VOIP, a tax cut for government

VOIP, a tax cut for government
Potential savings from federal, state, and local deployment

By Kenneth Brown, Tocqueville staff, and associates
October 25, 2004

Copyright © Alexis de Tocqueville Institution, all rights reserved

Tocqueville estimates federal, state, and local savings from VOIP deployment would equal $4.5 billion annually. This is a conservative estimate.

At these levels, government savings from VoIP could exceed the total amount of taxes governments collect from current levies on traditional telephone service by 100 percent or more, even after accounting for capital conversion costs.

AdTI studied data from more than 300 agencies, and performed or compiled more than a dozen case studies of agencies that converted to VOIP or similar technologies. These conversions yielded savings of between 20 percent and 60 percent, with most reported savings ranging between 30 percent and 40 percent.

State and local agencies were generally cooperative and make information about telecommunications expenditures relatively transparent; AdTI is grateful for their cooperation. The federal government is significantly less transparent.

Tocqueville recommends that all levels of government review their telecommunications expenditures in detail, as California and a number of other states are doing to determine how new technologies might save the taxpayers money.

1. Overview
2. Federal government savings from VoIP
3. Examples of state and local government savings from VoIP
4. Summary

1. Overview: Determining the government’s telephone bill
2. Empirical data
2.1 Eau Claire School District and City of Southfield
2.2 Data from other government projects
3. Methodology for estimating total bill
3.1 Employee estimates
3.2 Line usage
3.3 Averaging monthly line costs
4. Developing formulas for potential government estimates
5. Local, state, federal, and military estimates
6. Indirect telephone services costs to government
7. Summary – Annual savings estimates with VoIP

1. Overview: Tax Revenue from Telephone Usage
2. Accepted Total Usage Figure: $20 Billion
2.1. COST Study
2.2. PFF Methodology
2.3. Ernst and Young Methodology
3. Problems with Assigning $20 Billion in Losses to VoIP
3.1. Vague Definition of “Telecommunications” Taxes
3.2. Excess Taxes
3.3. Taxes vs. Fees
3.4. Business Taxes vs. Transaction Taxes (PFF# 50% issue)
3.5. Inconclusive Threshold
4. CBO VoIP Loss Estimates
4.1. Overview: CBO Only $3 Billion Annually
4.2. Contradicting Sales Tax Forecasts
5. Summary

1. Overview – Savings Estimates vs. Loss Estimates
2. Adjusting Estimates with Slower Migration To VoIP
3. The VoIP Tax Cut
4. VoIP as a Mandate to Change Local Sales Tax Policy
5. VoIP as a Refund to Taxpayers
6. VoIP Issue as a Template for Disturbance
7. Market Impact of Unregulated VoIP

VoIP detractors argue that the similarities between voice over Internet and voice over phone lines mean that both technologies should be taxed the same way.[2] Conversely, VoIP proponents argue that voice over Internet is an application, much like MP3’s, e-mail, instant messaging, etc., thus should not be treated (or regulated) differently.[3] The back and forth discussion over lost taxes has made VoIP taxation a hotly debated issue in Congress, the courtroom,the FCC, and within every regulatory body governing telephone service in the United States.

As quiet as it’s kept, however, an issue of equal significance intersects government and VoIP.

Government is a significant telephone services customer. Combined, U.S. federal, state, and local government agencies annually spend billions of dollars on telephone service. After closer study, it becomes apparent that government savings from VoIP have significant potential.

Before discussing the net loss/gain from VoIP, it is important to pre-empt a central cause of contention. VoIP and VoIP implementation can be defined at least a dozen different ways. For example, VoIP can be internal connections within a single office building or VoIP can be external connections, connecting offices on a campus or across the country.

At the consumer level, VoIP is commonly referred to as applications that move voice across the Internet with/without a phone number to other parties on the Internet. Services such as Skype and Kazaa provide free connections to users that wish to send voice directly to parties linked with the respective software, across the Internet.[4]

To add more confusion to the discussion, VoIP doesn’t necessarily have to mean voice over the Internet. VoIP is commonly described as Internet protocol or IP communications. Specifically, voice is converted into IP for many applications. Because technology enables voice conversion into IP protocol, with an IP network, voice can move office to office without the Internet.

Dwayne Goldsmith, CEO of Inflexion Communications elaborates on this point commenting, “VoIP, Voice over Internet Protocol does not necessarily refer to calls moving across the public Internet. Rather, VoIP often refers to voice being converted into an IP format, and then pushed across local area, wide area, private, or public networks. The underlying network architecture is what makes the difference. For example, a VoIP handset could be receiving Internet Protocol (IP) voice from a private network, analog voice converted to IP, or IP voice from the public internet. Unless you have all the details about the architecture, it is very difficult to tell what type of VoIP calls you are receiving.”

Thus, because there is such a wide range of definitions and configurations of VoIP, it is probable that any calculation of the total amount of VoIP the government is using (or could potentially use) could be argued as flawed. Today, even traditional telephone service providers either have VoIP backend networks[5] or even offer consumer and businesses VoIP services[6].

Each of these variations of VoIP make defining what is and isn’t VoIP extremely difficult. The approach of this paper is to focus the discussion on the pros and cons of the status quo; specifically, whether we should allow the marketplace to proceed, or alter the marketplace by implementing changes to accommodate new taxes. For the purposes of this paper, VoIP savings refers to any and all types of VoIP technology that can be employed by government for communications services and/or to reduce its total telephone bill.

Government has actively pursued VoIP implementation for its cost savings features for many years. Federal, state, and local IT and telecom managers are actively exploring ways VoIP can be used to replace older, more expensive communications infrastructure. Today, VoIP (in it’s diversely different formats) is either being researched or deployed in government offices across the country.

To date, almost every major state in the country including California, New York, and Texas has announced a VoIP initiative. However, the move to VoIP within the federal government appears slower and somewhat less reported. In addition, the discussion about cost savings from VoIP at federal agencies has been overshadowed by issues such as management, governance, privacy, security, and certification. Richard Grigonis of VON magazine wrote, “Indeed, what’s needed is standardization of the federal approval process for VoIP and other communications technologies. Right now it’s a mess…For the present, government acceptance remains a long and complicated process.”[8]

The U.S. Department of Commerce has been involved in one the most expansive VoIP rollouts in government, but has not widely publicized major cost savings from the implementation. Describing the project, Grigonis writes, “For many years, vendors eagerly awaited interest in VoIP by their largest potential customer–the US Government. Their patience is now paying off. The Dept. of Commerce, for example, has spent two years and $6.7 million to rejuvenate its communications infrastructure (consolidating 13 data networks into one Gigabit Ethernet network having a fiber-optic backbone), along with a $5 million investment in Cisco VoIP equipment serving 4000 users, thus eliminating and/or consolidating the previous 130 separate phone systems used in the vast, one million square foot J. Edgar Hoover Building.”

But today, government officials downplay cost-savings attributes from VoIP implementation. Karen Hogan, Deputy Chief Information Officer for the Department Commerce comments, “…the Commerce Department's implementation of VoIP was not done specifically to save money, but for other management reasons. The major sources of savings with this technology seem to be: (1) take advantage of a robust network infrastructure - which we have, but it was first widely used to support the VoIP implementation; and (2) long distance charges - our implementation is oriented to the main headquarters building for Commerce.[9]

Along with the Department of Commerce, various VoIP implementations at the federal level include Naval Sea Systems Command (NAVSEA),[10] the U.S. Census Bureau[11] (which is part of the Department of Commerce), and Central Command (Centcom), which oversees military operations in Afghanistan and Iraq.[12] Certification of VoIP equipment for military use demands rigorous technology testing and engineering. Recently, Cisco announced that it passed U.S. Department of Defense (DoD) Voice over Internet Protocol (VoIP) interoperability tests. This certification by DoD's Joint Interoperability Test Command (JITC) confirms that all aspects of Cisco's IP telephony solutions conform to the interoperability, reliability, resilience and security requirements of the DoD's multi-vendor voice network.[13]

In sum, the federal government is using VoIP, but not widely reporting its value as a cost savings tool. However, if VoIP were used throughout government, it would directly impact billions of dollars in expenditures on telephone service. Empirical evidence at the state and local level will help demonstrate this point more fully.

One of the best examples of how closely related VoIP and saving money has become with government is the news-making announcement in August by a California audit team. Concluding that implementing VoIP would save California government as much as $6.3 million a month, the report’s authors wrote:

“The convergence of voice, video and data communications on an IP network lowers the total cost of ownership and operation by enabling cost savings for long distance calls and by integrating the infrastructure and management operations. Replacing many different telephone systems with a single state-of-the-art system will improve efficiency and simplify system management and maintenance….The Governor should direct the California Inspector General to conduct an independent audit of all telecommunication services and respective billings received by the state. An appropriate measure of savings resulting from this one-time audit should be redirected to acquire and implement a Telecommunications Management System to control telecommunications costs and optimize the state’s network. The Department of General Services or its successor should initiate a pilot project to further test and confirm the advantages of VoIP within a statewide network infrastructure by January 1, 2005. It is anticipated that this project would take 90–180 days. Based on results of the pilot, the Department of General Services or its successor should coordinate the replacement of all appropriate departmental telephone systems with VoIP, using existing infrastructure components to the maximum extent possible beginning July 1, 2005.” [14]

A recent essay in Network World called “State IT Execs Betting on VoIP”, provided a number of reasons state IT and telecom managers are pursuing VoIP as a low cost alternative for infrastructure and communications in West Virginia, Texas, New Jersey, Minnesota, Alaska, California, and other states across the country.[15] The essay listed a number of reasons including cost savings as a major consideration for government managers choosing to deploy VoIP.

Within state and local government, applications for the cost-cutting technology are widespread and diverse. VoIP projects include custom conference calling, internet/voice mail capability, call center management, and trunk usage consolidation. Another principal use of VoIP technology has been to construct autonomous phone services centers. In many cases, managers are able to use the elasticity of VoIP platforms to become local phone service operators, eliminating much of their dependence on the local telephone service provider. Managers are able to use VoIP systems to create and assign new phone numbers, voice mail accounts and customize billing. In addition, another benefit of VoIP has been its use as leverage to renegotiate existing telecommunications agreements with providers. To discourage agencies from switching to VoIP, providers offer significant savings and discounts to keep customers. These and other attributes have enabled government agencies to use VoIP to lower their phone bills.

Although there are many success stories, there has also been a “stop and go” use of VoIP in state and local government-- mostly because of technology concerns. In an article entitled, “Buying Into VoIP” Shane Peterson of Government Technology Magazine best summarized this point writing, “Voice over IP (VoIP) technology may be maturing, but it hasn't run plain old phone systems out of government offices yet. More enterprises, however, are switching to VoIP as the quality of service improves and CIOs and IT managers discover ways the technology saves money. Government agencies that use VoIP like the return on investment, but acknowledge that implementing these solutions involves more than buying a bunch of IP phones and plugging them in. Making sure the network is ready to handle the increased traffic is critical, observers say, and that can turn into a hefty expense up front. The price tag can be daunting, but current VoIP users say it's worth it.”[16]

In Alaska for example, VoIP was ineffective in its first run, but officials are still pursuing the platform. Network World reported, “Alaska was an early supporter of VoIP and had plans to install a pilot program of 269 Cisco voice phones more than a year ago, but the project ran afoul…The service provider to which the state had outsourced its telecom network failed to use Cisco's best practices, and the system didn't work. The replacement provider fixed the problems, and the state is looking to expand its VoIP use. In Anchorage, the state is building new facilities and outfitting them with Category 5e and 6 wiring that has the characteristics specified by the IEEE to support VoIP. "It's more cost-effective to deploy voice over IP because of reduced wiring costs and lower ongoing operations costs," says Stan Herrera, Alaska's CTO.”[17]

VoIP is often chosen primarily for its management applications, but as an indirect benefit providing significant cost savings. VoIP News reported in July, “Next-generation applications are already having an impact, "Users notice features they didn't have before," said Andy Kuykendall, information technology director for Campbell County, Ky., which introduced VOIP technology last fall using 3Com Corp. equipment. "Some of our people don't stay chained to a desk," Kuykendall said. "Their jobs have them traveling throughout the county. They like [the] ability to change settings themselves, so they don't have to wait for someone else to reprogram speed dials or forward calls to cell phones." Savings are an added bonus. Kuykendall estimates that the county saves $100,000 per year by consolidating telephone business lines and eliminating service fees associated with adding or moving phone lines. He expects to have a full return on his VOIP investment in less than 18 months.”[18]

As local and state agencies more aggressively pursue VoIP for cost savings, it can be expected that information and studies on their success will be more widely publicized.

It can be expected that VoIP deployment in local, state, and federal government will continue to increase. The next step would be to forecast how much money government will save annually with VoIP deployment.

At the state level, telephone service contracts are often negotiated in a decentralized fashion. Telephone service decisions tend to be independently decided by agency leadership and/or its information technology directors. At the federal level, the General Services Administration manages agency telecommunications costs[20], but does not necessarily track in detail telecommunications costs for the military and agencies managed by the Department of Defense.[21]

The research in this report reflects interviews and data collection from over 300 government agencies. Chart A presents a brief data set that provides a snapshot of degrees of differences in government agencies, agency size, number of lines, telephone usage, and line per employee ratios.[22]

      2.1a City of Southfield, Michigan

     Project Coordinator: Gerald Werner
     Staff: 830 employees
     Original stystem: 1,050 Centrex lines
     Post VoIP configuration: subscribe to just 100 Centrex lines
     Capital cost for system: $600,000
     Change in telphone bill: declined to $150,000 from $338,000.
     Annual savings from VoIP: 49%. (Does not include capital upgrade.)

     Project Notes: Uses just 10 PRIs, high speed lines that manage VoIP voice traffic[23]. Manager and staff very pleased with project. Southfield is one of the few city-wide projects in the country described government described as 100% VoIP.

      2.1b Eau Claire School District, Wisconsin

     Project Coordinator: Don Johnson
     1,300 Employees
     Originally had 225 Centrex lines
     Went to just 75 Centrex lines
     Keep 40-50 lines dedicated for security and security systems
     Total cost of system $1,000,000
     Annual phone bill went from $176,000 to $76,000
     Subscribe to four PRI’s
     Annual discount from VoIP system around 57%

     Project notes: Before VoIP implementation; moving, adding, or changing an extension incurred a cost of roughly $100. VoIP has virtually eliminated this expense altogether. In addition, the ECASD has gone from having approximately 25 percent of the phone system users on voice mail to a full 100 percent with the VoIP system. Eau Claire is planning on expanding the project to include the city, county, and the library system in Eau Claire (including both police and fire services).

Agencies around the country provided helpful information to establish average monthly line rates and line per employee ratios. Data, interviews, and an assortment of published research provided tools for developing a model to estimate local, state, and federal government telephone bills. Chart A is a more comprehensive look at agency telephone bill data.

Consistently, every agency surveyed had considerable monthly and annual bills for telephone service. This chart includes a range of agency sizes and annual phone bills. In the case of the California Highway Patrol, there are 10,400 employees that include staff and uniformed officers. Excluding uniformed officers, there are just 3,400 employees. In each case, we see line per employee ratios that were very revealing. The widest spread reported was a 1:16 ratio in the case of California Department of Corrections. In the case of California’s Employment Development Department, it is noticeable that that there are almost 3 lines (2.6:1) for every full-time employee in the department. Monthly line costs on average were between $50 and $80 monthly.

The purchase of major switch products such as Centrex, PBX, and Key systems are frequently based on line per employee calculations.[24] Telecommunications expert Lawrence Harte of Althos provides helpful background on this point.

“Because it takes time for appropriations, many government agencies use older, more established systems. These include multi-line key systems and legacy PBX systems. However, there is a trend to use more advanced systems such as PBX and IP Telephony systems in some agencies, as these new systems are less expensive than their legacy predecessors. Line usage will vary with the type of agency. Consumer focused agencies such as child support and social services require a larger number of telephone lines per government worker than less-consumer focused departments. If the typical worker talks for approximately 1 hour per day, this will require a ratio of approximately 5 extensions for each (1) public telephone line. Offices with a lesser number of stations will also require a lower ratio (more outside telephone lines per worker) as long call holding times and a limited number of lines may result in some calls being blocked even with a limited number of staff in the offices. Newer systems can automatically select the best... or lowest-cost routes....”

Harte and other experts point to the standard practice of telephone companies to use line per employee ratios to estimate telephone usage at large organizations. Using the following variables 1) # employees, 2) # lines needed per employee, and 3) total annual cost per line, a baseline, rough estimate of the government’s annual telephone bill can be produced.

Bi-annually the Census Department publishes employee numbers for the local, state, and the federal government. In 2002, the U.S. Census reported that there are 2,715,500 federal civilian employees, 3,679,940 state employees and 22,758,780 local government employees. [25]

3.2. Line-per-employee ratio

Approximating line usage involves estimating the amount of lines needed per employee. While switching equipment is able to give each employee a different number, it does not mean each employee has their own line per se. Typically, agencies do not purchase a new line for every employee because it is not necessary. For example, an organization with 500 employees would not necessarily need 500 lines unless each of the employees remained perpetually on the phone. So, to keep costs down, agencies subscribe to much less than 500 lines depending on the amount of usage there is in the office.

Monthly line cost data provides the final variable to estimating the annual total bill. The FCC benchmarks consumer use in its study, “2003 Trends in Telephone Service”, reporting the average local line charge was $23.38, and the average business line charge was $42.39[26]. Empirical data and interviews suggest that average monthly costs for government are varied, but in many cases much higher than business and consumer monthly averages.

Long distance per line estimates for government agencies are not widely published. However, without long distance data, per line cost averages are still useful in the model. For example, a baseline cost of $30 per month with a $10 long distance charge would produce a $40 per line total. A $40 line cost with a $20 long distance charge would produce a $60 per line total. Regardless of the composition of telephone charges, the total telephone bill reflects an average per line estimate pretty accurately.

Using empirical data and census data, a range of estimates can be produced to estimate the government’s phone bill. For example, Chart B represents a range of estimates for the total local government phone bills of Indiana, Iowa, and Kansas.[27]

A $20 per line, per month[28] estimate produces an annual line cost of $240. Multiplying the annual line cost by the number of employees reported by the U.S. Census department provides an estimated bill for each state if every state employee had a phone line.

The chart provides a range of line per employee ratios from 1:1 to 1:10[29]. Respectively, if all 234,916 local government employees in Indiana had a phone line at an average cost of $20 per month, the estimated annual bill would be approximately $56,579,840. The chart provides estimates for fewer total lines as well. For example, with 1 line per two employees, 1 line per 5 employees, and 1 line per 10 employees, the respective annual estimates are much lower at $28,189,920, $11,275,968, and $5,637,984.

There are a total of 11,379,390 local government employees. Local government estimates are much higher than federal and state government estimates. There are 87,525 local governments in the United States. Of these local governments, 38,967 are general purpose local governments, 3,034 country governments, 19,429 municipal governments, and 16,504 town or township governments. There are an additional 48,558 special purpose local governments, including 13,506 school district governments and 35,052 special district governments. This number can be used to provide create an estimate of what the total telephone bill would be for all local governments (including the District of Columbia) in every state in the country.[30]

Chart C uses the same model, multiplying U.S. Census state employee totals to produce various estimates. In Charts D and E, it becomes immediately noticeable that the numbers are much lower. This is because there are considerably fewer total state government employees per state than there are total local government employees per state.

Chart F uses the same model to approximate the total federal government phone bill. These numbers are based on the U.S. Census estimate of 2,715,500 federal government employees.

Military employees are significant users of telephone service as well. On bases throughout the U.S., there are approximately 1,500,000 members of the armed forces. That list can be divided as 978,000 that are active and 554,000 of which are civilian.[31] The military’s telephone usage (Chart F) is more difficult to estimate due a number of distinct internal differences in its operation. For example, due to security precautions, parts of military operations have their own telephone communication systems. Due to other security aspects of the military, it remains uncertain how their telecommunications system is configured. However, due to the large military/civilian operations, the military remains a significant user of private sector telecommunications support.

Statistics of the number of National Guard and Reserve estimates are widely available. However, U.S. National Guard and Reserve totals are also difficult to estimate. The nature of their status removes them from counts as office workers that would depend on telephone service. Most U.S. National Guard and Reserve members are employed full time in other professions. Thus it would be difficult to use U.S. Census totals to estimate the on-premise user telephone bill.

Some National Guard and Reserve operations are funded by the state and others by the Department of Defense. This factor also makes it difficult to produce a line per employee ratio, because it is unclear how many people are using the National Guard’s network.

However, empirical data indicates that National Guard and Reserve operations are significant users of telephone service. For example, the Maryland National Guard reported that during the state fiscal year ending June 30, 2003, their annual phone bill was $398,788.26.[32] To estimate the annual telephone usage of the U.S. National Guard and Reserves, it would be necessary to rigorously survey every National Guard and Reserve operation in the

U.S.

In addition, to use data collected from Department of Defense, National Guard offices, and state agencies on National Guard phone usage variables, it would have to be differentiated thoroughly to eliminate double-counting problems.

Chart H represents the total of all local, state, and federal government phone bills. Empirical data suggests that line per employee ratios are closer to 1:1, while average phone monthly line costs are more varied. The purpose of this chart is to consider a range of possibilities. From the highly unlikely number of just 1 line per 10 employees in every government agency in the country, to the price for all long distance and local service for every line in government offices averaging only $20 per month, the minimum totals included in the charts help substantiate the much higher annual telephone bill estimates.

All local, state, and federal government agencies contract outside firms for various services. In 2003, the federal government paid over $300 billion dollars to outside firms for various services.[33] A government contractor’s bill reflects an assortment of costs. Supplies, labor, and transportation are examples of different services the government seeks from outside firms. Because each supplier to government pays a percentage of these monies to telephone service providers, indirectly, government spending on telecommunications services includes the aggregate of all contractor telephone expenditures.

Every five years, the Bureau of Economic Analysis[34] (BEA) completes a survey that estimates the percentage of total revenues each industry sector spends on various commodities. This study includes telecommunications services. [35] As a percentage of total expenditures, industry in the U.S. spent 2.4% of its total revenues on communications. Using this figure, it can be calculated that the federal government indirectly spends an additional $7.3 billion annually on telecommunications services.

State and local government procurement is sizeable as well. U.S. Department of Census reported in 2000, the total operating budget for all state and local municipalities in the U.S. was $1.28 trillion. Empirically, it can be assumed that as much as 30%, or approximately $386 billion of this budget on external contracts are from purchases from private firms. Using this figure, it can be estimated that local and state government indirectly spend at least $9.3 billion on telecommunications services.

Chart I estimates the total of government spending on telephone services via procurement. As a customer, the government has a significant influence both vendors and the market. For example, government agencies can issue vendor requirements for VoIP services, provide credits to VoIP users, or establish firmer rules around how it reimburses its vendors for communications and telephone services. The Federal Acquisitions Regulation System outlines government’s interest in controlling its internal costs as well as encouraging vendors to provide competitive bids to government. Rules outlined in the contracts process include:

Because the federal government can directly change the rules to the contracts process, it has the ability to decide internally to reward contractors that employ VoIP or influence them to consider deploying the new technology. Any number of government policies could reward, increase, or decrease vendor use of VoIP technology to fulfill government contracts.

For example, Section 16 of the Federal Acquisition Regulations outlines various incentives that the government puts in place to reduce costs, improve delivery and improve performance.[37] VoIP usage policy could become a part of an incentive policy. Considering the enormity of government spending on telephone service via its procurement process, government’s option to develop policy focused on VoIP savings is as germaine to the savings question as its direct spending on telephone services. Section E will explore policy recommendations and VoIP.

Considering empirical data, it is probable that estimates that support average line costs above $60 annually are more probable than line costs below $60 annually. This produces a considerable range of estimates. However, considering that it is improbable that line per employee ratios are much greater than 1:2 on average, we are left with a smaller range of estimates: from $6 billion to as much as $19 billion annually. Chart J illustrates this point. Average military estimates increase these numbers on average as much as $1 billion annually.

Indirectly, the government supports an enormous telephone bill annually to meet its operational needs. The estimate of $16 billion annually pushes the average telecommunications bill of government in the range of $23 billion to $36 billion every year.

The government cannot directly affect (or reduce) its indirect telecommunications costs. However, it can directly influence its own telephone bill. A savings of just 30%, would produce an annual savings of $2.1 – $6 billion annually.

Members of the private sector also embrace this number. Rowland Curry, an Austin, Texas-based telecom policy consultant to state public service commissions and consumer groups, noted in a private interview that it is also very much about money: The states have $20 billion a year at stake in the telecom Universal Service Fund and other fees, a sum they are loathe to lose to VOIP.”[39]

Understanding the origins of this estimate is actually helpful in understanding the question of regulating VoIP.

On September 7, 1999, Committee on State Taxation (COST) published its “50-State Study and Report on Telecommunications Taxation.” The report compiled tax data from fifty states and the District of Columbia and compared each of the existing sale taxes on telecommunications to existing sales tax. The study concluded that the average total effective tax rate for the telecommunications industry (including federal excise taxes and fees) of 18.15% is almost three times the average effective tax rate applied to general business (approximately 6.31%). Sales by general businesses are typically subject only to state and local sales taxes. The high effective tax rates applicable to telecommunications are due to the multitude of state and local taxes targeted toward either the telecommunications industry or public utilities – the total effective rate of state and local tax for ten states exceeds 20%.”[40]

Extrapolating the tax rates published in the COST Study, Progress and Freedom Foundation (PFF) published a paper entitled, “the Tangled Web of Taxing Talk: Telecommunications Taxes in the New Milennium” one of the first papers to use the COST study methodology to estimate the aggregate amount of sales tax on telecommunications services collected by government. The following excerpt described their methodology:

“According to FCC data, telecommunications providers reported receiving just over $230 billion in revenue in 1997, which translates into estimated total revenue in the year 2000 of just under $270 billion-- if one assumes an annual growth rate in revenues of 5 percent a year. Roughly 45 percent of this total comprised revenue from local calls, while revenue from all intrastate calls (including local calls) made up 55 percent of the total. As a rough first estimate, these data suggest that state and local telecommunications revenue, or about $135 billion in the year 2000.”

Treating the average burdens (see PFF tax chart attachment) as average tax rates, and multiplying these tax rates by 50 percent of total project telephone revenue in each state, suggest that state and local telecommunications taxes and charges will impose a total burden of

$22 billion in 2000. This estimate equals roughly 8 percent of total telecommunications revenue.[41]

A report by Robert Cline of Ernst & Young (EY) published in 2002 further delved into the COST study data, as well as using FCC, state and local data sets. The report called: Telecommunications Taxes: 50-State Estimates of Excess State and Local Tax Burden used a 2000 update published by COST to and reported: “Telecommunications providers and consumers of telecommunications services paid a total of $18.1 billion in state and local taxes in 1999: $5.3 billion in provider business taxes and $12.8 billion of transaction taxes.[42]

Reading the COST, PFF, and EY studies closely, in many ways the $20 billion ‘total’ derived from the reports is not as discernibly linked to the VoIP tax issue as many would like. After a closer look, there are 5 independent problems with linking a $20 billion decline in taxes to VoIP.

As an overview, each of the studies estimated the approximate amount of total taxes collected also point out that there are significant problems with how telecommunications taxes is reported by state and local municipalities. Each author indicates that the vague and opaque nature of state and local tax collections create inherent difficulties with discerning telecom tax totals.

PFF wrote, “It is much harder to characterize the scope of state and local telecommunications taxation because of the multitude of different jurisdictions that collect these levies…There are upwards of 20 different broad types of state and local taxes and fees levied on telecommunications providers; and the range of specific state and local variation in these charges is almost as broad as the number of different taxing jurisdictions that assess these charges. Elaborating on the challenge of discerning how states and local governments were reporting telecommunications taxes, PFF wrote, “Estimating the total amount of revenue collected by state and local governments from the taxes and fees described above is complex because some of these taxes and charges apply only to revenue from local calls, others to revenue from intrastate calls, and still others to all revenue…”

This problem has an impact on the VoIP tax question another way as well. While the $20 billion estimate refers to total telecommunications taxes, VoIP (in its present form) almost exclusively impacts landline telecommunications. The EY study indicates that its study does not separate wireless taxes, writing: “In this study, telecommunications transaction taxes are defined as those imposed by state and local governments on local, wireless, and long-distance (intrastate and interstate) service providers and their customers.”

It is clearly not the intention of the authors of the COST, PFF, or EY reports to suggest that states and local governments should rightfully collect $20 billion annually in telecommunications taxes, but quite the opposite. Each of the authors respectively, provide tedious detail and analysis to establish that in fact, the $20 billion figure is much too high, as states and municipalities should be collecting substantially less in telecommunications taxes.

All three of the reports were excess study reports. Specifically, the COST, PFF, and EY reports vehemently argued that telecommunications taxes were significantly higher than sales taxes for other goods. Each study produced excess tax estimates, state by state numbers that calculated the difference between sales tax amounts for products and sales tax amounts for telecommunications services. The premise of each report argued that: 1. consumers and businesses were being overtaxed and, 2. better policies should be implemented to more fairly adjust taxes for telecommunications taxes.

The COST report begins with this point writing, “The experience of the telecommunications industry, document in this study, indicates that the tax system imposed on this industry is no longer manageable, and therefore, no longer serviceable. The report demonstrates that the existing telecommunications tax system is even more burdensome and unmanageable that the complicated transactional tax system applicable to general businesses. The study indicates that a higher nationwide average effective rate of transactional taxes applies to telecommunications services (18%) than to sales of goods by general business (6%).”

EY wrote, “An estimated 39% of all telecom taxes - $7 billion – are excess taxes that exceed taxes generally imposed on other business and their customers. Excess taxes exceed 50 % of telecom taxes in 12 states. Summarizing the problem of excess taxes, PFF wrote, “…Since sales and use taxes are important sources of state and local tax revenues, there is a rationale for taxing telecommunications services – but not more heavily and in a more cumbersome manner than other goods and services. State and local tax reform efforts should thus focus both on simplifying telecommunications taxes, and on reducing their burden to levels that are commensurate with taxes levied on other goods.”

In conclusion, while it is not certain how much reform would lower overall taxes, it is quite apparent that each author insists that the total amount of taxes collected is exceptionally higher than it should be. To this extent, the $20 billion total VoIP detractors use is out of context. It is not credible to validate the $20 billion statistic from these reports and exclude their central thesis - telecommunications taxes are unreasonably too high and should be rolled back considerably.

Another problem with the $20 billion calculation is that local and state governments collect both fees and taxes from telephone companies. Considering the broad range of fees listed, it is quite acceptable to ask:

Chart K lists the litany of taxes and fees associated with telecommunications taxes.

The PFF report summarizes this point adding, “determining when a charge that is levied on a telecommunications provider “crosses the line” that divides fees from taxes is not easy. Nonetheless, many telecommunications charges do not appear to finance specific services provided to telecommunications providers or their customers by state and local governments, but instead are either relics of the days when telecommunications providers were heavily regulated by state and local governments, or are charges that effectively defray “public benefits.” Some of the issues that arise are illustrated by right-of-way charges and 911 fees…”

Another problem with the $20 billion figure is that almost 30% of the taxes are not directly correlated to traditional phone bills. This is seen most clearly in Chart L from the EY report. The EY study breaks state and local telecommunications taxes into two categories: transactions taxes and business taxes. If they are different types of taxes, can they be linked together? More specifically, would VoIP threaten both types of taxes equally?

The everyday phone bill lists transactions taxes such as the gross receipts tax, consumer sales and use tax, a 911 tax, and other transactions taxes. However, the total amount of taxes included in the EY estimate of $18 billion includes property tax, capital stock tax and sales tax on business inputs. Business taxes imposed directly on firms in the industry, including property taxes, capital stock taxes on net worth, and sales and use taxes on inputs purchased by telecom companies in the EY study come to 29.2% or $5.3 billion of the total estimate.

The EY report does not predict how much of the $5.3 would be directly impacted by unregulated VoIP. However, it is very likely that they would be significantly less interrupted because they are not impacted directly by consumer or business VoIP usage. How does VoIP impact property taxes? How could VoIP possibly impact capital stock taxes of a company? These points suggest that applying the entire $20 billion figure is problematic.

VoIP detractors insist that unregulated VoIP presents a direct threat to $20 billion in state and local taxes arguing: a) unregulated VoIP diminishes the sale of all telecommunications services, thus all state and local tax revenues will diminish as well, b) unregulated VoIP will drive this number critically downward enough to adversely impact state and local treasuries c) unregulated VoIP will injure the $20 billion imminently; and without immediate regulation, the $20 billion in revenue faces a serious shortfall.

However, this event does not happen tomorrow. It has to start and begin at some date 2005, 2006, etc. Meanwhile, VoIP detractors have not provided any qualifying events, signposts, or indicators to show that the total taxes collected are diminishing or have diminished due to VoIP; and when this tax decrease will accelerate to more significant numbers.

In sum, it inconclusive 1) what the total tax collected from states and municipalities really is (or should be), 2) what amount would be impacted by VoIP 3) how the fast or slow the $20 billion number diminishes due to unregulated VoIP.

To date, the Congressional Budget Office[43] (CBO) in Washington is the only agency that has estimated how fast state and local taxes would diminish due to use of unregulated VoIP. Ironically, in a letter to Senator Lamar Alexander, CBO also referred to the $20 billion tax total, but completely distanced itself from arguments that the entire $20 billion would be threatened by VoIP, commenting, “…However, most industry experts expect that less than one-third of current voice telecommunications services will move to the Internet over the next five years. This suggests that over that period, less than $3 billion annually in state and local telecommunications taxes could be affected by the enactment of ITNA (Internet Tax Non-Discrimination Act).”

The CBO letter is careful to use the words “up to... $3 billion” -- emphasis added. Hence the amount could be substantially lower.[44] Regardless, a $3 billion loss over five years is still a tremendous departure from an immediate $20 billion loss. Meanwhile, if the losses were $3 billion a year beginning in 5 years, it would take almost an additional 6 years (over 10 years) for cities to lose a total of $20 billion dollars.

This is significant also because the $20 billion dollar figure is an annual estimate not a total. Thus, according the CBO estimate, local and state governments would still collect $20 billion annually until 2009, and $17 billion annually thereafter.

Another point is that some would argue that 5 years is more like 10 or 20 years in the Internet age. For the $3 billion loss to occur, we must assume that no other products of innovation will enter the market and generate new sales taxes. For example, one obvious new source of sales tax revenue could be from new VoIP equipment sales.

A recent article in New Telephony Magazine highlighted the associated increase in equipment sales from VoIP use, reporting, “Rebounding in the last quarter of 2003, sales of voice-over-IP equipment rose 31 percent over the previous quarter and totaled $338 million worldwide, according to figures from Infonetics Research's quarterly worldwide market share and forecast service, Next Generation Voice Products. According to Infonetics, "vendors benefited from a surge of budget clearing and [an] overall trend to VoIP." For the long term, Infonetics predicts the worldwide market will grow 305 percent to $5 billion from 2003 to 2007, which is a compound annual growth rate (CAGR) of 42 percent.[45]

While equipment sales worldwide grew, the U.S. market and

U.S.

markets represent a significant portion of the total VoIP business-to-business (b2b) marketplace. Synergy Research Group reported in May 2004, “According to Q1 2004 preliminary market analysis by Synergy Research Group (SRG), North American Carrier VoIP equipment sales were up 6.8 percent, while EMEA and Asia/Pacific equipment sales were up 2.7 and 2.6 percent, respectively. In the quarter, the

United States

represented the largest share of the Worldwide Carrier VoIP opportunity holding a share of 54.9 percent. "The last 5 quarters have been significant for VoIP as a number of high-profile Service Providers, such as ATT, British Telecom, Cablevision, France Telecom, MCI, SBC, Sprint, Verizon, and Bellsouth, either began deploying VoIP or announced their respective strategies to migrate," said Jeremy Duke, President & CEO of Synergy Research Group. "The Carriers' willingness to move forward with VoIP has resulted in some long awaited market momentum for the Carrier VoIP space."[46]

It is safe to speculate that a significant increase in VoIP usage would lead to an increase in the sale of new VoIP equipment and a respective increase in sales taxes from that equipment. In sum, it is fair to question whether these products or others will offset concerns of tax losses.

Each study that has been credited with the $20 billion tax estimates also insists that the number is either too high, inaccurate, or not associated directly with taxes that VoIP would directly threaten. Meanwhile, the CBO letter contradicts the $20 billion analysis directly arguing that at best, we would only see a $3 billion yearly loss in taxes.

With “best guess” estimates on both sides attributed to VoIP savings and VoIP losses, the next logical step is to compare the benefits and losses to government from VoIP deployment.

As discussed earlier, it is not certain that government tax losses from VoIP will not be replaced by sales taxes from other new products. The purpose of promoting VoIP deployment would mean that government could more closely study how VoIP in the marketplace promotes economic development and new sales taxes within respective jurisdictions.

Cognitive dissonance is often used by psychologists to describe when individuals have a belief in one reality but reach a state of denial when they encounter a truth that torpedoes their beliefs. In order to assimilate inconsistent information with their beliefs, individuals experiencing dissonance will often rationalize and even dismiss the facts presented altogether.

This paper asks which two actions are more financially advantageous for government: regulating VoIP with new taxes, or supporting the status quo and allowing the new technology to continue as it has been in the marketplace. While this paper is not conclusive in its findings, the enormity of the potential savings to government and taxpayers suggests that this issue should immediately receive a much closer look.

The idea that government is focused on preserving its $3 billion annually in revenues is understandable. But if its internal IT departments are of the opinion that VoIP deployment can save them twice the amount of money they are losing, we have to question whether government is being prudent when it is moving opposite ideological directions.

This is not to deny local and state government its regulatory authority to regulate VoIP. However, if the government’s responsibility includes the stewardship of trillions of tax dollars, shouldn’t its policy decisions take into account the best interest of its taxpayers – the contributors of every dollar the government collects? As a steward of public dollars, shouldn’t its decisions (as often as possible) provide cost reductions to the public?

Just like any other entity, government must generate income and manage expenditures. Over the last two decades to tackle debt, government has implemented policies to decrease the size of government (and government spending) and lower taxes. In the case of the potentially tremendous savings to government VoIP may afford, it is essential that government study if reducing its spending on traditional telephone service is a faster and more fiscally sound approach to increasing its operating revenue.

Sub Part 16 of the Cost Reimbursement Contracts lists a number of ways that government can change its policy to encourage contractors. Cost plus award fees, cost plus incentive or even cost plus fixed fees are a few of the varied ways government presently structures it vendor relationships.[48]

As opposed to focusing only on sales tax revenue, the government could structure its procurement policies to 1. increase its deployment of VoIP and 2. increase its savings from VoIP, 3) increase its total spending on procurement.

DOC: To view (left click) or save (right click, then follow menu) the tables for this paper as a ".doc" Microsoft Word file, click here.

[1] Government began regulating telephone service with the Mann-Elkins Act of 1910, legislation that mandated common carriers to provide service even where it was unprofitable.

[2] “VoIP Regulation. Regulation Is a Political Not a Technical Problem for VoIP,” Jim Barthold, July/Aug 2004. See http://www.vonmag.com/issue/2004/julaug/features/voip_regulation.htm

[3]“ …That's because applications will run on that broadband pipe and VoIP -- as much as the Bell guys don't like to hear this, VoIP is an application.” Interview with David Dorman, CEO of ATT, September 1, 2004. ZDNet. http://www.zdnet.com.au/insight/toolkit/networking/voip/0,39023893,39157908,00.htm

[4] http://www.skype.com/, http://www.kazaa.com/

[5] http://www.americasnetwork.com/americasnetwork/article/articleDetail.jsp?id=81419

[6] http://www.techweb.com/wire/story/TWB20040722S0005

[7] VoIP is Key to Capturing Government and Education Managed IP VPN Business, Tech Update, David Parks, April 20, 2004. http://techupdate.zdnet.com/techupdate/stories/main/VoIP_Key_Capturing_Government.html?tag=tu.arch.link

[8] http://www.vonmag.com/issue/2004/julaug/features/government_voip.htm

[9] Interview with Karen Hogan, Deputy Chief Information Officer, Department of Commerce and Ken Brown of AdTI, Department of Commerce, September 27, 2004

[10] http://www.washingtontechnology.com/news/16_1/features/16351-2.html

[11] http://www.gcn.com/research_results/avvid2.html

[12] http://www.eweek.com/article2/0,1759,1644594,00.asp

[13] http://home.businesswire.com/portal/site/google/index.jsp?ndmViewId=news_view&newsId=20040927005432&newsLang=en

[14] http://cpr.ca.gov/

[15] “State IT Execs Betting On VoIP, August 23, 2004, Tim Greene, Network World Fusionhttp://www.networkworldfusion.com/news/2004/082304nastd.html

[16] "Buying into VoIP," October 23, 2003, www.govtech.net.

[17] “State IT Execs Betting On VoIP, August 23, 2004, Tim Greene, Network World Fusionhttp://www.networkworldfusion.com/news/2004/082304nastd.html

[18] VoIP’s Second Act, Alan Joch, July 19, 2004, VoIP News http://www.voip-news.com/art/3p.html

[19] Although the research team for this report was able to collect empirical data from a number of very cooperative and helpful agencies around the country, it did not have the resources to collect the totals for every state and federal agency in the U.S. Research team solicited responses from about 200 agencies and received 47 responses over a 4 month period.

[20] Federal Technology Service Overview http://www.gsa.gov/Portal/gsa/ep/channelView.do?pageTypeId=8199&channelId=-13291

[21] Federal Acquisition Policy divides purchasing into two divisions Civilian Agency Acquisitions Council and Defense Acquisitions Council. GSA and DOD are its respective representatives. See http://www.arnet.gov/far/loadmainre.html

[22] See Table section.

[23] www.webopedia.com/term/pri.html

[24] Interview with Lawrence Harte, author, president of Althos Publishing. http://www.althosbooks.com/, http://www.lawrenceharte.com/

[25] See http://www.adti.net/kenarbeit/www.census.gov.

[26] www.fcc.gov

[27] State by state estimates are provided in the notes section

[28] The variances of $20, $30, $60, and $90 average line rates were chosen based on empirical data and anticipating large long distance bill discounts and/or bulk rate discounts. These amounts are contrasted with of 1:1, 1:2, 1:5, and 1:10 line per employee ratios. A spread of ratios was chosen to account for a range of considerations in various types of government offices around the country. For example, government field office workers are not in the office and are less likely to have an individual phone line, thus accounting for lower line per employee ratios.

[29] Conversely, upon surveying different government agencies, empirically, it is was not unusual for offices to have higher than 1:1 ratios. Higher ratios were due to the prevalence of private lines, conference room lines, lobby lines, and a wide assortment of lines used for emergency and redundancy purposes. Erring on the side of being conservative, these estimates predict that most organizations have no more than one line per employee.

[30] U.S. Census, 2002

[31] http://web1.whs.osd.mil/mmid/M02/fy03/M02_2003_Personnel.pdf

[32] Freedom of Information Act and Privacy Act Request, AdTI, April 26, 2004. (FOIA request 04-081)

[33] Federal procurement Data System, Federal Contract Actions & Dollars by Executive Department and Agency. "Actions reported on SF279 and SF281, Fiscal Year 2003 Through Fourth Quarter."

[34] http://www.bea.doc.gov/

[35] http://www.fpdc.gov/fpdc/fpr_2003/FPR2003.pdf

[36] Federal Acquisition Regulations, 1.102 Statement of guiding principles for the Federal Acquisition System. See http://www.arnet.gov/far/farqueryframe.html

[37] http://www.arnet.gov/far/farqueryframe.html, Subpart 16.4- Incentive Contracts, Federal Acquisition System

[38] http://www.adti.net/kenarbeit/www.alexander.senate.gov/news/205709.html.

[39] http://www.findarticles.com/p/articles/mi_zdewk/is_200406/ai_ziff130114

[40] COST study, available online.

[41] http://www.adti.net/kenarbeit/www.pff.org/publications/communications/pop7.12tangledweb.pdf

[42] "Telecommunications Taxes: 50-State Estimates of Excess State and Local Tax Burden," Robert Cline, State Tax Notes, June 3, 2002.

[43] http://www.cbo.gov/. http://www.cbo.gov/showdoc.cfm?index=5040&sequence=0 Letter to Senator Lamar Alexander in notes section

[44] CBO interview with Kenneth Brown of AdTI, September 17, 2004

[45] VoIP Equipment Sales Up 31%, New Telephony, September 1, 2004. http://www.newtelephony.com/news/809.html

[46] “Carrier VoiP Equipment Ready for Substantial Growth”, May 21, 2004, http://biz.yahoo.com/iw/040521/067660.html, www.sgresearch.com

[47] "The Excessive State and Local Tax Burden on Wireless Telecommunications Service," July 19, 2004, page 6. See, http://www.adti.net/kenarbeit/.

[48] http://www.adti.net/kenarbeit/www.arnet.gov, sub. part 16.