopoly results in pricing where marginal revenue equals marginal cost leading to an inefficient undersupply at higher prices than would result under competition. Gould and Ferguson mention that natural monopoly exists when:

the minimum average cost of production occurs at a rate of output more than sufficient to supply the entire market at a price covering full cost?. The term 'natural' monopoly simply implies that the 'natural' result of market forces is the development of monopoly organization. [Gould and Ferguson, 1980, p. 248]

Figure 5-7 depicts the representative demand and cost that would face a natural monopoly. The monopolist would offer Q_NM to the market at P₂, while a competitive outcome would be more output (Q^*) at a lower price P₁.

Under natural monopoly, fixed costs represent an enormous proportion of total costs. Therefore, a new entrant could offer Q₁ at price P₃, or a much smaller quantity for a far higher price. At first glance, Q₁ might appear to be a ridiculously small amount in comparison to Q^*. However, the difference is deliberately sizable to illustrate three important points regarding telephone networks and natural monopoly. First, the telephone network must be built for peak use so that it is important to realize that the demand illustrated in Figure 5-7 is a static representation of demand at a given time period. The uncertainty surrounding demand at peak times leads the network to be built to accommodate peak demand.

Secondly, the units represented in Figure 5-7 are also important. On a per subscriber basis, some have argued:

the telephone industry is one which is sometimes pointed out as being operated under conditions of increasing costs as the size of plant increases. It is said that the cost per subscriber in a system of 10,000 subscribers is greater than in one of 1,000. The reason given is that the number of possible connections increases in a geometric ratio. [Koontz and Gable, 1956, p. 209]

Per call-minute-mile is a better unit of analysis than per subscriber as Koontz and Gable continue:

if the service given is actually taken into account, the telephone industry still seems to be one in which costs decrease with increasing size. While the cost per subscriber may increase, the cost per possible call or per call-mile-minute would seem to decrease. [Koontz and Gable, 1956, p. 209, fn. 40]

Therefore, the units of analysis have great bearing on the definition of natural monopoly.

The applicability of natural monopoly to hypercommunications access and transport is substantially different from application to the telephone network because network effects and network technologies have changed since the days of the Bell System. The technical changes that have occurred since computer networks were developed and sophisticated OR network management algorithms came on the scene may indicate that natural monopoly theory no longer applies to communication networks. Instead of a single PSTN and wireline infrastructure, new analyses of natural monopoly must consider a converging hypercommunications market in which former telephone, electricity, and cable TV monopolies compete with new entrants in multi-platform voice, data, and video communications. Until complete convergence occurs, multiple sub-markets are likely to result in competition among multi-product firms. However, there is the possibility that multilateral rivalry in multiple local markets combined with mergers and acquisitions could result in a new mega-natural monopoly.

Baumol (1997), Baumol and Blinder (1991, pp. 657-672), and Jamison (1997) each extend natural monopoly theory to multi-product firms with regard to single and multiple product markets. Jamison rigorously shows that "firms that produce products that the monopoly does not can produce some of the monopoly's products and have economies of joint production in doing so" [Jamison, 1997, p. 2]. Hence, a given firm is a natural monopoly "if and only if its production economies are greater than the economies these firms can offer" [Jamison, 1997, p. 2].

Additionally, regulatory costs and cost savings from single vendor providers of multiple services may not accurately reflect market prices according to Jamison. However, just as measurement difficulties led standard utility regulators to resort to Ramsey pricing and fully distributed costs [Baumol and Blinder, 1991, p. 661-664]. as proxies for per product marginal cost, the mega-natural monopolist's costs will be hard to calculate. Jamison admits as much by stating that testing for such mega-natural monopolies could be impossible due to the vast number of possible outcomes with multilateral rivalry and unstable technologies [Jamison, 1997, p. 9].

Historically, just as with its definition of universal service, AT&T was able to use its own definition of natural monopoly. As Stone puts it:

After the turn of the century, progressivism, one of the tenets of which was firm government control over large business enterprise, had become the dominant public philosophy. One of the critical moves that AT&T undertook . . . was to seize upon progressive sentiment and turn it to the Bell system's advantage. To do this, AT&T first embraced the theory of natural monopoly that had become fashionable in academic circles at the turn of the century. Although today an industry is considered a natural monopoly if production is done most efficiently by a single firm as output increases, the term had a somewhat different meaning in the Progressive Era. Based on the wave of bankruptcies and deteriorated service that had occurred in the traditional public service industries when multiple franchises were granted, the earlier view held that service to the public is best undertaken by a single franchisee or 'natural monopoly'. It is important to note the distinction between that conception and the later economic one. Better and more comprehensive service based on empirical observation is different from a theoretical conception that purports to predict what is most efficient. AT&T sought to clothe itself in the older natural-monopoly conception, the better to portray itself as the bearer of the public interest. [Stone, 1997, p. 30]

It is the sub-additive cost-technology relationship that modern economics recognizes as natural monopoly. However as Chapter 2 and Chapter 3 show, rapid changes brought on by the replacement of the telephone network's engineering-based model with computer-OR hypercommunication network models have created a new economic theory of natural monopoly. Therefore, to diagnose natural monopoly now, costs must be forward-looking and based on multilateral rivalry, structural changes, and other costs of "weightless economics" to diagnose natural monopoly now [Kwah, 1996, 1997]. Economides found that "the appropriate measure of cost (maximizing allocative, productive, and dynamic efficiency) is forward-looking economic cost and not the historical, accounting, or embedded cost of the incumbent's network" [Economides, 1998, p. 1]. The recognition of this fact is responsible for the use of forward-looking cost models like the HPCM by regulatory agencies today [Bush et al., 1998].

While the balance of evidence indicates that traditionally defined natural monopolies are dead in modern communications, some attempts have been made to argue that transport level costs of fiber optic hardware and conduit have resulted in natural monopoly's resurrection. Stanbury notes:

Technological change and economics, according to Huber et al. (1992), have (1) 're-created' the natural monopoly in interexchange (long-distance) telecommunications due to the economics of fiber-optic cables (high fixed costs and almost zero marginal costs), and (2) begun to make competition in local services feasible, largely due to cellular and radio technology, but also by converting cable-TV systems into LECs. The Huber et al. conclusion re: natural monopoly may hold in relation to transmission costs, but these costs are a modest fraction of all variable costs of long-distance service. Other evidence filed with the FCC indicates that MCI's total variable costs are less than AT&T's. This contradicts Hubert et al. On the matter of emerging competition in the local telephony, Huber et al. are correct, although their conclusion that there is no bottleneck or essential facilities, controlled by the telcos can be challenged. . . . [Stanbury, 1996, p. 152]

Once actual costs of fiber began to fall in the late 1990's, this idea of a resurrected natural monopoly became even less convincing.

However, as Baumol and Blinder argue: "What matters is the size of a single firm relative to the market" [Baumol and Blinder, 1991, p. 579]. A natural monopoly may be a national telephone system or a local exchange if a single firm can produce the entire output at a lower cost than under competition. This point may have special bearing on underserved rural areas.

Two additional reasons support the contention that the traditional POTS PSTN was a natural monopoly due to inherent technical and engineering reasons. First, the traditional POTS PSTN was part of a larger telephone connecting system that included physical lines and poles to telephone terminals as well as systemwide indirect costs such as research and development. The enormous fixed and indirect costs (invested on a sunk basis) discouraged entry, effectively prohibiting competition. Costs of maintaining the local loop, telephone wires and plant, machinery, equipment, and research and development were so high that only a single firm (a natural monopoly) could operate.

Second, under past technology, local competition was physically unworkable in addition to being inefficient. Historical photographs show parallel telephone lines of competing telephone companies straining poles in places such as New York City. Subscribers had to have different telephones and multiple access lines in order to receive calls from customers of every telephone company. The access and transport networks from switches down to wires on telephone poles were all duplicated [Oslin, 1992]. According to these arguments, natural monopoly is based on technical factors that result in system, scale, or scope economies.

The recent consensus is that there are three general markets within the telephone system, tending to reduce system, scale, and scope economies. First, there is a market offering access to the telephone network, second a market for switching calls between parts of the network, and third a market for intermediate transport of calls between nodes. Often, economic analyses of the telephone network have failed to recognize these complementary yet separable features of the telephone system because the traditional network architecture was part of a monopolistic telephone system that was considered a natural monopoly. The recent view of the telephone business as a set of hierarchical networks has evolved from a system orientation because of deregulation and technological change. These three markets (especially access) are involved in the hypercommunications market for voice, enhanced telecommunications, private data networking, and the Internet as well.

After an early competitive period, AT&T (named the Bell System after the telephone connecting system it owned) had monopolistic control of all three parts of the U.S. telephone market (access, switching, and transport) as well as owning all telephones. This control was partly based on a network engineering foundation. "The primary idea was that since the telcom system had to be as large and comprehensive as possible in the geographical sense, its technical integration should be based on some kind of central coordination" [Verhoest, 1998, p. 4]. Each level of the network was considered part of an overall connecting system that could break down (or lead to injury to equipment or workmen) if varying technical standards or non-standard equipment was used. It was thought that the only possible way to ensure system economies was to have a natural monopoly telephone company.

Others argued that the reason the "natural" monopoly label was attached to the traditional POTS PSTN era came when AT&T's anti-competitive market conduct combined with successful lobbying and the "capture" of the FCC and various state regulatory authorities. AT&T "slashed rates, bought competitors out, and tried to dent financing and equipment to others. An independent local phone company that was refused interconnection was cut off from the entire country" [Mansfield and Behravesh, 1990, p. 395]. AT&T itself worked out the Kingsbury agreement with the federal government in 1914 to guard against anti-trust action.

AT&T aggressively tried to prohibit competition during its entire history while venturing into telegraph, radio, television, and other businesses, retreating only when challenged in the courts. AT&T's massive research and engineering budget was spent on problems far broader than the telephone "system", including radio communication, television, development of cellular and other forms of wireless telephony, computers, and other interests. Under this system, politics, regulation, and predatory practices resulted in a technically inefficient telephone network run by AT&T.

It is also suggested that even if historically the telephone system or individual sub-networks were natural monopolies, new technology and deregulation have lowered costs and increased innovation making such a structure impossible. However, others argue that natural monopoly has merely become more complex in an era of convergence, multilateral competition, and multi-product production [Jamison, 1997].

In reality, it is debatable whether the telephone network (or the connecting system) was indeed a natural monopoly. This point is made on several grounds, regardless of any network engineering concerns. First, the consensus is that costs have been badly measured by traditional engineering and regulatory analyses. Again, because the telephone system is a collection of technologically changing interconnected networks, "forward-looking economic cost and not the historical, accounting, or embedded cost" is most likely to achieve an "appropriate measure of cost (maximizing allocative, productive, and dynamic efficiency)" [Economides, 2000, p. 1].

Second, there is empirical disagreement as to whether particular parts of the network and even the entire connecting system are indeed natural monopolies. Shin and Ying (1992) found that local exchange carrier cost structures were not natural monopolies, and, surprisingly that "it is doubtful that the pre-divestiture Bell System was a natural monopoly" [Shin and Ying, 1992, p. 171]. However, Wilson and Zhou (1997) cite a number of studies that have found that such empirical results depend on the functional form of the cost function and other econometric issues.

5.4.2 Public Infrastructure Investment Rationale

A second rationale for government involvement in hypercommunications can be expressed through the argument that today's low infrastructure growth rate is tomorrow's decayed competitive position. If that argument is true, it may be possible to stimulate economic growth (or prevent economic decline) in areas with low infrastructure growth rates using government investment (or a public-private mix) to increase the infrastructure growth rate. If the policy is calibrated correctly, both government and private enterprise can expect a revenue stream capable of repaying the government with the new tax revenue generated and providing private firms a reasonable return. This rationale is used to support a mixture of social and economic goals.

According to this argument, if government investment is used to encourage infrastructure development several effects are likely. First, there is a direct effect. The direct effect may be measured solely as benefit achieved in a particular area by the use of hypercommunications by existing firms, organizations, and individuals. Efficiencies and benefits that flow to firms from using the same amount of communications but with the new infrastructure are counted here.

Secondly, there is a multiplier effect.

The telecommunications 'multiplier effect' refers to the level of indirect economic activity created by investments in telecommunications infrastructure expansion and modernization. These indirect benefits are measured by various cost-benefit analyses and input-output calculations done to determine the value of infrastructure spending. While measurement methodologies sometimes differ, the indirect economic benefits that result from telecommunications infrastructure investments are among the highest measured. . . . A recent analysis in the US, for example, determined that investments in broadband telecommunications infrastructure, at current planned rates, will add US $191 billion to GNP over the next 16 years. [According to Cohen, 1992], policy reforms to accelerate infrastructure investments in broadband networks would add an additional US $321 billion in GNP growth over the same period. [Hubert, 1996, p. 83]

The multiplier effect may appear in several ways. First, it comes as firms that that use hypercommunications as their main inputs such as high-tech companies, call centers, e-commerce, or electronic commerce are able to locate in the local economy. Second, the multiplier effect occurs as existing firms change their business models to take advantage of positive network effects inherent in hypercommunications. Third, the multiplier effect occurs as benefits from connected educational institutions and individuals trickle down to create a more skilled information-literate workforce and stimulate job creation from new firms. The multiplier effect also depends on a holistic, positive interaction of these three ingredients.

The first two benefits of public investment in hypercommunications infrastructure are shown in Figure 5-8. The figure shows the hypothetical economic impact (gross) over ten periods from three separate sources. The impact shown might be for a region, a town, a county, a firm, or a group of firms. The periods represented might be years, quarters, or other planning periods. The figure does not show how time compression (see 2.5.2.5) is shortening the periods over which effects may be felt as a direct result of better communications and processing power. A base of one is used so the results may be clearer in this hypothetical case.

The bottom line shows the result that would occur without an infrastructure investment. Note that over ten periods, a growth of 25% has been achieved by doing nothing. This assumption may be quite overstated, especially in locations where it is probable that a decline would occur because of the lack of infrastructure. The middle line shows the direct effect of hypercommunications infrastructure investment. Instead of assuming a constant growth rate, an increasing growth rate has also been assumed. At the end of ten periods, over a 75% increase in activity has been registered. Finally, the top line shows the multiplier effect. Total gross activity has almost tripled.

Third, there is also a leverage effect of communications infrastructure development (not shown in Figure 5-8).

While the economic multiplier for telecommunications infrastructure investments is substantial, the impact of these investments in the North American environment is further enhanced because they are, for the most part, privately funded. The economic and social benefits discussed can be achieved with minimal public sector funding. The 'leverage effect' refers to the level of private sector infrastructure investment that can be stimulated by public sector investment. Telecom infrastructure programs can be triggered by policy and regulatory initiatives and by tax incentives or investment credits because investment is largely provided by the private sector capital commitments. While $1 billion in government investment in highways or airports will produce $1 billion in additional infrastructure, $1 billion in government commitments in the telecommunications sector could produce $5 to $10 billion in additional infrastructure. The impact per dollar of public sector investment on job creation and other economic gains would be even more dramatic. [Hubert, 1996, p. 83]

Thus, another policy exists in addition to direct transfers of tax funds to carriers serving rural or high-cost areas. Under the leverage effect, it is possible for the government to obtain more impact per dollar invested using tax incentives and other policy mechanisms as levers to loosen more private investment than would otherwise occur. The leverage effect can occur as a lead-lag relationship (with public funding establishing a skeleton infrastructure on which private enterprise constructs a body). This method was used to jump-start the Internet. Alternatively, the leverage effect may occur simultaneously with regulatory concessions and tax incentives.

To support the second rationale, many statistical studies are offered. Hubert mentions " a series of recent US studies" that highlight the "linkages between telecommunications and economics" [Hubert, 1996, p. 81]. He makes seven points shown in Table 5-7.

While the results found in Hubert's review (Table 5-7) from the leverage and multiplier effects are attractive, they are not without controversy. The speed with which a national economy can grow while reducing or holding unemployment down, growing real incomes, and provide for government spending and borrowing are complex economic policy problems. Many authors of macro-communications studies automatically extrapolate national economic data on to states, counties, and precise localities.

However, there can be a tendency to extrapolate more than is warranted by enthusiastic proponents of proactive government infrastructure development policies. Hence, caution must be placed upon crucial estimates of the rate of hypercommunications penetration, concomitant growth of ideas and inventions, and the economic growth potential. This becomes important when nationally based economic growth and impact estimates are projected onto narrowly defined regions. In particular, many studies assume a base economy that is urbanized.

Source: Hubert, 1996, p. 81.

Even when evidence from regional and local models is used, there is still a danger of over-extrapolation. The Pennsylvania Dutch country and Arizona's Navajo Reservation are both "rural" areas, but the extrapolation of results from one area to another should not be automatic. Within Florida, some areas have little traditional agribusiness, yet have timber reserves or rich fishery areas. In still other places, such as Homestead, large-scale agriculture and urbanization are on a collision course with ecological constraints and urbanization.

Finally, the rationales for universal service and universal access (already introduced in Table 5-6) include more than economic arguments about multipliers and leverage. Universal service is also a social program designed to eliminate the "digital divide" that separates information haves and have-nots [NTIA, "Falling Through the Net: Defining the Digital Divide." July 1999]. The right to get service at a "fair" price is enforced by cross-subsidizing rates or by other regulatory methods to cover the greater cost of providing service to rural areas, the elderly, or the poor. Universal access rights include the choice of carrier, carrier interconnection rights, and access to enhanced telecommunication services and the Internet. Some universal access rights are designed to ease the transition into deregulation for hypercommunications firms, while others are clearly normative attempts to add the right to high-speed Internet access to the existing rights of universal telephone service.

5.4.3 Normative or Social Rationales

A third rationale for government involvement in hypercommunications is purely based on normative or social goals. Information apartheid is a term used by Don Tapscott in 1994 to "splitting society into those who can communicate with the world and those who cannot," because they do not have access to the information superhighway at a "reasonable price". Social scientists argue that there are negative network effects (perhaps externalities since market prices do not reflect their costs to society) involved with increasing the penetration of hypercommunications.

However, economics is often taken as a positive (values-neutral) science capable only on analyzing costs and benefits in dollar terms designed to describe rather than proscribe the desirability of social policies such as equity or fairness [Lutz and Lux, 1988]. Other disciplines, such as rural sociology, are better disposed to rule on the normative desirability of upgrading the rural infrastructure. It is true that many Florida farmers may never use the Internet and that many in rural America would prefer to live in isolation from high technology. All an economic investigation can do is to evaluate the values-neutral positive economic pros and cons. However, just as some were suspicious of every innovation in American history from railroads and electricity to radio and TV, so now are the dangers of the Internet and hypercommunications proclaimed. Psychologists and others might argue that economics does not factor in evidence about the long-term effects on crime from watching gratuitous violence on TV.

Economists might respond that the market acts on all publicly (some say privately) available information. Many economists would argue that to the degree that the psychological evidence is true, the market would react with adjustments in relative prices. However, market prices may be unable to protect human life and property based on the danger resulting from new and ubiquitous forms of information that become more available with infrastructure development. Not all network effects are positive, as the discussion of network economics in 3.7.3 recognized.

However, there are limits to the view that economics alone provides all the rationale necessary for government involvement in hypercommunications. For example, the choir of voices singing the praise of positive economics is almost silent when the discussion turns to the "child pornography" market on the Internet, or Judge Posner's famous baby sales market [Landes and Posner, 1978]. Recent attempts to auction human organs, children, cadavers, human eggs, and genetic material on e-bay have been thwarted by e-bay itself. There is a growing outcry that some legislation is necessary to prevent many forms of economic activity from fraud to gambling to child pornography that can easily exist in virtual space. The legal issues of jurisdiction, harm, and the appropriate case law to apply are further thwarted by the international nature of the Internet.

In this sense, the hypercommunications industry's information content production and transmission markets may be seen as unique markets in which both a normative economic side and a positive economic side can be seen. The protection of children from pedophiles, pornography, graphic violence, and websites that secretly gather information on parents is a particularly strong rationale for government involvement. This could be because society as a whole (including economists) values children in non-economic ways as well as in economic ways. Furthermore, psychology and medicine are positive sciences also, but few economists would argue that legislation requiring reporting of child abuse by doctors and social workers (in spite of normal professional confidentiality) represents market failure due to governmental meddling.

5.4.4 Other Rationales: Federal, State, and Local

The fourth rationale for government involvement is actually a set of rationales derived from many sources. Federal anti-trust statutes, the desirability of e-commerce sales taxes, the national security implications of encryption export restrictions, protection of children, and free speech issues are other areas of hypercommunications regulation. The government's multiple roles of law enforcement, preservation of national security, and encouragement of domestic commerce and international trade bring in many additional rationales for regulation such as discouraging fraud and other enemies of competitive markets. Finally, often forgotten in the debate over de-regulation is that taxation policies are regulations that affect hypercommunications more than other kinds of businesses [Cordes, Kalenkoski, and Watson, 2000]. These actions can bring possible re-regulation in the form of new taxes, laws, and policies in spite of de-regulation.

Before turning to specific mechanisms by which policy is achieved, federal, state, and local agencies with some form of involvement in hypercommunications are listed. Table 5-8 shows federal agencies involved in hypercommunications policy along with specific rationales used for involvement in some phase of hypercommunications. On the federal level, there are numerous special issue publics involved in setting policy. NARUC, the National Association of Regulatory Utility Commissions is one of many national non-governmental organizations and pressure groups that seek to affect federal policy. NECA, the National Exchange Carriers Association, administers the Universal Service Fund. Rural telephone and electric co-operatives are considered one of Washington's most powerful lobbying groups [Stockman, 1986].

These groups provide legislative impetus for new rationales for regulation and re-regulation in hypercommunications. Indeed, Posner argues that "a number of standard features" of utility regulation may be the product between the regulated industry and consumer groups" [Posner, 1974, p. 351]. Most of the entries in Table 5-8 are self-descriptive. Agencies such as the FBI, IRS, and U.S. Customs Service may not regulate hypercommunications directly, but their missions provide rationales for regulations that extend into cyber space.

The TCA also was designed to allow states to continue to regulate telecommunications carriers under their own current laws, except where state statutes prevented competitive entry to the marketplace or otherwise affected FCC mandates. The TCA phrases for this requirement are given in the following passage:

Preempts any State and local statutes, regulations, or requirements that prohibit or have the effect of prohibiting any entity from providing interstate or intrastate telecommunications services. Preserves a State's authority to impose, on a competitively neutral basis and consistent with universal service provisions, requirements necessary to preserve and advance universal service, protect the public safety and welfare, ensure the continued quality of telecommunications services, and safeguard the rights of consumers. [1996 TCA]

Florida is the nation's fourth most populous state behind California, Texas, and New York. Large and small state agencies, cognizant of Florida's especially high need for communications due to its strategic location, regulate many aspects of the state's hypercommunications industry. Florida's gigantic international and domestic tourism industry, ports, and status as a Caribbean and African trading center often involve coordinated regulatory activities of federal, state, and local agencies. Table 5-9 lists the main Florida State agencies with responsibilities in communications along with their rationale for involvement.

Most important of these is the FPSC (Florida Public Service Commission). During the Broward administration in Tallahassee, the Railway Commission (1897) "became a body with real power to exercise control over all transportation and communication companies" [Patrick, 1945, p. 100]. Today, the FPSC has over 380 employees and an annual budget of just under $40 million. It is the most important state agency involved in communications, though it does not generally have jurisdiction over cable TV, wireless telephony, or the Internet.

According to the FPSC, the "primary responsibility is to ensure that customers of regulated utility services receive adequate service at fair and reasonable rates. At the same time, the Commission is required by law to see that the regulated companies are allowed an opportunity to earn a fair return on their investments in property dedicated to providing utility service" [FPSC, July 1998, p. 16].

As in the federal case, the jurisdiction of the FPSC is limited to regulated or tariffed services of particular firms. Most regulatory attention focuses on ILECs and the POTS market, on the intrastate long-distance carriers (ILECs and IXCs), or on tariffed enhanced telephone services provided by ILECs. The FPSC is particularly concerned with the actions of BellSouth, the RBOC that serves much of Florida. ALECs do not require FPSC permission to set rates on services, but each carrier must register with the state and provide a price lists.

Cablecos, electric utilities, and ISPs that offer telephone or enhanced telecommunications services must register as ALECs. The Department of Agriculture and Consumer Services (DOCS) has jurisdiction over consumer complaints regarding wireless communications, pagers, and cable TV, as well as general consumer protection. The missions and jurisdictions of other state agencies shown in Table 5-9 provide similar rationales for involvement on a case-by-case basis as with the federal agencies in Table 5-8. The mechanisms of the rural telemedicine program (health care) and universal service programs (high-cost support, low-income support, and schools and libraries E-rate program) are discussed in 5.5. The other agencies mentioned in Table 5-9 do not fulfill market regulatory responsibilities and are not discussed further here. However, that should not be taken as an indication that their contributions are not important.

The regulatory players in hypercommunications are by no means limited to federal and state agencies. Many local agencies have specific regulatory functions as detailed in Table 5-10. The rationales behind the involvement of local agencies are also shown. Again, the mechanisms through which those rationales are achieved will be mentioned in 5.5, especially under taxation. Tables 5-8 through 5-10 broadly summarize the variety of functions (and some of the rationales) for government involvement in hypercommunications. However, to appreciate government involvement in hypercommunications, it is necessary to review the main mechanisms by which government is involved in hypercommunications.

Importantly, many of the agencies listed in Tables 5-8 through 5-10 reappear through specific mechanisms of involvement in the next section (5.5) where the economic impacts of some of the main mechanisms of government involvement are analyzed. Then, section 5.6 considers how convergence and technological change alter the economics of regulation, deregulation, and re-regulation.

5.5 Mechanisms of Government Involvement

This section does not purport to trace the precise mechanism of every federal, state, and local governmental agency's role in hypercommunications. Instead, it will highlight the economics of four main mechanisms of the TCA. This entails showing economically how the regulatory mechanisms are designed to work in theory along with problems encountered as asymmetric regulation and convergence alter the regulatory mechanisms.

5.5.1 Universal Service Programs and Mechanisms

The first mechanism to be covered in 5.5 is the universal service fund, which is actually four separate programs (some with multiple mechanisms) used to accomplish policy goals. The Universal Service Fund (USF) predates the 1996 TCA, but two new programs covering new were added by the 1996 legislation. The USF is a transfer program or an indirect regulatory fee whereby certain telecommunications carriers pay money into a fund administered by an independent agency. Funds are transferred from the carrier to the fund based upon a complicated series of revenue and traffic-based formulas. Before and after the 1996 TCA the USF has subsidized carriers that serve high-cost areas and underwritten programs designed to help low-income telephone customers pay for and establish service (LifeLine and Link Up).

The 1996 TCA expanded the USF's mission so that it now pays for rural telemedicine programs, provides discounts to schools and libraries for telephone, data networking, and Internet service in addition to the traditional low-income and high-cost programs. There has been considerable debate about what kinds of services and which carriers should receive support as well as which carriers must contribute to the fund [Kaserman and Mayo, 1997].

The 1996 TCA established the universal service mechanisms with the following language:

Requires all carriers providing interstate telecommunications services to contribute to the preservation and advancement of universal service. Authorizes the FCC to exempt a carrier or class of carriers if their contribution would be "de minimis."

Provides that only designated eligible carriers shall be eligible to receive specific Federal universal service support. [1996 TCA, Sec. 254]

One aspect of the TCA continues the USF as a mechanism to transfer money from urban carriers to carriers serving rural areas. Carriers who receive high-cost support directly from the fund may do so for three separate purposes connected with local loop costs. The transfers are meant to allow high-cost carriers to be subsidized by low-cost carriers, thus furthering the goal of rate equity between urban and rural areas.

Five mechanisms provide USF funds to high-cost areas. Rules regarding qualification are complicated and change from year to year. Loop costs are calculated on a NTS (Non-Traffic Sensitive Basis). During 1999, local loops with annual NTS costs of $259 or greater (just over 1.1% of all loops in the United States) were eligible for consideration to receive support from the HCL (High-cost Loop) mechanism from the USF. In 1999, under $10 million in HCL support was awarded to Florida rural-carrier certified ILECs. The LTS (Long-term support) mechanism (also designed for NTS costs) is a second area where rural-certified carriers may qualify for USF subsidies. Over $5.2 million was received by qualifying Florida rural carrier-certified ILECs in 1999 for LTS. Finally, the LSS (Local Switching Support) mechanism is a traffic-sensitive charge meant to help compensate high-cost area switching costs. In 1999, $3.6 million was paid to qualified rural-certified Florida ILECs for LSS. The total carrier-to-carrier transfer for these three high-cost programs in Florida reached $18.7 million in Florida in 1999 [FCC, CCB, 98-202, December 1999].

Two additional high-cost support mechanisms have begun in the year 2000. First, incremental support under the forward-looking cost mechanism (HCPM) for non-rural-certified carriers will allow high-cost areas served by those carriers to obtain USF monies. Second, a new IAS (Interstate Access Support) mechanism was initiated by the FCC in May 2000 to replace the indirect funding of universal service through access charges paid by IXC customers to ILECs [FCC, Sixth Report and Order in CC Docket Nos. 96-262 and 94-1, May, 31 2000].

Another universal service program subsidizes carriers who provide Internet access, private data networking services, CPE, and enhanced telecommunications services to schools, libraries, and educational consortia. The following TCA language establishes the program and the discounting mechanism:

Requires a carrier, upon receiving a bona fide request, to provide telecommunications services: (1) which are necessary for the provision of health care services in a State, including instruction relating to such services, to any public or nonprofit health care provider that serves persons who reside in rural areas in that State at rates that are reasonably comparable to those charged for similar services in urban areas in that State; and (2) for educational purposes included in the definition of universal service for elementary and secondary schools and libraries at rates that are less than the amounts charged for similar services to other parties, as necessary to ensure affordable access to and use of such services. Permits a carrier providing such service to have an amount equal to the amount of the discount treated as an offset to its obligation to contribute to the mechanisms, or receive reimbursement utilizing the support mechanisms, to preserve and advance universal service. [1996 TCA, Title I, Subtitle A]

Generally, the USF schools and libraries transfer is from the fund to the carrier that provides discounted services to the qualifying institution. A specialized e-rate program was set up to allow qualifying educational institutions to obtain discounts of 20% to 90% (based on school lunch program poverty guidelines). Qualified carriers include ILECs, ALECs, IXCs, and ISPs. Deserving institutions were identified by the SLC (Schools and Libraries Corporation), a specially-created administrative entity operating under the following section that:

Directs the FCC to establish competitively neutral rules to: (1) enhance access to advanced telecommunications and information services for all public and nonprofit elementary and secondary school classrooms, health care providers, and libraries; and (2) define the circumstances under which a carrier may be required to connect its network to such public institutional telecommunications users. [1996 TCA, Title 1, Subtitle A]

This mechanism and the FCC rulemaking that followed allowed both private and public schools to benefit from the universal service e-rate plan. Rural schools were granted preferential consideration. In 2000, the FCC reported that over 90% of the nation's schools and libraries had Internet access (partly because of the program) and over 63% of U.S. classrooms are connected [NTIA-RUS, June 2000, p. 34]. The final USF program, for rural telemedicine uses a mechanism similar to the schools and libraries discount scheme. The Rural Health Care program funds are administered by USDA's Telecommunications Office.

The states have specific powers and responsibilities of their own with regard to the USF. The 1996 TCA:

Grants States authority to adopt regulations not inconsistent with the FCC's rules. Requires all providers of intrastate telecommunications to contribute to universal service within a State in an equitable and nondiscriminatory manner, as determined by the State. Permits a State to adopt additional requirements with respect to universal service in that State as long as such requirements do not rely upon or burden Federal universal service support mechanisms. [1996 TCA, Sec. 254]

This allows specific state regulations to be established concerning universal service support. Under FCC rules, the universal service fund relies on a mix of state and federal funding from a mix of interstate and intrastate traffic. Other state regulatory policies that are relied on to a greater degree to achieve universal service goals are analyzed economically in 5.6.2.

Rates paid for universal service are established in the following language that:

directs: (1) the FCC, within six months, to adopt rules to require that the rates charged by providers of interexchange telecommunications services to subscribers in rural and high-cost areas shall be no higher than those charged by each such provider to its subscribers in urban areas; and (2) such rules to require that a provider of interstate interexchange telecommunications services provide such services to its subscribers in each State at rates no higher than those charged to its subscribers in any other State. [1996 TCA, Sec. 254]

This passage has been responsible for an enormous amount of research, hearings, studies, and debate about precisely how to implement equitable rates. Much of the controversy surrounding the program results from charges on the bills of ILEC, ALEC, IXC, and wireless carrier customers. While not required to (and though they never did before the 1996 TCA), local and especially LD carriers began to place fixed or ad valorem charges for universal service on their customers' bills, increasing the controversy.

Representative Joe Scarborough (R-Florida), who represents the first congressional district of Pensacola and the western panhandle, addressed the U.S. House of Representatives on June 16, 1998, to protest the universal service fund, which he called taxation without representation. Known unofficially as the e-rate tax this charge began to appear on cellular, local, and long-distance telephone bills as: "Telecommunications Access System Act Surcharge", "Federal Universal Service Fee", "Universal Connectivity Charge", or by other names made up by the carrier. According to Rep. Scarborough, "the FCC used heavy handed tactics to try and stop phone companies from telling their consumers that a 5 percent tax had been passed onto every one of their phone bills secretly" [Congressional Record, June 16, 1998, p. H4575]. Through the efforts of Scarborough and other congressional critics, together with lobbying actions, the FCC ended up dramatically reducing the scale of USF contributions in 1998, but funding rose again in 1999 and 2000 when Congress became more supportive.

The TCA prescribes which carriers were allowed to receive support from the universal service fund in section 102 of the 1996 TCA.

Specifies that a common carrier designated as an 'eligible telecommunications carrier' shall: (1) be eligible to receive universal service support; and (2) throughout the service area for which the designation is received, offer the services that are supported by Federal universal service support mechanisms either using its own facilities or a combination of its own facilities and resale of another carrier's services, and advertise the availability of such services and the charges therefor using media of general distribution.

Requires a State commission to designate such a carrier for the service area. Authorizes (in the case of an area served by a rural telephone company) or requires (in the case of all other areas) the State commission to designate more than one common carrier as an eligible carrier for a service area designated by the State commission. . . . [1996 TCA, Section 102]

Especially important here is the fact that some carriers who do not pay into the USF are able to receive funds. For example, ISPs are exempt from paying into the fund, but qualified school and library ISP customers receive discounts with the difference paid from the SLC to the ISP.

Some USF programs have complicated mechanisms. Three steps may help in analyzing how the USF works in general. The first step is to concentrate on where "contributions" for the fund originate. Figure 5-9 shows the source of USF contributions by carrier type. Of the $253 million collected in total federal USF contributions in Florida during fiscal 1999, fully four-fifths came from IXCs or long-distance carriers. Fifteen percent came from interstate-related charges of LECs, with the balance made up of contributions from ALECs and wireless carriers. New high-cost mechanisms will change these sources considerably in fiscal 2000.

Although universal service contributions are not completely new (having originated after the 1984 breakup of AT&T), they are new to the bills of most customers [Kaserman and Mayo, 1997]. Before the 1996 TCA, carriers simply absorbed the charges as part of business expenses. This is because most carriers (of the types shown in Figure 5-9) began passing on the full cost of USF contributions to customers only after the 1996 TCA.

The charges were shifted to consumers and businesses as USF subsidies began to leave the traditional telephony industry and go to ISPs, enhanced telecommunications providers, and data networking connections and CPE. With passage of the TCA, universal service expanded from a basic telephone orientation to include discounts for more advanced services and technologies than telephony that are received by schools, libraries, and rural medical facilities. The total amount of contributions increased as well.

The next step in understanding the USF mechanism is to see where support dollars go. This may be looked at by program, by mechanism, in geographical terms, and by what kind of hypercommunication carriers receive funds. First, begin by considering where funds go by program. There are four universal service programs: high-cost support, low-income, rural health care, and the schools and libraries program. Each program may have more than one mechanism. Figure 5-10 shows the fund size projections for the entire United States (for the fourth quarter of 2000) for all five mechanisms of the high-cost program and for the three other programs [FCC USAC, August 2000, Appendix M5, pp. 1-2].

Roughly speaking, annual spending will be slightly less than four times the quarterly figure because of administrative expenses. Additionally, not every dollar collected will be spent in the quarter (or year) in which it was collected. USF funds for all four programs are projected to total $1.35 billion for the fourth quarter of 2000. Of this amount, $657 million will be available for high-cost support through the five mechanisms shown on the left (HCL through IAS). The schools and libraries program will be the next largest recipient with an expected $542 million. Several low-income mechanisms (Lifeline, Link-Up, incremental toll limitation, and PICC reimbursement) will represent almost $152 million. The rural telemedicine program will have $2.8 million available for the fourth quarter.

Another way of following where USF funds go is to trace the geographical and sub-market destination of each dollar collected. Figure 5-11 shows the per-dollar breakdown of USF contributions collected in Florida (and allocated to the schools and libraries) program. According to Universal Services Administrative Company data, over $105 million in total schools and libraries contributions were received from Florida sources during the July 1, 1999 through June 30, 2000 funding period [FCC CCB, Section 4, Table 4.4, p. 4-11, 2000].

Of this total, almost forty percent of total SLC contributions went out of Florida. Of funds staying in Florida, over $5 million went to ISPs, while almost $25 million (29% of total contributions) went to enhanced telecommunications carriers (ILECs, ALECs, and IXCs). Over $34 million (32%) was spent on private data networking and internal connections [FCC CCB, Section 4, Table 4.2b, p. 4-8]. In Florida, SLC commitments went to 813 different projects during the period.

The third step in understanding universal service is to study how particular mechanisms operate economically. For example, agribusinesses in underserved areas may be especially interested in how mechanisms in the high-cost program work. While each high-cost mechanism has a slightly different function, the general answer to this question is that all five mechanisms transfer part of the total USF charge from urban customers to rural customers.

Figure 5-12 shows how this works using data for the three high-cost program mechanisms that were in effect in 1999 (HCL, LSS, and LTS). Suppose that the right panel represents the urban local telephony market. The left panel depicts the much smaller rural local telephone market (not correctly scaled to quantity). The discussion in Chapter 6 on market boundaries will show that there are not simply two such markets, but assume that is the case for now. The example also assumes that the total is collected on access lines alone. While Figure 5-9 shows that reality contradicts this, recent reforms will make the scenario shown in Figure 5-12 more realistic.

A free market would allocate a quantity of Q₁ at price P₁ in the urban market with a far higher price of p₁ and much smaller quantity of q₁ in the rural market. USF high-cost loop contributions in Florida averaged $0.87 per month per USF-qualified local telephone line [Eisner, FCC, 2000, Table 1.5, p. 21]. Carriers (not the government) decide how customers will be charged. If the carrier figures the contribution on a percentage of total bill basis rather than a uniform monthly amount, the contribution would resemble an ad valorem tax. Hence, the supply price plus contribution would result in an upward shift from S₁ to S₂ and a steeper slope. The result would be that the urban market quantity would fall from Q₁ to Q₂ (Q₂ <Q₁). A further assumption is that carriers can simply pass all the USF charges onto consumers, something that has been observed in practice possibly because demand for local service is inelastic, often as low as -0.25 [Taylor, 1994, Sappington and Weisman, 1996].

In Florida, the entire $0.87 per loop contribution does not reach the rural market to create the changes seen on the left-hand panel. Instead, $0.14 of the USF high-cost loop support would stay in the state, with $0.72 leaving Florida to support high-cost loops in other state. Nonetheless, the large number of urban lines paying contributions is responsible for an average subsidy (USF support payment) per qualified rural-certified carrier loop of $8.75 in Florida [Eisner, FCC, 2000, Table 1.10, p. 26]. as shown in Figure 5-12. The result would be a drop in price from p₁ to p₂, accompanied by a greater number of rural telephone customers with quantity rising from q₁ to q₂.

Several additional points should be noted. First, only the high-cost USF support mechanism is shown in Figure 5-12. Importantly, an additional $0.24 is assessed on average per loop for low-income support and $0.80 per loop for schools and libraries, along with charges of under $0.05 per loop for rural telemedicine and $0.13 for the telecommunications relay service. Figure 5-12 considers the direct effect of the high-cost support USF mechanism alone. Other regulatory policies (5.6.2 and 5.6.3) are used to balance rates between urban and rural subscribers of non-rural-certified carriers. The new HCPM mechanism is the first direct USF method of intra-carrier transfers for non-rural-certified ILECs that have ten times or more rural lines than rural-certified ILECs do. Analysis of such intra-carrier rate averaging is covered in 5.6.2.

The three USF high-cost support mechanisms shown in Figure 5-12 affect up to 193,000 customers of rural-certified ILECs in Florida. In some cases, the USF does not lower the rural price to the same price as the urban price in Figure 5-12, nor would that normally be the case. The FPSC sets per-carrier access rates for rural-certified carriers that are slightly higher than for non-rural-certified ILECs. USF is not the only mechanism used to work towards the goal of approximate equity between urban and rural rates. The precise rates would depend on other regulatory policies (covered in 5.6) that have several goals. Among these are general rate regulation, assurance of a fair return (or reasonable price) to the serving ILEC (whose conduit and facilities are used whether the ILEC or a competitor provides service), and approximate urban-rural rate equity.

Additionally, as was mentioned, court decisions have been made ordering future contributions to be collected from the interstate portions of carrier revenue instead of a 75-25 mix of interstate and intrastate. Hence, Figure 5-12 would have to be redrawn to show the transfer from urban IXC customers to rural local customers. While this can be done on a loop basis, long-distance carriers vary by how they charge customers for USF contributions, complicating the analysis.

Furthermore, there are second-order effects not shown in Figure 5-12. The decrease in customers in the urban telephone market because of the price increase due to the USF would be mitigated by low-income support programs. To the extent that the movement from Q₁ to Q₂ was due to a loss of low-income customers, the $0.24 per loop low-income support USF contribution ($0.09 of which stays in Florida) could be expected to subsidize low-income customers. The result might negate part of the urban exodus (and further increase rural service).

5.5.2 Taxation and Regulatory Charges

Taxation and regulatory charges are a second mechanism of government involvement that includes taxes, fixed charges, and proportional (ad valorem) fees. Some of these come from the 1996 TCA or other federal legislation, while others are assessed at the state and local levels. Since over 90 different taxes and regulatory fees are assessed by several hundred different jurisdictions on different services, different carriers, and different customer classes, a series of tables is needed to summarize these mechanisms. Table 5-11 gives codes to keep track of the markets, carriers, and customer categories that are subject to the regulatory fees shown in Table 5-12 and the taxes in Table 5-12. Table 5-14 is meant as an approximate summary of the various taxes and regulatory charges that can appear on communications bills for Florida agribusinesses.

The first entries in Table 5-11 (labeled M1 through M8) denote regulatory attempts at defining separate markets. In reality, the regulatory definitions of each market are more involved, but those listed represent the clearest way of separating them. Many taxes and non-tax charges affect only a particular market or markets.

The second group of entries (SP1 through SP11) represents the service providers. For most taxes, the service provider collects the tax from customers and forwards it to the appropriate government agency. For most other charges, the service provider pays the charge to the appropriate jurisdiction and has a choice as to whether absorb the charge or pass it on to customers. There is an immense burden on service providers of taxes and regulatory fees. Nationwide, there are 10,857 jurisdictions that tax communications, with carriers required to fill out 55,748 tax forms [Cordes, Kalenkoski, and Watson, 2000].

The third group of entries shows specific groups of customers. While each tax or charge depends on the market or provider, the customer groups (defined in POTS and enhanced telecommunications terms) are shown separately to make three points. First, many taxes and charges are paid only by business customers. Second, some taxes or charges (especially in the highly regulated telephony and enhanced telecommunications markets) will depend on how the agribusiness makes purchases from a service provider.

For example, a firm with twenty separate telephone lines (purchased on a multi-line basis) has a different tax and charge exposure than a firm that buys twenty ALEs (Access Line Equivalents) or derived lines on a trunkside basis. Finally, under old rules, rural customer groups pay the same amount that similar urban groups do in taxes and charges except that there are no city jurisdiction charges in the rural case. New regulations may allow rural and low-income customers (or all customers including businesses of smaller ILECs) to pay smaller amounts for particular charges.

Table 5-12 details non-tax financial mechanisms of government involvement in hypercommunications. Nationally, over 300 separate taxes and fees are assessed on almost 700 tax bases, so any table is a simplification [Cordes, Kalenkoski, and Watson, 2000]. For each regulatory charge, columns identify the main payer or payee (if not a direct customer transfer), the affected customer groups, and markets. Most of the charges listed there are collected for wireless or wireline local or long-distance telephone service.

Access charges originated as a mechanism with the breakup of AT&T designed to compensate local telephone monopolies (ILECs) for the use of the local telephone infrastructure in originating, completing, switching, and holding lines open for long-distance calls. These flat fees charged to each local subscriber for connection to the long distance network were intended to prevent burdens on local phone companies that would otherwise have been covered by increasing business rates even more above residential ones.

Until recent FCC action [FCC 00-193, 2000], two access charges were assessed on all residential and single and multi-line business customers of ILECs. Businesses with local or long-distance voice (switched ISDN T-1's or point-to-point non-switched T-1's) or with local or long-distance special access lines (such as data T's or frame relay connections) did not pay such charges directly. Instead, prices for such derived lines were subject to special caps on pricing that could be used indirectly to collect access charges or by 5:1 or 9:1 ratios.

Both the SLC (Subscriber Line Charge) and the PICC (Presubscribed Inter-exchange Carrier Charge, pronounced pixie) compensate local carriers for access to LD networks. Specifically, both compensate ILECs for traffic-sensitive (TS) and non-traffic sensitive (NTS) costs of operating the local PSTN for interstate calls. The SLC is assessed on each local CCL (Common Carrier Line) or may be charged in ratio-derived units for blocks of derived lines (ALEs). The SLC also is called a federal subscriber line charge, since it is regulated and capped by the FCC. However, the SLC is not a tax since all monies go from customers to ILECs.

Until July 2000, primary residential line SLCs were capped at $3.50 per month, non-primary residential at $6.07, and multi-line business lines were capped at $9.38 per line per month. A monthly fee of up to these amounts would be charged to all lineside customers of ILECs and embedded in prices used to resell ILEC lines to ALECs. Since the charge did not vary with traffic, it was meant to recover TS costs of using the ILEC local loop for interstate traffic.

The PICC charge is paid to LECs by IXCs. The PICC charge originally replaced a TS charge (the CCLC). However, the PICC became a mechanism through which LECs could recover NTS costs of interstate use of the local access and local transport networks from IXCs that had not been recovered through the SLC. As such, the PICC was an indirect subsidy because long-distance carriers were assessed the PICC charge by the local telcos on a per-line subscribed basis, regardless of how much that line used long-distance.

While all IXCs must pay PICC charges, the PICC charge has been passed on to customers in diverse ways. Some carriers do not collect it at all; others collect it based on call volume, while still others collect a fixed amount on a per-line basis. Additionally as shown in parentheses in Table 5-12, an ILEC may assess a PICC on a local customer who does not choose a long-distance carrier. The scheduled PICC cap for primary residential lines in July 2000 is $1.56, while non-primary residential lines may have a PICC of up to $3.58 per line per month. The multi-line business PICC cap had been scheduled to rise to $5.90 per line per month in July 2000 (before the CALLS program was enacted).

Under the CALLS (Coalition for Affordable Local and Long-distance Service) plan for access charge reform (approved by the FCC in late May 2000), all residential and single line business PICCs have been removed. Additionally, the multi-line business PICC cap is lowered to $4.31. However, residential SLC caps rise to $4.35 for primary lines and $7.00 for non-primary lines, while multi-line business SLC caps fall slightly to $9.20 per line per month through 2004. The net effect of the CALLS plan is to shift the burden of access charges away from long-distance company customers (especially from low-volume customers) and towards local carrier customers.

Indirect access cost recovery is an even more complicated subject when trunkside or special access services are considered. On the surface, it might seem as though the ALE (Access Line Equivalent) basis would be the most efficient way to collect the access subsidy on switched and non-switched trunk lines. As was discussed in Chapter 4 (4.7 through 4.9), many kinds of trunks and multiple trunks use ILEC conduits, poles, and equipment to transmit local voice, long-distance telephony, data, and Internet communications. Furthermore, many businesses purchase services from carriers who utilize the ILEC's local loop or bypass that loop entirely with their own facilities. However, different kinds of access technologies have different impacts on local access and local transport networks.

For this reason, the FCC defined certain "baskets" of services provided by local carriers to IXCs or to customers. The service baskets are an attempt to clarify the sources and uses of direct access charges (such as the SLC and PICC), along with indirect access charges such as price-cap regulations. Based on whether a telephone, enhanced telecommunications, or data circuit fell into a common line (CMT) basket, TS switched interstate access basket, trunking basket, interexchange corridor basket, or special access basket, an appropriate mechanism(s) would be chosen to recover network access costs [FCC 00-193, p. B-26-B-27].

Consider, for example, three examples of business connections that use some part of the ILEC's local loop. Suppose that one business has twenty-three telephone lines through a switched ISDN or T-1 trunk, while another had twenty-three individual lines. For technical reasons, the access costs related to the trunkside services are a small fraction of those of the multiple-line (lineside) costs. Lineside costs are mainly in the CMT and TS basket, while trunkside costs have no TS component for the ILEC if the trunk's long-distance calls are not switched through the CO, but instead are switched at the IXC's POP. Most trunkside services are carried over ILEC facilities (whether sold by the ILEC directly or resold by an ALEC). However, the cost per trunked ALE might be below 5/23^rd to 9/23^rd the multi-line total (for technical reasons explained in Chapter 4). Other kinds of lines such as a frame relay T-1 do not even carry telephone calls and have a different underlying cost structure for using ILEC facilities to access the ILEC or its competitor's frame relay network. Such data lines or other kinds of special point-to-point connections are examples of services from the special access basket. Competition for these enhanced telecommunication and private data networking customers is fearsome enough that the PICC and the SLC are often not charged by IXCs and ALECs.

The issue of derived lines is so complicated it cannot be explained completely here. Importantly, though, customers of retail ALECs along with certain special access and trunking basket services are regulated indirectly through wholesale price-caps on ILECs. After adoption of the CALLS proposal, the Commission used a special X-factor (whose definition changed under CALLS) to seek indirect reimbursement for trunked and special access costs through changes in price-caps for certain ILECs [FCC 00-193, 2000, pp. 54-78]. Undoubtedly, this was in response to what the Commission had worried about in 1997:

We are concerned that assessing PICCs on multi-line business lines may create an artificial and undue incentive for some multi-line customers to convert from switched access to special access to avoid the multi-line PICC charges. A migration of multi-line customers to special access could significantly reduce the amount of revenue that could be recovered through per-minute charges, and would result in higher PICCs for the non-primary residential and multi-line business lines remaining on the switched network. We tentatively conclude that we should therefore apply PICCs to purchasers of special access lines as well. [FCC, 97-158, paragraph 103]

However, after seeking comments on how to do this, the commission failed to discover a mechanism by which the special access basket could be so regulated at the time.

LNP is the next special charge in Table 5-12. LNP is a fee charged per multi-line to allow telephone numbers to stay the same even if a business or residence switches local telephone companies. One barrier to competition before the 1996 TCA was the fact that the numbering system was set up so that a business' telephone number would have to change if its local carrier changed. LNP charges vary based on the costs incurred by the ILEC. ILECs are allowed (but not required) to charge a LNP charge. For trunkside services, the nine (ALEs) to one common line ratio is not used for businesses with PBXs. Instead, the LNP is per ported line [FCC, "Local Number Portability Factsheet", January 1999]. LNP applies only to changes in LECs, not to changes in location. Therefore, a business that moves outside local exchange boundaries cannot be guaranteed the same telephone number whether it changes LECs or not. LNP charges can apply for a maximum of five years.

The next charge is the 911 charge. This charge is made on all single or multi-line business customers. For switched trunked lines, it may be charged using the nine to one ratio, though lines sold by certain ALECs or multiple trunks may be exempt. The amount varies by county and depends on the cost for the ILEC and county to set up the 911 system. The per-line charge is normally less than $0.50 per month.

The telecommunications access system surcharge may be charged by ILECs to fund the telecommunications relay system for the deaf. It normally is under $0.20 per month per line for single or multiple lines. It may be collected by ILECs to reimburse their costs of running the relay system.

The universal service charge has already been discussed. The CALLS proposal adds a new USF mechanism (the IAS) to replace the implicit access charge subsidy from universal service in interstate rates. Hence, in addition to providing intrastate high-cost support and schools, libraries, and rural telemedicine support, a new mechanism is created to ensure interstate access universal service, while funding for low-cost support is absorbed (rather than charged to customers) by IXCs. The new mechanism will collect $650 million per year from LEC customers rather than the current method where long-distance customers indirectly support universal service through access charges [FCC, Sixth Report and Order in CC Docket Nos. 96-262 and 94-1, May 31, 2000].

Several minor mechanisms are shown in the remainder of Table 5-12. The first of these is the payphone surcharge. This charge was established to compensate payphone service providers for completed interstate and intrastate calls that use their payphones. Most calls from payphones are under contractual arrangements between pay providers, IXCs, and LECs. However, calls to toll-free numbers and calls that use payphones to access long-distance carriers (dial-around, calling card, or toll-free IXC lines) receive a 28.4 cents per-call completion charge from the IXC handling the call. In some cases, these charges appear on a business telephone bill. Pager users are also charged a monthly charge for 800 number calls (or other toll free numbers) and for use of 800 numbers through an NTW-800 charge.

Calling plan or dial-around surcharges sometimes appear on bills with the word surcharge to suggest that government regulations are responsible for the fee. That is not the case. Calling plan charges are assessed by IXCs for certain plans whether or not any minutes are used while dial-around surcharges occur when an IXC's PIC code (10-10 plus a unique carrier number) is used to dial-around the pre-subscribed IXC carrier.

Finally, wireless telephone service has a landline local access fee (essentially an SLC) and a national access fee (similar to a PICC). These charges are payments from the wireless provider to a LEC or IXC to compensate those carriers for calls that originate or are terminated on wireless telephones but also use other carriers' networks.

The taxes on hypercommunications (shown in Table 5-13) are more easily figured than the special charges on hypercommunications in Table 5-12 because they are based on the total bill for specific services, rather than on any kind of indirect method. Both taxes and charges are summarized with their effect on lineside and trunkside telephone service in Table 5-14. Before summarizing the financial impact of the taxes and charges on agribusiness, specific taxes and charges paid to government (listed in Table 5-13) are described.

The first tax is the federal excise tax on communications services. This tax originated as a temporary luxury tax on telephone service in 1898 to pay for the Spanish-American war. The rate rose as high as 25% on long-distance calls in 1944, falling to 1% in 1982. In 1983, the federal excise tax was raised to 3% for toll calls and 2.7% on local service. Federal excise tax does not apply to private communication service including accessory services provided with an office PBX. Therefore, special rules apply concerning which enhanced telecommunications services are taxable [IRS, Publication 510, 2000]. The tax does not apply to Internet access, private data networking, or to WATS charges.

Florida has two state taxes that are assessed on particular hypercommunication markets. The first tax is the sales and use tax. During 1999, of $17.4 billion in gross sales by communications companies and $11.2 billion in net sales, over $736 million in sales and use tax was collected. The tax rate is 7% for toll calls, cellular, beeper, SMR, and two-way cable TV services. A 6% sales and use tax applies on installations, sale and rental of communications devices, or leases of communications equipment such as telephone systems [State of Florida, Department of Revenue, 2000]. In October 2001 a new communications services tax (s. 202.12, F.S.) will take effect designed to tax the actual cost of operating a substitute communications system and on satellite service as well. Under a 1997 law (s. 212.05(1) (e) 1. a., F.S.), sales tax cannot be imposed on Internet access, e-mail, or other on-line services. Nationally, goods and services purchased online are exempt from sales tax until October 2001. In May 2000, federal legislation passed the U.S. House extending a moratorium on Internet taxes until 2006. A similar Senate bill has not yet passed.

The second state tax is the Florida gross receipts tax. The gross receipts tax raised $304 million from telecommunications in Florida in fiscal 1998-1999. The gross receipts tax does not apply to private data networking, WATS, or Internet access. The tax rate is just above 2.5%.

Under current law, Florida counties and municipalities may tax local telephone service (net of access charges and other fees) up to 10% or broad-based intrastate telecommunication may be taxed up to 7%. The collection of local option taxes (also called municipal public service taxes) is based on the service address. The collection of such municipal taxes and the definitions of service providers and markets vary considerably. In some locations, tax authorities argue that the way municipal tax law is written it favors ILECs over ALECs for private line services (enhanced telecommunications). Sharon Fox, Tampa's Tax Revenue Coordinator, stated in 1996 that "barriers to entry" such as taxing an ALEC private line at 7% and an ILEC's at 0% seem to be "forbidden by the Telecommunications Act of 1996". However, she added, "municipalities must follow all provisions of the (state tax) law. . . ." [Fox, 1996, p. 4].

The result has been a system under which taxes on hypercommunications are haphazardly collected in some parts of the state. Problems with address locations, service provider types, and markets are too complex for some local agencies to deal with. Generally, most municipal 10% taxes exclude Internet or any other service where communication does not originate and terminate within Florida. Under the new communications service tax scheduled to take effect in 2001, no municipality may collect a higher tax rate than was collectable on July 1, 2000.

Franchise fees are collected by local jurisdictions for right-of-ways for cableco and wireless providers. Additional local city or county property or special assessment taxes and surcharges may be collected in some instances in addition to the taxes mentioned.

Some argue that many benefits of deregulation have been canceled by increases in taxes and fees. Long-distance savings have been "virtually eliminated" by increases in the FCC equal access charge, PIXIE charge, and a 6% tax on long-distance from the ten-year old state tax on telecommunications (local and LD). Often such local taxes are not itemized as to city or county and are often billed incorrectly even when numbers or jurisdictions change [Taylor, 1999].

Table 5-14 attempts to summarize the impact of taxes and regulatory fees on Florida agribusinesses depending on whether multi-line POTS or trunked enhanced telephony have been purchased. The impact of taxation and fees on the agribusiness' hypercommunications bill is multi-faceted. Taken together, the monthly costs for businesses of taxes and charges can approach 40% of the local telephone bill and exceed 25% of the long-distance bill. Florida has one of the highest average burden of taxes and fees of any state, estimated at 28.5% of the average residential bill (higher for multi-line low-volume LD businesses). The result is that Florida is ranked second nationally in taxes on telecommunications [Cordes, Kalenkoski, and Watson, 2000]. The tax and fee burden is less for enhanced telecommunications services, still less for private data networking, and almost non-existent for Internet access.

First, ILEC customers are likely to pay higher charges for local telephone service than ALEC customers. Local single-line and multi line telephone customers tend to pay more than customers who choose switched ISDN T's or T-1 point-to-point lines even if the ALEs are equal (each has the same number of lines).

As Table 5-14 shows, the 1:9 (a 1:5 ratio in some cases) ratio will keep trunked costs lower for businesses equipped with a suitable PBX than for businesses with multi-line or key systems. Second, businesses that bypass the ILEC switch and use special trunked access to switch long-distance through an IXC POP benefit as well. For example, a business that purchases a long-distance T or fractional T avoids PICC charges entirely (except through a small allowance on the rate that the ILEC charges the IXC for calls over the local loop). However, long-distance rates are competitive enough that the volume of calls ensured through this arrangement guarantees the business a still lower rate. Furthermore, since ALECs do not have the same interstate obligations, they sell such private trunks for a lower price, allowing businesses to save taxes on the difference in total charges in the process.

A third effect of taxes and regulatory charges is that enhanced telecommunications market benefits and the POTS market (in particular the ILEC POTS market) suffers. Another result of asymmetric taxation is that private data networking and Internet access markets benefit even more than enhanced telecommunications. Frame relay connections and Internet T's still use ILEC infrastructure. However, since these "special access basket" lines cost ILECs less than switched or dedicated voice lines and since total charges are lower taxes as well. Many fixed fees do not apply at all for special access circuits.

Finally, another result of high taxes is that Florida (with the second highest tax rate on communications in the country, 28.46%) disadvantages agribusinesses (and other businesses) located in the state. State and local taxes in Florida represent an average of 24.47% of the average residential single-line telephone bill. Idaho (the state with the lowest burden) has an average state tax of 3.94%, which amounts to less than 8% of an average residential telephone bill when federal charges are added in. Georgia's state burden is 18.98%, Alabama's is 19.89%, and California's stands at 15.99%. Only Texas, with an average state and federal tax rate of 28.56%, is higher than Florida [Cordes, Kalenkoski, and Watson, 2000].

Short of leaving Florida, agribusinesses can reduce tax burdens by switching fax and data traffic off POTS lines or enhanced trunks. By doing so, businesses further reduce overall communications costs, avoid PSTN-based taxes and regulatory charges, and lower their total tax bills. In addition, firms that use IP PBX solutions based on VOIP and VOFR (see Tables 4-39 and 4-40) cut bills, taxes, and charges even further. The greatest savings occur for agribusiness communications that originate and terminate within a VPN (4.9.7) since private networks are subject to fewer taxes. However, many such alternatives have still have significant QOS issues, making them questionable for some business-class communications. Even if the technologies improve, most PSTN calls and fax traffic still terminate on PSTN lines (at the recipient's location) so that some PSTN costs still must be borne by the calling party.

A more specific breakdown of the costs and savings involved will be found in Chapter 7. Before providing a more general model of the policy implications through several economic analyses in 5.6, two last mechanisms are used by governments in hypercommunications.

5.5.3 Carrier of Last Resort

Another mechanism of government involvement is the COLR (Carrier Of Last Resort) provision. Currently, COLR obligations exist for ILECs in Florida to provide basic telephone service. As competition occurs in the LEC market, universal service and other subsidies (such as the charges shown in Table 5-12) tend to be charged only by ILECs. However, customers (especially those who are most heavily responsible for financing cross-subsidies) will tend to leave the ILEC if prices are lower with ALECs (who do not have COLR responsibilities).

An important question concerns whether COLR obligations should be set at high or low levels as competition occurs in local service. This is particularly relevant when considering competition in hypercommunications between ILECs and other COLR utilities (such as cablecos and electricos who provide telephone service using cost structures that arise from COLR schemes for electricity or cable TV). If an ILEC COLR carrier is obligated to reach all customers in a service area, then COLR providers of electricity or cable TV might have to be able to serve all hypercommunications demand for their COLR customers.

According to Weisman, the level of the COLR obligation sets the reliability of the COLR carrier and its competitors [Weisman, 1994]. According to the Weisman model, the optimal COLR obligation is less than 100%, but greater than 0% because if the COLR obligation is zero, competitors will undersupply reliability. However, if the COLR is set to 100%, competitors will oversupply reliability since they do not take into account the prices needed to cover the COLR carrier's cost of excess capacity [Weisman, 1994, p. 97]. Convergence would tend to exacerbate the inefficiencies of COLR provisions.

There are additional risks for COLR carriers due to convergence. For example, ILECs are already required to provide universal service. However, they could also be required to provide high-speed Internet access and other hypercommunication offerings (including traditional and enhanced telecommunications services) if states or federal authorities expand the obligation to universal access. Florida law and FPSC policy requires COLRs to extend service to distant but not extremely remote places. It is not easy to differentiate between distant and extremely remote.

One case that shows the difficulty of the distant versus remote COLR distinction concerns Dog Island, Florida. In 1985, homeowners on the island off the Florida panhandle (near Carrabelle in Franklin County) requested telephone service from their ILEC, the St. Joseph Telephone, and Telegraph Co. It was not until 1993, that fifty island residents made known their (still unacted upon) requests to the FPSC. Due to a variety of concerns regarding the environmentally sensitive storm-prone island, underwater cable facilities from Carrabelle (four to six miles away) could not be constructed. Such a submarine cable would cost $750,000 to extend service to a maximum of 125 households at a cost of six thousand dollars each. However, the ILEC could only charge the normal per residential line charge of under $20 per month to recover costs. St. Joseph's was ordered to extend service (in keeping with its COLR obligation) using a radio transport link to shore from a wireline CO on the island at a cost of several hundred thousand dollars [Dog Island VS. St. Joseph Telephone Co., Docket #950814-TL, Order No. PSC-97-1196-FOF-TL, FPSC, October 2, 1997].

The FPSC has held that ILECs are required to provide non-discriminatory service throughout their territories. Hence, ILEC facilities must be extended to offer Universal service to "consumers in all regions of the nation, including . . . those in rural, insular, and high-cost areas". Furthermore, "those areas should have access to telecommunications and information services . . . that are reasonably compared to services provided in urban areas. . . ." [47 U.S.C. § 254(b)(3), 1996 TCA]. Yet the FPSC does not appear to think that the current FCC definition of universal access (200 kbps) will become workable in Florida anytime soon.

Recently, the commission staff found that "over 60 percent of Florida wire centers were incapable of providing xDSL service. Until the LECs are further along with deployment of this capability, no redefinition of local exchange service" or requirement for high-speed service "should be considered" [FPSC, 1999, p. 1]. The report points out that the "average data rate available to consumers will rise to 56 kbps" in the next three to years. According to FPSC staff, all that is needed is monitoring of the situation "to determine whether a redefinition of basic service might be appropriate in the future" [FPSC, 1999, p. 1].

5.5.4 Spectrum Allocation

Another important mechanism used to affect hypercommunications is spectral allocation. Because there can only be one user of a particular frequency at one time, spectral allocation has been left to the FCC (Federal Communications Commission) in the US. According to Francois in 1978, there are four essential concepts that provide the rationale for spectral allocation: public ownership, scarcity, media differences, and fiduciary [Francois, 1978, pp. 436-437]. By this multi-pronged rationale, the federal government was assumed to have the right to allocate spectra ranging from sub-sonic physics particle generators through infrared light.

Many of the dimensions of spectral allocation can be seen in Chapter 4 in Figures 4-30, 4-32, and 4-33 where the location of various wireless services in particular bands can be noted. Spectral allocation is being used in several ways to ensure competition and increase high-speed services. First, in recent years the FCC has transferred from government and scientific use huge parts of the frequency band to accommodate new services. The new upperband frequencies (generally above 20 GHz) are especially attractive because they offer greater bandwidths so that high-speed services may be accommodated.

A second way spectral allocation serves as a mechanism comes from specially controlled auctions held by the FCC when new frequencies are released. The auctions are set up to maximize the economic efficiency of the results. A third way spectral allocation is used as a mechanism is the relaxation of the spectrum cap in rural areas. According to the FCC, the "spectrum cap governs the amount of CMRS spectrum that can be licensed to a single entity within a particular geographic area" [FCC, Wireless Bureau, September 15, 1999, p. 1]. By relaxing the spectrum cap in rural areas, the FCC feels that certain firms will be encouraged to enter those areas to provide 3G high-speed wireless services. Additional discussion concerning the boundaries used by the FCC to determine a rural area within which different kinds of wireless carriers may be licensed is found in Chapter 6.

5.6 Convergence and the Economics of Regulation

This section starts by presenting a general conceptual model (5.6.1) that is then used to examine three topics in convergence and the economics of regulation of hypercommunications. The issues of rate rebalancing and rate averaging to provide universal service in a dynamic converging marketplace are covered in 5.6.2. A general summary of the problems of using incentive and traditional utility regulation to achieve policy goals in hypercommunications is given in 5.6.3. Finally, a general analysis of the possible economic effects on agribusinesses of status quo policies involving asymmetric taxation and regulatory charges is the subject of 5.6.4.

5.6.1 Conceptual Model

Until now, the discussion of rationales for government involvement in hypercommunications and the mechanisms used to make policy have focused on complexities in the current marketplace save one important exception. The converged hypercommunication market has been ignored in favor of separate markets for POTS, enhanced telecommunications, private data networking, and Internet. Each market has been examined technically (in Chapter 4) and in regulatory terms (in Chapter 5) as though it could be separated from the others. As technologies progress, such separation becomes less possible. However, regulators continue to treat one group of services differently from others.

The problems with this asymmetric approach to regulation are illustrated through Table 5-15. There, an example given by Fox in 1996, is extended to show the ways that a person employed at a firm located in Tampa might communicate with an associate in the Jacksonville branch office [Fox, 1996, p. 9]. The example is important to agribusiness managers because it encapsulates the technical, economic, and regulatory factors of hypercommunication networks that have been established from Chapter 2 until this point. Table 5-15 (and the conceptual model it gives birth to) highlight two important issues for agribusiness managers. First, hypercommunications policy has an impact beyond rural infrastructure development. Regardless of the rural infrastructure, taxes and regulatory charges appear on bills each month, even up to forty percent of the total cost in some cases. Competitiveness, QOS, and pricing are all affected by utility regulation policy as well.

Second, because of these regulatory impacts on agribusinesses' bills many counterintuitive results occur. Rural businesses pay higher rates than urban residents even as urban businesses and residences pay higher rates than they otherwise would to subsidize rural access. Competitors of local telephone monopolies become more attractive since they may not collect certain regulatory charges. Furthermore, as convergence allows businesses to communicate by voice, video, Internet, fax, and e-mail without using the telephone network, additional advantages can flow to innovative firms. If an agribusiness can buy more communication for less than half the price and reduce communication taxes by an even greater percentage, it bears looking into. However, there are QOS and financial risks involved.

Table 5-15 displays four classes of communications ranging from real-time voice conversations (CLASS 1) to fully interactive collaborative two-way simultaneous voice, text, data, graphics, and video hypercommunication (CLASS 4). Within each class, several examples of service offerings are listed that are capable of delivering that overall class of service.

In general, the older the hypercommunications technology (within each class or among classes), the greater the amount of regulation and taxation. For example, single-line local or long-distance telephone calls (Example 1A) have far more regulation and taxation than simultaneous voice, text, data, graphics, video, and collaborative networking using Internet access.

There are three ways to analyze Table 5-15: within classes, among classes, or on an offering-by-offering basis (as hypercommunication bundles). First, consider comparisons among the service offerings (examples) within classes. Here the comparison is among different technologies capable of delivering the same message type. Suppose the price of one service offering equaled that of an extremely close substitute offering (capable of delivering the same message) within the same class. If the tax produced significant differences in costs, the higher-taxed provider would be expected to lose customers to less-taxed competitors. The same would be true if due to regulatory reasons, one offering arbitrarily charged higher prices.

Of course, an important assumption is the word "close substitute". As Chapter 4 examined in detail, the QOS characteristics of a particular message type can vary dramatically by technology and service provider. Since the equivalent service offerings are provided by different technologies, all costs (economic and accounting) must be considered. This point is important because most service offerings within classes or among them are substitutes, but rarely are they perfect substitutes.

When the second level of analysis, comparison among classes is considered, units of analysis (both prices and quantity) can vary among classes, making analysis difficult. Instead of calls or call-minute-mile, other units, such as kbps transmitted (data rate) or bandwidth (capacity) may be used. Here the comparison is among different message types. However, if the alternatives to one class were close communication substitutes and offered additional services for no additional charge, highly regulated and taxed classes would become less popular because customers could pay less and receive greater value.

The variable cost of communication tends to fall for entries as the table is descended, both within each class and among each class. For example, within example 1, a ten minute overseas telephone call (via POTS single-line) might have a low monthly recurring cost, but a high variable cost when compared to any alternative voice communication method below it. However, as successively lower voice examples in the list are considered, the greater the total monthly recurring costs would be relative to variable or per message costs.

The third level of analysis would compare hypercommunication bundles within a class and among classes (each entry in the table with all other entries). The advantage of this characteristics-based method is that within-class and among-class differences could be analyzed, along with any interaction effects. A simple conceptual model (interaction effects not shown) of the economic costs of choosing a hypercommunications bundle from a given carrier is given by equation (5-1):

Where:

• COST represents the economic cost to an agribusiness of service offering i within class j for carrier k, during the recurring period
• C&I is the LR fixed cost of CPE and installation over ijk
• RACC is the SR recurring cost for network access over ijk
• RTR is the SR recurring cost for network transport over ijk
• VACC is the VSR variable cost for access over ijk
• VTR is the VSR variable cost for transport over ijk
• OTHCOSTS are costs of available substitutes and complements
• TAX is a set of taxes and charges over ijk
• QOS is a vector of Quality Of Service characteristics over ijk.

Given the differences in VSR, SR, and LR and time compressed periodicity as technology progresses (2.5.2.5), the agribusiness hypercommunication outlays have many cost levels that would influence demand for a particular offering, class, or carrier. Each variable would affect costs differently. Furthermore, even if the variable units (those associated with VTR and VACC) could be derived into kbps the price per unit might not be easily convertible. Additionally, the units the associated with recurring charges (RTR or RACC) would still differ among alternatives. For this reason, rather than trying to separate prices and quantities, equation (5-1) may be thought of as a relationship among costs.

C&I refers to the capital cost of purchasing suitable CPE, software, employee training, and installation costs for the local infrastructure the agribusiness uses at its location(s) to communicate. It is an example of a long-run cost. Long run in this sense refers to the fact that equipment upgrades occur infrequently. Such long-run costs are often forgotten except when agribusinesses switch from one service to another.

Recurring access (RACC) and recurring transport (RTR) costs typically use monthly (recurring period) charges that are also influenced by length of run of contracts. Recurring charges are examples of SR non-traffic sensitive costs. However, by encouraging the use of long-term contracts, many service providers effectively make recurring charges more "long run" than CPE decisions. For example with dial-up Internet service, RACC is the monthly cost of the telephone line paid by the customer to an ILEC or ALEC, while RTR is the monthly cost of Internet service paid to an ISP or OSP.

Variable access (VACC) and variable transport (VTR) costs are per message, per call, per minute, or other traffic sensitive charges that may vary according to time of day, distance, or network congestion loads. These measures may be VSR for two reasons since they can vary day-to-day or hour to hour. With a single-line POTS line used for long-distance (LD) telephone calls, VTR would be the per-minute toll rate for use of the IXC LD network. VACC is the charge per-call or per-minute the IXC pays the originating and terminating LEC. In the case of telephony, VACC is usually embedded in VTR. The PICC was originally intended to replace the CCLR to compensate ILECs for local VACC incurred in LD calling.

However, in other cases VACC could be broken apart from VTR. One example is ISDN-BRI Internet service. The LEC that provides the ISDN line may charge a per-minute or per kbps rate (VACC) for traffic over the ISDN access line from the agribusiness to the ISP's POP in addition to RACC. Some ISPs charge per kbps charges (in addition to a monthly recurring fee, RTR). Even though such (VTR) charges are called Internet access charges, they are essentially transport charges because they cover costs incurred by the ISP for the ISP POP to Internet NAP (backbone) link.

Taxes (or regulatory charges) may be placed on any or all elements. Federal excise, Florida gross receipts, and most municipal taxes are ad valorem taxes assessed as a percentage of the total bill. These are rates on the total (RACC, RTR, VACC, and VTR) for service examples or classes that qualify for the tax. Other taxes or charges are based on the number of connections. For example, the PICC and SLC are assessed on a per-line basis. Special charges such as these may be collected only by certain carriers (non-rural-certified ILECs) on regulated service classes or specific examples. Later in 5.6.4, tax aspects of the conceptual model in (5-1) will be analyzed. The OTHCOSTS variable, which represents the set of available substitutes and complements, is important because the choice set is governed by what is available in a given location.

QOS (Quality of Service) variables include factors such as reliability, delay, capacity, and speed that affect the quality of hypercommunications. These are normally economic costs, but not accounting costs except when pricing depends on SLAs (Service Level Agreements). Table 4-3 in 4.2.4 provided a list of fifteen QOS dimensions.

The conceptual model helps agribusinesses realize that hypercommunications is governed by numerous network interrelationships. C&I is a proxy for the local network investment made by the agribusiness, RACC and VACC capture costs at the access level, and RTR and VTR represent the transport level. Each level is subject to different regulatory restrictions. The general QOS reference model shown in Figure 4-14 (in 4.2.3) may be consulted for review. A more specific version of the conceptual model will be used in Chapter 7 to aid agribusinesses in understanding the specific economic costs associated with service examples or classes and each carrier. There, the conceptual model can be extended to each level as well over class, form, and provider.

This remainder of 5.6 will use the general conceptual model just introduced to consider three regulatory issues. For each issue, the traditional regulatory approach is presented along with an analysis of how convergence changes the economics of that issue. It is likely that the structure of hypercommunication markets throughout Florida will be affected for some time by regulatory lag or the tendency for regulation to lag well behind technological change.

5.6.2 Rate Rebalancing and Universal Access

The first issue concerns the idea of rate rebalancing and universal access. Rate rebalancing is concerned with aligning prices and production costs more closely. When rate averaging has been followed as a policy (such as geographical averaging of rates for universal service), rate rebalancing seeks to make the average rate more economically efficient, often as a preface to the complete deregulation of rates. According to Sappington and Weisman, "if competition is encouraged in the telecommunications industry . . . rate rebalancing can limit cream skimming, and can help ensure that services are purchased from the least cost supplier" [Sappington and Weisman, 1996, p. 119]. Traditional approaches have focused on universal telephone service, seeking to balance access and transport prices with access and transport costs, and on balancing urban and rural rates with appropriate costs.

Figure 5-13 illustrates one aspect of rate rebalancing, the differences between the geographic averaged rate and marginal cost of providing telephone service as population densities fall. On the left panel, the example shows the monthly rate for telephone service on the y-axis, along with the amount per line that would be need to bring rates into equity at the averaged rate. Subsidizers pay a subsidy equal to the difference between the cost-based price and the average rate, while subsidized customers receive a similar amount.

When rates are rebalanced (as shown in the right-hand panel), subsidizers gain an area equal to that lost by subsidized customers. While a subsidy still occurs, it is smaller (more balanced) than under the initial average rate. Figure 5-12 also depicted rate averaging (in price-quantity space) during the discussion of universal service mechanisms in 5.5.1. Of course, one problem (and the reason rates may need rebalancing) is how the average rate line is arrived at. However, when convergence is combined with rate rebalancing it becomes especially difficult to see who should be subsidized and who should be the subsidizer. Beyond the obvious issue of locating the rebalanced rate line, there are many other issues regarding rate rebalancing (even before convergence) that cannot be covered here.

However, several pertinent points need at least to be listed before the complications caused by convergence to rate rebalancing are considered. First, as was discussed in 5.4.1, the unit of analysis matters. The rates to rebalance can be calculated on several bases. For example, local telephone service can use CCLs or ALEs while long-distance can use minutes, call-minute-miles, or total bill. New technologies cause problems when units of analysis are improperly combined. Unit differences between technologies can be equalized using derivations such as ALE-CCL ratios. However, a single common carrier line and a set of virtual ALEs carried on a single trunk have non-linear differences in costs and prices that a ratio alone cannot control for. Adding a further level of complexity, derived enhanced telecommunication channels can be used for local and long-distance telephony as well as for data networking and Internet access. In the case of "customer reconfiguration controlled" smart T's (discussed in 4.7.4), deriving channels is made harder since the circuit automatically makes virtual adjustments among voice, data, and Internet channels based on customer traffic patterns. Not only are derived lines capable of different uses, but they are capable of dynamically configuring themselves to those different uses, making comparison with a POTS line impossible.

Convergence changes analysis of rate rebalancing in two chief ways. First, under convergence, universal access to hypercommunication networks rather than universal telephone service becomes the objective of rate rebalancing. Second, the unit of analysis, therefore, shifts from local telephone access lines to some kind of kbps per month (VACC) or bandwidth/capacity measure (RACC). Figure 5-14 illustrates this through a hypothetical situation with four carriers who provide a hypercommunications service (rather than telephone service) with the price line now showing monthly price per kbps. The location of the average rate line (not shown, but left to the imagination) would affect each carrier differently. An important question to be kept in mind is whether recurring or variable access charges (or both) need rebalancing.

The four carriers are a cableco, an ALEC, an ILEC, and a wireless provider. The cableco and the ILEC are COLRs so that their costs represent at the very least the fact that their infrastructure (if not their operating costs) benefit from a combination of size and possible cross-subsidies. The ALEC exhibits classic skimming behavior. As a smaller size firm without the burden of a COLR obligation, the ALEC targets the ILEC's most profitable customers. Since it can resell ILEC access facilities (such as the local loop) at regulated interconnection rates, the ALEC can skim away the cream of the most profitable customers from the ILEC. The wireless firm is smaller still. Its costs of serving a dense area are dependent on technical issues such as those mentioned in Chapter 4. The wireless firm in Figure 5-14 is able to serve the sparsest area at the lowest cost to customers.

Each carrier has a density level (shown on the x-axis as persons per square mile). The cableco can serve customers more cheaply in the densest area (over 5000 persons per square mile). The ALEC is able to serve a density just above the medium density more efficiently since the cableco costs begin to rise at that point. Next, the ILEC is the low cost provider until the low-density zone is reached where the wireless carrier remains the most efficient provider as densities fall.

It could certainly be argued that rebalancing of averaged rates (under the pre-convergence regulatory structure) is no longer necessary since market convergence has removed the worst of the inequity between rural and urban. However, the effect of removing an averaged rate overnight could influence competitors differently. COLRs in particular might be disadvantaged. In absolute terms, the urban-rural differences appear smaller than the hypothetical case shown in Figure 5-13. Remember though that the units have changed. Furthermore, if the public infrastructure investment rationale (5.4.2) applies, the fact that the densest location is some six to seven times more cheaply served than the sparsest may support the case for continued rate averaging. However, if the cableco and the ILEC are both subsidized and are still enjoying the fruits of past supernormal profits due to regulatory inefficiencies, an anti-competitive bias may be present.

Yet even in the case of a single firm, rebalancing rates is not an easy exercise. In addition to trying to prevent the regulated firm from losing money or making supernormal profits, the effect of the rebalance on each density group must be considered. With four carriers, four density groups, and four costs, the effect of rebalancing one average rate (much less establishing that rate to begin with) would be an extremely difficult task that could easily result in new inefficiencies.

Rate rebalancing has some interesting dynamics if the ALEC buys the wireless carrier and the cableco buys the ALEC in response as shown in Figure 5-15. Here, the ILEC/cableco is the low-cost provider in the highest density area and in the medium to low zone. The ALEC/wireless is low-cost at middle-high densities and at low to very low densities. Again, placement of the average rate line would be important, but there is no clear-cut economic level where it "should" be placed. Interestingly, had the hypothetical acquisition been different, one carrier could be placed in the position of always subsidizing the other carrier. The effect of such mergers on price even in rural areas can be positive. For instance, Levin and Meisel found that when rural telcos buy rural cablecos monthly prices fall by eight percent on average [Levin and Meisel, 1993].

Rate rebalancing is also concerned with the differences among residential, business, and SOHO (Small Office Home Office) customers. Business rates may be two to three times higher than residential rates, even though it may be less costly to serve most business subscribers. Through such a demand segmented (price discrimination) approach a subsidy of residential service by businesses occurs because more inelastic business demand will bear a higher price. However, SOHO users are able to circumvent the division between business and residential markets. It is becoming difficult to identify business users from residential users since SOHOs and telecommuters can easily order circuits at residential rates. In both cases, costs may be cut in half; telecommuters would be reimbursed by their delighted firms while garage start-up SOHOs would save on entry costs and receive tax deductions.

Rate rebalancing is also concerned with the economics of balancing costs and rates between local access and long distance telephone transport. According to Sappington and Weisman,

rather than distort access prices to preserve artificially high intraLATA toll prices and artificially low basic local service rates in the presence of competition, it may be advisable to reduce regulated prices. When toll prices and basic local service rates better reflect marginal production costs, access prices can also be moved toward marginal production cost without encouraging excessive cream skimming. Undesirable bypass can also be limited by such a rebalancing. [Sappington and Weisman, 1996, p. 248]

A wider focus could be to expand rebalancing between access and transport among classes in Table 5-15 within the conceptual model, rather than the status quo approach of focusing on two examples (LD and local telephony) in a single class. Unless status quo universal service contributions are expanded to include data networking transport and Internet access and transport, rate averaging mechanisms cannot meet the "universal access" 200 kbps standard of high-speed network access defined by the FCC [FCC CCB, FCC 00-290, 2000]. If this is true, the luxury of rebalancing these rates may not be available to regulators.

Even if universal access in rural areas could be supported by universal service mechanisms, the policy will mean that urban telephone customers will pay for rural business and consumer access to non-PSTN hypercommunication networks. That would likely be a case where rates become more unbalanced than rebalanced. As convergence continues, rate averages may become more out of balance rather than rebalanced, unless new mechanisms are enacted. However, ISPs and data networking providers and their customers have fought extremely hard to see that no such result occurs. For these reasons, existing subsidies along with regulatory alterations may be necessary if universal access is to be achieved. As the next section shows, policy may not be able to implement universal access without deepening existing regulatory asymmetries and inefficiencies.

5.6.3 Convergence and the Application of Incentive and Traditional Utility Regulation to Hypercommunications

The next topic relates to rate rebalancing but is much broader. The literature on utility regulation is immense. All this sub-section can do is briefly cover four main issues involving agribusiness hypercommunications as shown in the conceptual model. More expansive treatments are found elsewhere on topics such as the fundamental postulates of telecommunications demand under regulation [Taylor, 1994] or the principles of utility regulation [Berg and Tschirhant, 1988].

Agribusinesses have two chief reasons to be interested. First, expansion of access to hypercommunication networks in rural areas is one way regulation influences Florida agribusinesses. On this first issue, concern focuses on the mere availability of classes of services. A second issue deals with the costs agribusinesses pay, service quality, and market competitiveness in general. On this second issue, concerns include pricing and the use of deregulation, regulation, and re-regulation to prevent market failures or negative network externalities. The objective of traditional and incentive utility regulation is to deal with concerns over pricing, access, and competitiveness as they influence rural areas and agribusiness constituencies, as well as urban businesses and consumers.

The first topic is to sketch out traditional regulatory theory and incentive regulation and apply them to a convergence scenario for agribusinesses. Traditional regulatory theory focuses on monopoly behavior under several regulatory regimes. Importantly, deregulation and technical change are erasing monopoly structures so that regulatory theory will need to apply in imperfectly competitive markets instead. According to Gong and Srinagesh, the economic model necessary to support the NII (National Information Infrastructure) would have "oligopolistic competition among a few large companies that invest in the underlying physical communications infrastructure" with "network and call externalities at the virtual network level". Additionally, there would be "large sunk costs and excess capacity in underlying transmission links" [Gong and Srinagesh, 1997, p. 66].

The first regime, strict rate-of-return regulation is depicted in Figure 5-16 [Adapted from Sappington and Weisman, 1996, p. 109]. Rate-of-return regulation is still used by the FPSC for several rural-certified ILECs in the state. The FPSC defines the rate-of-return method as:

the Commission determines the amount of revenue a firm needs in order to provide services. This determination involves establishing the appropriate rate of return and allowable rate base and expenses for the firm. Once the company's revenue requirement has been established, rates are set to produce that level of revenue. This process constrains the company's ability to act, or react, quickly to competitive changes. In order for the company to change rates either up or down, it must come to the Commission for approval to do so. [FPSC, "Understanding the Local Telecommunications Competitive Environment", 1999, p. 1]

The first step involves setting a TEL (Target Earnings Level) for each regulated firm. Rate of return regulation may be used to set indirectly any of the prices of the conceptual model (RACC, RTR, VACC, and VTR). Importantly, the effect on specific service offerings, among service classes, or on particular providers is uneven. Furthermore, unless every product of a particular provider is regulated, rate-of-return regulation requires that within-provider costs be "allocated" to particular services even for large, multi-product firms. Hence, it can be difficult to control prices through earnings for a multi-product carrier that has a mix of regulated and unregulated services.

Under strict rate-of-return regulation, the TEL is set by regulators with the idea that it represents a "fair" economic return. Information about what constitutes a fair return is necessarily derived from cost and accounting data from the regulated firm itself, presenting an important problem. The regulated firm has more information about these costs that the regulators. For this reason (and due to the tendency of the method to allow inefficiencies to become fixed) strict rate-of-return regulation has become much less used.

Another form of rate-of-return regulation is also shown in Figure 5-16. Banded rate-of-return regulation permits the regulated firm to operate within rate-of-return bands. TEL+d₁ becomes the upper band while TEL-d₁is the lower band. Hence, the band size is 2d₁. Under this policy, rather than attempt to recalibrate prices frequently to deal with inevitable variation in earnings, the firm is permitted to allow earnings to vary within a range set by the regulating agency. Banded rate-of-return regulation allows more flexibility, but suffers from the same problems the strict policy does. When costs are rising, banded rate-of-return regulation gives regulated firms a way to absorb higher costs through price increases [Ai and Sappington, 1998].

Traditional regulatory theory does not end with rate-of-return regulation, though rate-of-return still dominates regulatory practice for rural-certified ILECs. Accounting methods such as fully distributed costs and direct costing are used along with economic methods such as incremental cost to map the regulated firm's cost to a TEL or to a regulated price. Costing methods may be used outside rate-of-return schemes to set prices directly. However, convergence may doom cost-based pricing as Jamison concludes:

The difficulties of cost-based ratemaking have increased exponentially in recent years. This makes regulators' lives difficult enough by itself, but it is particularly troubling when one realizes that we were unable to resolve cost measurement issues even in the simpler world. [Jamison, 2000, p. 25]

Cost-based pricing schemes or certain demand-based schemes such as Ramsey pricing allowed inefficiencies and abuse of the regulatory process under pre-convergence regulation [Sheehan, 1991]. These effects would be magnified in the conceptual model of hypercommunications presented earlier.

Another set of regulatory devices is known as incentive regulation. Incentive regulation is defined by Sappington and Weisman as follows:

Until relatively recently, rate-of-return regulation was the predominant form of regulation in the telecommunications industry. Today, alternatives to rate-of-return regulation are commonplace. These alternatives take on many forms and are generally referred to as incentive regulation. . . . Incentive regulation can be defined as the implementation of rules that encourage a regulated firm to achieve desired goals by granting some, but not complete, discretion to the firm. [Sappington and Weisman, 1996, pp. 1-2]

The first form of incentive regulation is Earnings Share Regulation (ESR), a variation on rate-of-return regulation (also shown in Figure 5-16). According to Ai and Sappington,

ESR requires the regulated firm to share realized earnings with its customers according to a specified schedule. The typical schedule specifies an intermediate range of earnings in which no sharing occurs, so the regulated firm's actual earnings vary dollar for dollar with its financial performance in the marketplace. [Ai and Sappington, 1998, p. 5]

An earnings share factor, d₂, is used to construct an earnings share region as shown in Figure 5-16. Sharing occurs from TEL + d₁ to TEL + d₁ + d₂ using

a fraction (often half) of incremental earnings above (and often below) this intermediate range are shared with consumers. Once earnings reach this bound, the firm does not benefit financially from improved performance in the market place, much like RORR (Rate-of-Return Regulation). [Ai and Sappington, 1998, p. 5]

All profit above TEL + d₁ + d₂ is given to ratepayers under the ESR plan shown in Figure 5-16 [Sappington and Weisman, 1996, p. 111]. However, if profits fall below TEL-d₁-d₂, then rates are revised to allow the firm to earn TEL + d₁ + 0.5d₂or whatever the revenueshare amount would be. Sappington and Weisman say it is rarely a 50/50 split.

Another form of incentive regulation is price-cap regulation. Price-cap schemes are used to price telephone access in Florida for non-rural-certified price-cap ILECs such as BellSouth, GTE, and Sprint. Braeutigam notes that price-cap regulation removes any effort to map accounting or economic cost determinations to prices [Braeutigam, 1997]. Price-caps are meant as a more workable and less inefficient method of achieving a zero economic profit result.

Before considering how price-cap regulation may fare under the conceptual model of convergence in the hypercommunications marketplace, an outline of the status quo's ILEC price-cap regime is in order. According to the FCC:

Under price-cap regulation, rates charged by incumbent local exchange carriers are governed by caps on rates instead of traditional rate-of-return regulation. Price-cap regulation creates incentives for LECs to operate more efficiently and to introduce innovative new services. [FCC CCB, Competitive Pricing Division, 1999]

Price-cap regulation caps prices using a rate that creates falling prices over time on regulated services. Since firms cannot control the rate of price-cap decline through production costs or earnings, the regulated firm has an incentive to reduce operating costs.

Conceptually, the price-cap is set so that the regulated firm will get zero economic profit. This occurs if "the rate at which its output prices rise on average, is restricted to equal the difference between: (1) the rate at which the firm's input prices rise; and (2) the rate at which the firm's productivity increases" [Bernstein and Sappington, 1999, p. 12]. Symbolically this is expressed as CAP = RPI - X, where RPI is an input price index and X represents the x-factor, an average rate of price decline with inflation controlled.

Alternatively, the price-cap on output prices in the regulated industry, can be written as: (5-2).

Here, output prices in the regulated sector are allowed to increase at the rate of output price inflation in the entire economy less an offset, the basic x-factor . The cap is composed of multiple goods through a quantity-weighted index that changes yearly.

However, as Foreman discusses, the weights on items a basket of services under a common price-cap can be manipulated by regulated firms. Figure 5-17 (adapted from Foreman, 1995) shows how this can occur.

Suppose that there are two services, each with the weight 0.5 in the first period and each with an identical demand curve as shown. In the first period, the regulated firm can set each price equal to the cap (exactly) and satisfy the regulation. However, if the firm wanted to raise the price of one service, it would have to lower the price of the other service by an equivalent amount, unless the demand for one service was perfectly inelastic. By increasing and decreasing prices, the firm would lower its profit for the first period compared with the equal price case.

If the price equals p₀ for both products, then total revenue would be 2(BCDB'). Note that if one price is increased to p₁, then the other must fall to p₂. The first service would gain A but lose areas C and B'. The price of the second service would result in a loss of ABC and a gain of A'. Since A equals B and A' equals B', the net loss overall for the firm is 2C. Over time, however, such behavior would become profitable because the weights for the next period would change with prices since they are revenue-based. With a new, lower rate, the regulated carrier would be able to raise profits but stay under the cap. The more inelastic demand is the greater that a firm could use revenue weight manipulation. Quantity weighting would defeat this incentive to cheat [Foreman, 1995].

Another consideration is that for multi-product firms, joint products and non-allocable factors make it impossible to measure RPI or the x-factor for specific products. Thus, adjustment of the x-factor can be sued as a proxy. The adjusted x-factor also corrects for limited regulatory control of multi-product firms. If only some services produced by the firm were regulated, while others were not, then (5-2) would become (5-3).

In (5-3) , while represents an adjustment for a limited span of regulation [Bernstein and Sappington, 1999, p. 19, equation 3.6]. Hence, the cap is flexible enough that the constraint on capped price increases can be relaxed if uncapped service price growth is below the growth rate of all input prices and the total factor productivity growth rate of the firm. Similarly, the growth rate of capped services would need tightening if the reverse were true [Bernstein and Sappington, 1999, p. 21].

There are several disadvantages to price-cap regulation. First, there is the risk that regulated firms will suffer from "excessive or unduly meager earnings" [Bernstein and Sappington, 1999, p. 7]. Furthermore, due to regulatory lag and other factors, the choice of an appropriate X-factor can easily miss the socially optimal figure. In addition to the possibility that firms will strategically change revenue weights, there are possible inefficiencies with multi-product firms with only some products regulated. Nevertheless, price-cap regulation is currently an important part of the regulatory arsenal since it seems to be socially preferable to other methods. Overall, incentive regulations have been found to have effects that are more positive on the regulated market and infrastructure development than rate-of-return regulation [Ai and Sappington, 1998].

A third topic is the implication of deregulation on the availability of high-speed access to hypercommunications. Here, several new pricing regulations on interconnection have been designed to increase competition and improve availability and choice. There are two chief objectives: first, to consider how regulation, deregulation, and re-regulation may be used to expand the rural infrastructure; second, to consider how market structure and pricing may be altered as regulation and convergence interact.

Under convergence, this strain of regulatory theory may impact agribusiness hypercommunications by changing interconnection pricing, allowing competitors to use ILEC facilities to provide a range of services over the ILEC local loop or cable TV infrastructure. The ILECs are generally required to allow open interconnection, even with non-telephony services such as DSL. However, the cablecos do not have to provide open access to competing providers over their facilities [May, 1999].

The ECPR (Efficient Component Pricing Rule) is one example of how interconnection is treated by regulators. The ECPR encourages new entrants that want to access an incumbent's network to serve a limited area. Such fringe competitors will not enter unless they are able to profit from wholesale rates that have been set at such a level so that the incumbent cannot bar the new entrant. The ECPR yields an interconnection price paid by the entrant (such as an ALEC) to interconnect with the ILEC's network under the following relationship. The ECPR can be written as: (5-4).

Where:

• IP is the interconnection price
• RP is the retail price
• RC is the retail cost
• WC is the wholesale cost.

The idea is to pick "an access price equal to the difference between the telephone operator's price and marginal cost on the competitive segment" [Laffont and Tirole, 1996]. The IP would correspond to a fraction of VACC and RACC that most closely matched specific marginal costs. In this way, competition would be fostered.

Other related interconnection issues include the difference between resale and unbundling, traffic mix, traffic exchange, and other barriers. The difference between resale and unbundling will become clearer when bundling is taken up in Chapter 6. Essentially, this concerns how prices for providers who use an incumbent network (ILEC, cableco, or electrico) are structured given that some alternative providers have their own transport (or intermediate access) networks while others are purely resellers. The ECPR or another method is used to help set prices on UNE (Unbundled Network Elements). A UNE can be a hardware device or a virtual element of a more complicated composite service.

Within the conceptual model (5-1), the IP (determined through the straight ECPR method) becomes a fraction of the retail costs such as RACC, RTR, VACC, and VTR. With UNE, these prices cannot easily be compared among providers. At the wholesale level, the cost behind each applicable service is broken into as many as several hundred sub-components and priced individually. At retail, the resulting prices may appear comparable, but particular options on a service or class vary. For example, many ALECs use UNE to include so-called custom calling services that require AIN (such as caller ID, call waiting, etc.) in RACC, rather than pricing them as separate services or charging a separate VACC per-use [Jamison, 1999].

Traffic mix rules concern how to recover the underlying costs for a particular type of traffic from a converged network. For example, Chakravorti, Sharkey, and Srinagesh suggest that if each asynchronous ATM cell on a converged network were naively priced (using equivalent regulated voice rates), an arbitrage opportunity would exist. A local telephone call might be $0.01 per minute, but a two-hour video movie (at 45 Mbps) would be priced at $843.75. Even with compression and a T-1 speed (1.544 Mbps), the cost of the movie (again using regulated telephone rates) would be $30.

If the true costs of access and transport in computer networks are used instead, the results are dramatically different. The regulated telephony rate charges a VACC (with VTR embedded) capable of recovering all costs for a circuit-switched connection over the PSTN. Inherently included are the combinatorial, probabilistic, and variation technical characteristics of the telephone network (as explained in 3.3). However, through traffic mix arbitrage (where the market technologies internalize VTR) all costs could be recovered with a fixed price (RACC) of approximately $40 per month [Chakravorti, Sharkey, and Srinagesh, 1995, p. 83]. Arbitrage would be possible because computer networks have more flexible combinatorial, variational, and probabilistic properties than the regulated PSTN. This point was made repeatedly in Chapter 3 beginning in 3.4.1.

Hence, regulatory prices must be competitive across classes of the conceptual model.

On a per packet basis, a video packet will likely be heavily discounted relative to a voice packet. Voice signals might then be aggregated and packaged so as to resemble a video transmission, assuming that detection by traffic-distinguishing methods is imperfect. As a result, video channels could be used to carry voice traffic at a price that is significantly lower than that being charged by the local operating company for bulk transport of voice. [Chakravorti, Sharkey, and Srinagesh, 1995, p. 100]

Such inconsistencies can create opportunities for agribusinesses (with access to hypercommunication networks) to buy more communications (measured by capacity and bits) than they otherwise would be able to if access was restricted to PSTN service offerings.

Fourth, and finally, there are concerns about re-regulation necessitated by QOS issues, regulatory jurisdictions, and negative network externalities. While the FPSC does place standards on some regulated services (for example, a POTS line must be capable of sustaining modem speeds of 9.6 kbps), most QOS issues are left for the market to solve. Hence, there may be no recourse for agribusinesses that purchase new voice circuits from ALECs and then find they cannot receive or make telephone calls for several weeks because of a carrier software or hardware problem. The only recourse may be the recovery of charges for the period of time the circuit is down. Without a SLA from the carrier, they may be no guarantee of reliability, speed, or a host of other QOS dimensions. It is conceivable that as the regulatory mission broadens from universal service to universal access, specific performance-based rules (currently enforced on telephones and electric utilities) would be required on wider service classes. However, regulatory authorities are in no hurry to expand their missions.

In Florida, complaints concerning non-regulated hypercommunication services do not go to the FPSC, but to the Department of Agriculture and Consumer Services (DACS). It can be confusing for agribusinesses to understand where to turn if a serious problem arises. In some cases, there is no remedy other than court. For example, customers of the now-defunct Iridium project had no agency to turn to (federal, state, or local) when the carrier went bankrupt. Some customers were saddled with expensive CPE that was incompatible with other providers and lost pre-payments for up to a year of service. The untaxed, unregulated Internet offers many opportunities but also danger. Unlike the PSTN (where prank calls can be traced), there may be no way to catch the cracker that steals company financial data or the vandal that destroys a website.

5.6.4 Asymmetric Taxes and Regulatory Charges

Specifics of the third issue, direct taxation, were discussed in 5.5.2. However, it makes sense to cover the topic briefly in a more general way. By doing so, the tendency of taxation to alter competition among classes and among services in the conceptual model (5-1) can be reduced into a more apparent form. There are four general forms of taxes and regulatory charges. Table 5-16 shows each type, a general formula, and gives an example of each.

First, ad valorem taxes are assessed on a percentage basis of the taxable total of RACC, TRT, VACC, and VTR for a particular service. The tax rate, r is multiplied by the taxable prices V or R on quantities (Q or U). Ad valorem taxes or fees differ in competitive impact only when particular classes (such as Internet access) are excluded entirely, or when they do not appear in a particular cost such as to RACC, RTR, VACC, or VTR.

Fixed fees or taxes (F) are usually regulatory charges that apply to the units (U) on which recurring access charges (RACC and RTR) are based. Typically, either the units applicable to RACC (such as access lines) or those applicable to RTR (such as pre-subscribed LD lines) are used. In addition to the competitive effect of that distinction, the U or F among classes may differ. For example, the F is different between the single-line PICC cap of $4.35 per line and the multi-line cap of $9.20 per line. U's differ also, as when 23 or 24 derived lines on a T are one unit that is assessed 5F, while two lines are assessed 2F. Recall that charges or taxes may apply only to certain carriers such as ILECs, to certain customer classes, or to particular service offerings only. Hence, they can have an asymmetric impact on different firms as well.

The third example is the tax on a tax (or on a regulatory charge). The federal excise tax, for example, includes the USF contribution, SLC, PICC, and some local charges such as right-of-way and 911 charges. This compound taxation magnifies the effect of such taxes. Taxes on taxes and fees are applied most often to fixed regulatory charges that are transfers from customer to firm under the rationale that these fees are no different from recurring charges.

The last form is the indirect tax or regulatory charge. Indirect taxes and fees tend to differ according to how services are defined within a class or among classes. For certain special access connections for private data networking, federal SLC or PICC charges are incorrectly collected through modifications of the x-factor on local loop-based services that have no ALE or minute conversion unit. Customers may not realize that the price of the underlying service (always RACC or RTR) is slightly higher than it would otherwise be so that ILECs can indirectly collect an access subsidy.

The incidence of taxes and fees, while important, can only be touched on here. In general, the more inelastic the demand, the more the incidence of a tax shifts to buyers. Similarly, the more elastic the supply, the more the incidence is on the buyer. Taken together, the tendency for demand to be relatively inelastic on local access [Taylor, 1994]. and for supply to be elastic (by the cost arguments in Chapter 2), results in the forward shifting of many taxes and fees to customers. Of course, there can be variation among the four forms of taxes and fees. The indirect tax might seem to be easier to pass on to customers (or be backward-shifted to suppliers or employees of the carrier) than a direct tax or charge. However, the post-1996 TCA custom of adding fixed regulatory charges as line items on customer bills (rather than embedding them in prices) suggests it may be easier to pass along an explicit tax to customers rather than an indirect one.

An issue related to asymmetric taxation is the broader concern over asymmetric regulation. It is often argued that asymmetric regulation and taxation produces market inefficiencies on agribusiness demanders and hypercommunications suppliers. However, before presenting that argument, it is important to take the advice of Sappington and Weisman. One of the ten myths of incentive regulation is that "Absent direct evidence of harm caused by asymmetric regulation (that is treating different competitors differently), it is safe to assume that no harm has occurred" [Sappington and Weisman, 1996, p. 333]. This warning rings similar to the tendency for some to automatically equate provider size, mergers, or other changes in market structure with automatic losses in social benefits. As Weaver states, "the presence of bigness in a case at hand will convince many lawyers that they will find 'something bad' about the situation if only they look long and hard enough" [Weaver, 1977, p. 170]. Again, absent evidence of collusion, predatory pricing, or other anti-competitive actions, decreases in firm numbers or other structural changes do not automatically transfer into socially inefficient conduct.

However, it is clearly theoretically possible for asymmetric regulation to create negative impacts on efficiency. Again, possible impacts can be understood through within-class, among-class, and bundle-by-bundle analyses of the conceptual model. Even when an asymmetric barrier is withdrawn, the pre-barrier service structure tends to linger, prohibiting reliability and regional availability.

Importantly, technology is changing at a faster rate than regulation and policy can. This is leading to a new communications market for rural and urban areas as Bernard Courtois writes in 1996:

Throughout the world, there seems to be an accelerating trend in both developed and less-developed nations toward privatization and more competition. As well, there are numerous consolidations and alliances between players and across national borders and across industry borders. The whole notion of traditional lines of business, where you had various players in their respective little cubicles, is disappearing through these changes. [Courtois, 1996, p. 189]

Importantly, regulatory policies governing hypercommunications need to recognize that the important markets for agribusinesses are communication and information, not transport and access. As if to recognize this, Jamison notes that

telecommunications is becoming a commodity to be used by information services coming from the publishing, computer, and broadcasting industries. Trying to control information markets by controlling telecommunications markets may result in telecommunications playing only a minor role." [Jamison, 1995, p. 516]

The reasoning expressed above was ahead of its time and is more applicable now for agribusinesses than ever. However, as access and transport become more like commodities, it is hard to see where Florida agribusinesses are left. Regulation appears to have several important effects.

First, agribusinesses have to have competitive high-speed access to hypercommunication networks in order to profit from information and communication. Regulation may help bring high-speed access to hypercommunications networks in rural parts of Florida faster than would otherwise occur. However, far from guaranteeing the universal access standard of the 1996 TCA, the FPSC can only guarantee 56 kbps universal service within several years [FPSC, 2000]. The federal standard of 200 kbps is not even contemplated yet as a possibility by the FPSC.

Hence, the second benefit of hypercommunication regulations to businesses is a perverse one, the arbitrage opportunity arising from using derived channels from competing providers instead of PSTN or ILEC-based carriers. Again, access is needed to obtain this advantage and may not occur for years in the most sparsely populated parts of Florida. Thus, poorly located agribusinesses may become locked into the telephone and enhanced telecommunications network where bandwidth and bps are not yet commodities. However, many mitigating factors such as market competition and the fact that fully converged technologies cannot compete on an equal QOS footing may lessen negative impact.

5.7 Communication and Rural Development Measures Targeted to Agribusinesses

The discussion of how hypercommunications policy affects agribusiness would not be complete without mentioning how rural development policies help rural agribusinesses reach infrastructure development goals. Several state and local rural development programs can directly or indirectly help agribusinesses with hypercommunications. While not a major focus of agricultural policy, the development of rural infrastructure and special programs designed to help agribusinesses with their hypercommunication needs have received some attention.

Some federal programs are designed to support infrastructure development indirectly in rural areas, as part of agricultural policy through the Farm Bill. Some are demonstration programs or other small-scale efforts. Additionally, there are modest state efforts. Some of these initiatives are listed in Table 5-17.

Rural infrastructure development is one of four basic approaches to rural development policy. The rationale for rural infrastructure development was given in 5.4.2. However, federal rural development policy lacks cohesion and is especially splintered in communications infrastructure development (outside of RUS programs for rural-certified ILECs) [Fisher, Harris, Fletcher, and Brown, 1998].

AgVenture Services is a program of Florida's Department of Agriculture and Consumer Services (DACS). AgVenture is designed to promote the use of commodities from Florida agriculture in value-added products, while increasing agribusiness profitability within Florida. While mainly a marketing and promotional program, AgVenture services can help new agribusinesses with facilities and equipment (including hypercommunication-related products) as well as with developing business plans, competitive intelligence, and (in some cases) offers to share the cost of capital improvements.

Crossroads Florida is the state's rural economic development and workforce program. Crossroads Florida works in concert with Enterprise Florida, Inc. a public-private partnership responsible for overall economic development in Florida. Enterprise Florida can assist larger businesses with specialized infrastructure needs as part of a business relocation package. Crossroads Florida does not offer specific assistance for infrastructure development.

The RUS is the successor agency to the REA (Rural Electrification Administration) and is acts as a bank and technical resource for rural electric power co-operatives and rural-certified ILECs. Both kinds of utilities are assisted by the RUS with building rural communications infrastructure. Through loans from the RTB (Rural Telephone Bank), the RUS has helped finance ambitious infrastructure development efforts for rural carriers that choose to do so.

Importantly, the need for regulatory help in extending high-speed hypercommunications to rural areas is not universally agreed upon by ILECs. According to the FPSC, the two ILEC service areas in the state with the greatest percentage of loops capable of supporting DSL are Northeast Florida Telephone (100% of DSL-capable loops) and TDS Telecom (formerly Quincy Telephone, 67% DSL-capable) [FPSC, 1999]. The FPSC concludes that no state action is needed to improve outside plants of small rural-certified ILECs since the federal RUS modernization program requirements are taking care of the problem already.

However, such a conclusion ignores the fact that non-rural-certified carriers are hardly rushing to extend high-speed services to the rural areas they serve. Some rural-certified carriers are meeting the federal 200 kbps standard [FCC CCB, FCC 00-290, 2000] of extending advanced telecommunications services, but others have no DSL program. Furthermore, all of Florida's rural-certified ILECs combined serve only 193 thousand customers over a service area that comprises 14 percent of Florida's land area. More importantly, non-rural-certified carriers BellSouth, GTE, and Sprint serve at least ten times as many rural customers (depending on the definition of rural, see Table 5-3) and their actions will have more impact.

If a rural-certified ILEC (or non-rural-certified ILEC) decides not to upgrade its plant and no competitor is able to do it, there is no federal program to encourage deployment of high-speed service other than USF support. However, currently USF support is targeted towards universal service (telephony) instead of universal access (hypercommunications).

Under current regulations, RUS can lend to service providers only to

meet the 'basic local exchange telephone service needs of rural areas', i.e., provide wireline voice service.This prevents RUS from lending to providers that want to offer, for example, advanced telecommunications services without offering voice. Given Administration and Congressional interest in promoting the availability of broadband, RUS is proposing regulatory reforms to change these policies. This would allow RUS to use more of its $670 million in lending authority for rural telecommunications to encourage private sector investment in rural broadband services. [USDA RUS-NTIA Joint Project Team, 2000, p. 43]

Other USDA agencies and programs concentrate on rural development. Some such as the SBIR (Small Business Innovation Research) makes grants to qualified small businesses to support high technology research and other qualified projects. Agribusinesses can obtain loans or grants from several other USDA rural development programs. These include B&I Direct Loans, B&I Guaranteed Loans, IRP (Intermediary Re-lending Program), RBEG (Rural Business Enterprise Grants), RBOG (Rural Business Opportunity Grants), REDL (Rural Economic Development Loans), and REDG (Rural Enterprise Development Grants). Money from these programs can generally be used to help defray business expenses including communications.

The specific programs in Table 5-17 are not targeted only to agribusinesses or to hypercommunications. However, they do represent sources of funds for small agribusinesses seeking to upgrade the communications infrastructure on their own premises. Loans and grants are often made for capital equipment such as CPE devices.

Chapter 5 has concentrated on the implications of regulation on agribusiness hypercommunications. The topic is an enormous one. The regulatory rationales are wide-ranging, the mechanisms complicated, and the traditional and incentive regulation are not necessarily able to handle convergence. However, three main conclusions can be reached.

First, access to hypercommunication networks is important for agribusinesses and the communities around them. Until universal access occurs, some agribusinesses and some rural areas of Florida will not be able to compete on the same terms with other areas. However, neither the FCC nor the FPSC are ready to expand universal telephone service to high-speed universal access. Of all agencies, the RUS has had the best record of accomplishment of improving service in rural areas. However, the population served by rural-certified ILECs in Florida is small. Furthermore, RUS results depend on the ILEC's initiative, and, even then, are limited to DSL.

In Florida, the most important infrastructure that can be used for access is probably that of non-rural-certified carriers. However, other less regulated providers such as cablecos, and fixed wireless providers are showing some initiative in small towns and rural areas in parts of the state. Technological developments and competition are driving convergence. New, better, and cheaper technologies are being developed to link agribusiness premises over the "last mile" in the local access network to the hypercommunications networks. Regulatory efforts alone are unlikely to result in universal access.

Ironically, where there are multiple forms of access, asymmetric regulatory and taxation produce perverse incentives for businesses to desert PSTN and enhanced telecommunication technologies and use special access, private data networking, and Internet technologies instead for converged hypercommunication needs. Yet, these new technologies are in their introductory stages so that there is a tradeoff between high-tech, low-priced approaches with questionable QOS and low-tech, high-priced approaches with better reliability.