Archive for December, 2014
In recent days, the net neutrality crowd has seized on select, abbreviated versions of comments by certain executives of Internet service providers (ISPs) as evidence that ISPs are in fact supportive of the public-utility-style regulations being considered by the FCC for internet access service. Even the Chairman of the FCC made hay with the comments to advance his regulatory agenda.
As it turns out, the “gotcha” quotes were amplified in the media, while statements consistent with the “regulation-can-be-harmful” thesis were neglected. Even if we ignore what else those executives said, corporate financial officers (CFOs), or any executive for that matter, don’t have complete say over their firm’s investment decisions. That’s because external investors who lend money to ISPs are equally if not more important, particularly over the long run.
A small helping of investment theory is in order. Tim Karr at Free Press is fond of characterizing the ISP investment decision as an all-or-nothing affair, but in reality, investments (like any decision in economics) are made at the margin. Each project has a different expected return. And even within a project—say, fiber to the home (FTTH)—the expected return will vary depending on the city in which the investment would be made.
As any CFO knows, basic investment theory teaches that a firm invests in a project so long as the internal rate of return (IRR) on a project is greater than the minimum required rate of return, as measured by the firm’s the cost of capital. This is simple, folks: Line up your projects from highest to lowest IRR, and fund the ones that exceed your cost of capital.
So to believe that public-utility-style regulation would undermine investment at the margin, one needs only to believe that reclassification would either (1) increase an ISP’s cost of capital or (2) reduce the expected return of a set of ISP investment opportunities. Projects with an IRR above the pre-reclassification cost of capital but below the post-reclassification cost of capital are called the “marginal” investments—that’s a kind way of saying the forgone investment caused by reclassification.
Let’s start with the first mechanism–raising the cost of capital–and leave the returns constant. By one estimate, Verizon’s weighted average cost of capital was slightly below 6 percent in April 2014, right before the FCC’s light-touch approach was abandoned for political reasons and Title II gained traction. Because Verizon has largely stopped its FiOS investment, one can infer by the investment rule that the IRR associated with expanding FiOS into other cities within Verizon’s footprint is south of 6 percent.
As bullish as Verizon CFO Francis Shammo is about investment, Verizon’s investment decisions involve participants—namely, the investor community—not entirely under Verizon’s employ. (Notice that Mr. Shammo referred to the light-touch approach of section 706 as “section 765” in the “gotcha” quote that appeared in the Washington Post, which suggests that he isn’t following the net neutrality debate too closely.) More on Mr. Shammo in a bit.
External investors could demand a risk premium (over and above what they otherwise would demand) to compensate for the added risk associated with the new rules. An investor may ask: Why should I lend an ISP money for a new project if there is a heightened chance under reclassification that the ISP would be subject to rate regulation or mandatory sharing rules? Through the haggling between Verizon and investors, the new risk could manifest itself in the form of a higher cost of capital. While the CFO needs to allay investor fears related to regulation, no amount of corporate cheerleading can compensate for a perceived increase in risk.
Turning to the second mechanism, holding constant the cost of capital, reclassification could reduce the expected return of an array of investment projects by a certain percentage. This would not mean that all such projects would be abandoned. But if Project A’s IRR was reduced from 10 to 9 percent, while Project B’s IRR was reduced from 6 to 5.4 percent, and if the ISP’s cost of capital were 6 percent, then Project B would be abandoned.
In a seminal application of this theory, in 2002, Cambridge Strategic Management Group (CSMG) examined the potential effects of mandated unbundling on FTTH deployments by incumbent and competitive providers. CSMG projected that if unbundling were required, all-fiber deployments would pass only 5 percent of U.S. households in a ten-year period. In contrast, if unbundling of fiber loops was not mandated, CSMG estimated that by 2013 FTTH could be economically deployed in 31 percent of households. In 2003, the FCC relied in part on these findings to decide not to mandate unbundled access to FTTH loops, concluding, “We expect that this decision to refrain from unbundling incumbent LEC next-generation networks…will stimulate facilities-based deployment.”
When sifting between the disparate messages coming out of ISP executives, tech reporter Brian Fung pointed out that the musings of ISP executives should be credited because SEC rule 10b-5 compels executives to tell the truth—kinda like Wonder Woman’s lasso of truth. (Free from these restraints, ISP lobbyists take “extreme rhetorical positions.”) But publicly traded firms have an equal fiduciary obligation to shareholders not to lose money. Applied here, that means not to invest in projects whose net present value is negative.
Once you understand a bit of investment theory, it is not much of a stretch to see how heightened risk associated with reclassification could reduce investment at the margin. And that would be a terrible thing.
The whole episode of selectively quoting CFOs could introduce a new word into the lexicon: “Shammoed.” As in, you’ve been Shammoed for purportedly making an admission against your interest. To clarify his thinking, Mr. Shammo posted further comments to the Verizon Policy Blog:
“But as we and other observers of the net neutrality debate have made abundantly clear, experience in other countries shows that over-regulation decreases network investment. If the U.S. ends up with permanent regulations inflicting Title II’s 1930s-era rules on broadband Internet access, the same thing will happen in the U.S. and investment in broadband networks will go down.”
Per the SEC requirements cited above, one should give equal weight to these statements as the ones flagged in the media. Taken together, any balanced reading leads to the same conclusion as economic theory: Regulations that reduce the expected rate of return on a given class of capital investments or raise the cost of capital will cause such investments to fall at the margin.
Flaws in Free Press’s Alternative Estimate of New State and Local Fees Attributable to Reclassification
In reply to a PPI study that I co-authored with Bob Litan, Free Press issued an alternative estimate of the annual consumer burden associated with reclassifying broadband as a Title II service. Free Press is not arguing over the fact of consumer injury; they are just quibbling over the extent of the damages. Whereas our study placed the estimate at $15 billion in new state and local fees for broadband consumers, Free Press arrives at a mere $4 billion in new fees. As their report states:
If the FCC reclassifies broadband access as a Title II service, it will also (based on precedent) declare that broadband is a purely interstate telecom service. Because broadband access is interstate and not intrastate, none of the intrastate taxes or special telecom fees would apply. The only taxes that could apply would be state sales taxes levied on interstate telecom services. Even if you used PPI’s fuzzy math, this would amount to approximately $4 billion in total, nowhere near the $15 billion sum Singer and Litan cite.
Free Press makes at least six erroneous claims in its reply (as well a few personal attacks on me and on PPI that I will ignore). This response is meant to set the record straight.
Claim 1: “Because broadband access is interstate and not intrastate, none of the intrastate taxes or special telecom fees would apply.”
While state utility commission regulatory authority is limited to intrastate services, state taxing authority has never been limited this way. Accordingly, the only state or local taxes in our analysis that could possibly be avoided if the FCC were to declare broadband to be an interstate service would be the state-based universal service fees adopted pursuant to state utility commissions. Yet these fees make up one component of several state-based charges and taking them out of the calculations does not materially reduce the fee burden on consumers. For example, in California, in addition to a “Universal Lifeline Telecom Service Surcharge,” telecom service providers are assessed a State 911 Surcharge, a Local 911 Surcharge, a Public Utility Commission Fee, a California Teleconnect Fund Fee, a Universal Lifeline Telecom Service Surcharge, and a Utility Users Tax.
Notwithstanding the fact that state-based universal service fees represent one of several state-based telecom fees, we don’t know whether the FCC would declare all broadband services to be interstate. For example, in a proceeding involving reciprocal compensation for dialup Internet traffic, the FCC held that such traffic was “jurisdictionally mixed and largely interstate in nature.” Thus, it’s speculative for Free Press to claim that the FCC would declare all broadband service to be interstate here.
Further, there are already a handful of states that assess universal service fees on interstate voice revenues, including South Carolina and Vermont. This suggests that, when push comes to shove, there is no ironclad protection from taxation if the state decides it wants to grab those revenues.
Free Press’s claim that state sales taxes are the only state taxes or fees that would survive if the FCC declares broadband to be an interstate service is simply wrong. To the extent that a state adopts any tax or fee under its general tax authority, from property taxes to gross receipts, rather than under the authority of a public utility commission, the hypothetical interstate designation affords zero protection. The reason that states haven’t been able to tax broadband services isn’t because of any FCC interstate classification, but because of a federal statute that has prohibited taxes by states on Internet access, a point I’ll address in a bit.
Claim 2: “PPI’s argument also ignores the possibility that the FCC and Congress could take additional steps to remove or limit any future taxes or fees.”
Litan and I conceive that the FCC could in theory forbear from assessing federal universal service on broadband revenues but that would be at odds with the fact that the FCC has effectively converted the focus of the universal service fund from voice to broadband. And the agency’s willingness to increase the E-Rate plan by $1.5 billion this year demonstrates the political forces at work to expand program demand. How could the FCC decline a similar infusion for its LifeLine program once it expands the assessment base to include broadband in light of its recent E-Rate decision? One can hear the lobbyists knocking at the door.
Claim 3: “Whether or not the FCC does [forbear], adding broadband into the USF mix wouldn’t impact the fund’s overall size.”
Litan and I conceive that reclassification by itself (absent the powerful political forces) does not imply an increase in program demand for the federal universal service fund (FUSF). Indeed, when calculating the new federal fees from reclassification, we conservatively assumed no increase in program demand. Adding broadband revenues to the mix, however, does alter the consumer burden in two important ways that are lost on Free Press.
First, whereas the FUSF is largely financed today by voice customers (long-distance in particular), the fund would be largely financed by broadband customers (which might be a different set of customers) if the FCC proceeds to reclassify broadband as a telecom service.
Second, and more importantly, whereas the fund is split fairly evenly between business and residential voice customers, the fund would be largely financed by consumer broadband customers under reclassification. Indeed, Litan and I estimate that the consumer burden would shift from 50 percent to 62 percent of the contribution base, as the consumer contribution (compared to business) of broadband revenues is proportionally greater than the consumer contribution of long-distance revenues.
Claim 4: “Just as the FCC can decline to extend USF assessments to retail broadband access at this time, it also has the authority to preempt states from doing so.”
It is not clear that the FCC has the authority to preempt states from taxing in whatever method the states see fit. Indeed, the FCC traditionally has been loathe to preempt states from acting in areas even where the agency has been given authority to act. Section 253 of the Act authorizes the FCC to preempt state laws that would impair a carrier from providing interstate or intrastate telecom services. It’s hard to see how assessing fees on broadband would impair any carrier from providing broadband services. At most, such fees would reduce broadband penetration by squeezing out marginal (price-sensitive) customers. But that’s a far cry from impairing entry by a provider.
Claim 5: “If Congress doesn’t renew this [Internet Tax Freedom] Act, which expires on Dec. 11, this all becomes moot: There will be new taxes no matter what the FCC does.”
If the Internet Tax Freedom Act (“ITFA”) is not renewed—that is, if nothing prevents the states from taxing broadband access—the states will be largely free to apply taxes to broadband access. As noted above, this has been the most important obstacle (besides reclassification) that has blocked states from grabbing a portion of broadband revenues. What Free Press fails to recognize, however, is that if the ITFA is not renewed and the FCC reclassifies broadband, then states will be encouraged to assess such taxes since the feds will have defined broadband as a telecom service on which states already have assessed taxes and fees.
Claim 6: “If Congress wants to renew this special exemption, making sure it applies to Internet access after the FCC reclassifies is a very easy legislative fix.”
Free Press admits that, as the ITFA is currently written, if the FCC were to reclassify broadband as a Title II service, a legislative fix is needed to ensure that the current protection would extend to telecom services. Stated differently, Free Press admits that the ITFA as currently written would not prevent these new taxes if broadband were reclassified. If there was any doubt on their views, Free Press estimated that the application of state taxes to broadband revenues would generate $4 billion per year, an admission that the ITFA offers no protection as written.
In regards to the likelihood of an amendment, the language of the ITFA has been roughly the same for more than 15 years. The notion that there could be changes to the ITFA to avoid this outcome is farfetched.
* * *
Free Press concludes that “If the FCC does nothing more than stick with precedent and designate broadband as an interstate telecom service, the average potential increase in taxes and fees per household would be far less than PPI estimates.” It seems like they have conceded the big issue: Free Press and PPI are now arguing over the extent of the pain inflicted on broadband customers. To reduce the estimate from $15 to $4 billion, however, Free Press has badly misconstrued the telecom landscape. Dialing back the vitriol in their response might permit them to think more clearly.