The wireless broadband business in India, led by the Rajan Raheja-owned Hathway Cable & Datacom, may be heading for a downhill ride. Recent data suggests broadband penetration in India increased to 27.3 per cent in October 2017 from 25.9 per cent in September 2017, which translates into net addition of 15.3 million subscribers. Interestingly, telcos have been the biggest beneficiaries, led by Mukesh Ambani’s Reliance Jio.
According a research conducted by Citi in Dec 2017, broadband subscriber base of telcos grew by 15 million as of October 17, with Jio leading the pack with 7.3mn new subscribers, followed by Bharti (4.6 mn) and the Idea-Vodafone combine (3.9 mn).
Being the leader of the pack, the heat is clearly visible on Hathway, which holds 52 per cent share of the total MSO cable broadband market in India. The September quarter saw broadband subs growing from 0.66 m to 0.69m (QoQ) as it saw net additions of 30,000 subscribers. Net subs additions for the past three quarters have been in this range, as competitive activity from wireless players continued to remain high.
And it I not just about the subscription base; broadband players have also seen a steady decline in their average revenue per user, or ARPU, which is down to Rs 717 from Rs 730 on a quarter-on-quarter basis. Industry watchers say the trend is likely to continue for a while, as established players cut tariffs to retain long-term subscribers.
In a conference call after the September quarter numbers were released, Hathway’s management had indicated that it does not expect any increase in ARPU over the next 18 months.
According to Ambit Capital, Hathway’s broadband business is likely to generate a mere 6 per cent post-tax RoIC, despite generating a robust 39 per cent EBITDA margin, thanks to high capital intensity. Quite obviously, cable business stocks have been laggards in the bull-run on Dalal Street. The Hathway Cable stock has seen an erosion of around 43 per cent in its market value, at a time when the benchmark Nifty gained almost 32 per cent.
Analysts say the increasing reach of internet broadband, backed by an aggressive tariff play by telcos, could sound the death knell for the business. Rohit Dokania, Research Analyst at IDFC Securities, says Hathway’s underperformance may be linked to the lack of increase in net realization across business verticals. Mr Dokania, however, acknowledges that Hathway’s execution rate on the wireline broadband side among the best in the MSO pack.
Another area of concern could be the high net debt ratio, which has been on a steady rise because of the high capex involved in the broadband business.
MANDEL NGAN/AFP/Getty Images
Graphic by Daniel Kleinman, Forbes Staff
Broadband has been in the news a lot lately. Last week, President Trump traveled to Tennessee to announce an executive action on rural broadband; earlier this week bipartisan legislation to “dig once” was introduced; and just yesterday a group of House Democrats introduced legislation supporting municipal-provided broadband (“muni broadband”). As of today, there are 21 states with restrictions on community broadband and a federal court has once again ruled that the FCC has no statutory authority to preempt these laws, which adds weight to the proposed federal legislation, though it is unlikely to garner much Republican support.
This version of the Bytes Chat discusses the wisdom of those restrictions against muni broadband with three economists who are following the issue closely. Robert Seamans is an Associate Professor of Management and Organizations at the Stern School of Business at NYU, and the author of Fighting City Hall: Entry Deterrence and Technology Upgrades in Cable TV Markets. George Ford is the Chief Economist at the Phoenix Center for Advanced Legal and Economic Public Policy Studies, and the author of The Impact of Government-Owned Broadband Networks on Private Investment and Consumer Welfare. Kyle Wilson is an Assistant Professor of Economics at Pomona College, and the author of Does Public Competition Crowd Out Private Investment? Evidence from Municipal Provision of Internet Access. The Chat was moderated by the editor of Washington Bytes, Hal Singer.
It has been edited lightly for readability. Hal Singer: Welcome fellow econs! Let’s get right in.
When considering whether a town should own and operate a local broadband network, from whose perspective should we be looking? Consumers or taxpayers? Does it matter?
I’m looking for the right analytical lens. Rob Seamans: I think the key is the long-run benefit of both. A town’s “consumers” are likely its taxpayers.
And a focus on taxpayers doesn’t necessarily mean lowering taxes. For example, people willingly move to “high tax” areas to access better schools. The key is providing something of benefit to residents in return for taxes.
George Ford: From an economist’s perspective, we look at economic welfare and efficiency. Of course, there are many angles from which to view this issue and economic welfare and efficiency is a view enjoyed by few. Most people look at it through the lens of broadband hysteria.
I think if you just look at the performance of these systems, you’ll see muni broadband projects are a poor way to spend government money. Failure is almost certain and the cross subsidies are huge. And, I think we are one lawsuit away from the whole thing shutting down–a legal challenge under the American Railroads decision.
Kyle Wilson: If a town is deciding whether or not to install a broadband network, I think we would hope that it does so on the basis of whether it represents a net improvement in the well-being of the town’s internet users and the gain or loss in revenues from the town’s revenues, which may include drawing upon taxpayers. Singer: Suppose your town wants to get in the broadband business. What do you tell them?
In particular, what are the specific economic factors that should influence the decision-making? Ford: Mainly, why isn’t there broadband now? What explains that?
Why isn’t there adequate demand that justifies the costs? How can we increase demand? How can we reduce costs?
Is there private investment in the market already? Where are the subsidy dollars going to come from? Wilson: In most cases when a municipality considers entry into the broadband market, there is already broadband but its quality may not be satisfactory.
The existing firms appear not to find improving the network profitable. But perhaps the municipality values more than just profits. Maybe it values the well-being of its consumers and other benefits the town might reap.
Singer: In addition to George’s list, I am curious whether the current finances of the municipality should be a factor. Suppose, for example, that a town is in debt: Should it go take on even more debt to fund a broadband network? Or is this consideration not relevant?
Will it have to pay higher interest rates to service its debt? Seamans: I think each town needs to make its own decision, Hal. Ford: These systems are built in markets with municipal electric utilities, now almost exclusively.
The debt is placed–or hidden–in the electric utility where profits are abundant and easily raised. The municipal electric utility is the only remaining source of subsidy, which brings up legal and fairness issues. As a general principle you take on debt when you can afford to repay it with the returns.
It makes no difference how much debt you have, except that the rate you pay will be lower if you have a demonstrated ability to meet your debt. Seamans: If the town is in a lot of debt, muni broadband probably makes no sense. If it isn’t, and if there’s a need, then it probably makes sense.
Do you have a municipal electric utility or not? If not, don’t consider it. If you have a municipal electric utility, then consider it.
Do you have enough physical plant that you can enter and operate at relatively low cost? If so, you might be able to enter at lower cost than a private provider, serve more customers, and reap benefits from broadband. Singer: Here’s another consideration for our hypothetical consulting project for the town: A dollar spent on broadband is one dollar not spent on roads or schools or pensions.
Is there some way to rank the social returns of these potential investments? Or is there no limit to what our town can spend? Seamans: Well I suspect each town would rank these differently.
This is up to the residents of the town or city to decide. Not a bureaucrat in a state office. Ford: It would be interesting to rank returns.
I don’t think the returns of a third, or even a second, broadband network are very high relative to schools and roads. Seamans: But we willingly pay for hospitals, schools, and other government services that private providers won’t. Why is broadband different?
Ford: Public education, public roads, are the norm. Private alternatives take a small piece of the market. Public and private hospitals have long operated alongside each other, and the market is quite large.
In broadband, we have private provision for decades with large investments being made. Billions are invested in private broadband networks. Broadband is not a public good.
Wilson: It’s worth remembering that unlike schools, installing a municipal network creates a new stream of revenue, even though it may not be enough to break even. Singer: I would suggest that the state of local competition, which varies tremendously across the nation, is also relevant to the town’s cost-benefit calculus. For example, in an “underserved” area where consumers face a choice of (high-speed) cable modem or a (slow-speed) DSL provider, a second high-speed wireline provider could generate substantial welfare gains.
But in an area where consumers face two high-speed wireline providers, the potential upside from a third wireline provider seems remote. Conversely, in an area with no high-speed wireline options, the so-called “unserved” areas, it seems like the cost of muni entry (in terms of crowding out private investment) would be small. Can we at least agree on that premise?
Wilson: Agreed! Though, I’ll add that the areas we label as “unserved” depends on our definition of “high-speed.” As consumers demand faster and faster speeds, the number of “unserved” areas will increase if investment doesn’t occur one way or another. Ford: To be clear, if there is no broadband, it may make sense for a city to subsidize a broadband network.
It’s the use of muni broadband to “increase competition” that is problematic. Seamans: So we agree that if there is no broadband, it’s okay for a city to provide broadband? Singer: More consensus!
Yay. It looks the bone of contention, then, is how to advise towns that are underserved. George says “no go” on muni broadband, and Rob says yes.
Before we dive into Kyle’s paper, I want to hit one more background question: The issue of “selection.” There is a reason why the telco has avoided deploying fiber by the year 2017 in a given town. These aren’t randomly chosen areas. If the local telco thinks the private returns aren’t sufficient to cover the private costs, why should we be certain that the calculus is any different for the municipality?
In other words, could selection be leading these municipalities into unprofitable investments? Wilson: If we believe that private firms and municipalities value the same things, then this thinking is correct. It’s safe to assume that private firms only deploy fiber if it’s profitable.
But is this really what is driving municipalities to invest? It seems likely they are also taking the consumer surplus of their constituents into account. So perhaps municipalities are making investments that profits alone can’t justify, but if profits plus the benefits to consumers outweigh the cost of investment, then I would argue that the municipality is making a sound investment.
Ford: You can’t pay debt with consumer surplus. Weighing consumer surplus makes for privately uneconomic decisions, and these system typically compete with private providers. Seamans: Why is the test “what private firms do or don’t do”?
Wilson: A private firm’s lack of investment in a town signals that it isn’t profitable for them, not that it isn’t profitable for anyone. If an existing private firm deploys a fiber network in an area where they already operate a DSL network, they might attract some new customers. But largely, they are going to cannibalize their existing market share.
For a new entrant, all of their customers will be new, so the investment is more likely to be profitable. Ford: These areas are underserved for a reason–bad economics. Start there and ask how we solve the bad economics problem.
Let’s not pretend that by some miracle there are profits to be had in Nowhere, USA. Seamans: There is something different about a municipal electric utility in a city. They can potentially provide broadband at lower cost.
Ford: I doubt it. How can a city of 20,000 people be more efficient than Comcast, Charter, AT&T, Verizon? The cost of equipment and programming will be significantly higher.
Seamans: On equipment, some of the costs may be higher. For cable TV programming, there are coop buyers that buy in bulk. But the biggest costs have already been made by the municipal electric utility–for example, most municipal electric utilities have a lot of dark fiber, they already send bills out each month to customers, they have teams of service people and customer service reps.
Ford: The coops still pay much higher prices than companies like Comcast. And dark fiber is part of the broadband network, not the electric network. Singer: Let’s move to Kyle’s new paper on muni broadband, which caught the attention of two famous econs (Rob and George) and one obscure econ (me).
Kyle, can you explain what hypothesis you were testing in your study? Wilson: There’s really three questions I set out to answer: (1) To what extent do municipalities value profits vs. consumer surplus when making their decisions?; (2) What effects do municipal networks have on private investment?; and (3) What effect does a ban on municipal provision have on economic welfare?, taking into account the profits of both private firms and municipalities, and the welfare of consumers. Singer: And what dataset did you use?
And how did you go about assembling it? Wilson: I primarily used data from the National Broadband Map (NBM). For every census block (think city block) in the U.S., this data set indicates what firms offer services there, and for each firm indicates their technology and speeds offered.
It’s available twice per year from 2010 through 2014, so I downloaded all of the data sets and matched up the census blocks and providers over time. I was then able to identify instances where a new provider entered a market, an existing provider upgraded its technology, or an existing provider exited the market. Seamans: Kyle, how does the data get aggregated to the NBM?
Wilson: The NBM data is collected separately by each state through the State Broadband Initiative. Then, the NTIA presents it in the NBM, still at the block level. Ford: NBM data can be pretty filthy at that level.
And make sure you keep an indicator for who collected it. The quality varies widely. Wilson: I think you’re probably right about that.
The map says I had two options at my old apartment in Tucson, and that definitely wasn’t the case! But I aggregate it up to the zip code level, so that should take care of a lot of the noise in the data. Singer: Data nerds!
Ford: I helped with the data in Alabama. There are many smaller cable operations that don’t want the government to know they exist. Seamans: Interesting.
Why? Ford: Tom Wheeler and his ilk. It’s more of a regulator thing.
No good comes from it. Singer: Kyle, what about your broadband pricing data. Seems like it involved some web scraping.
Wilson: I collected it by querying a price comparison website for every zip code in the U.S. Interestingly, there’s lots of variation in prices across firms and plans, but I found no variation across geography for the price of a particular firm’s plan. Singer: You identified two specific factors that determine whether muni broadband increases investment on net: the “crowding out” effect and the “preemption effect.” Can you explain those to our readers?
Wilson: It’s important to consider the effect of muni networks on the decisions made by existing private firms. It’s likely that there are cases where a private provider would have invested in improving its network in the future. But once a municipality installs its own network, the private provider no longer finds it profitable to upgrade its network.
This is the “crowding out” effect. Singer: And the preemption effect? Wilson: It’s also likely that there are cases in which the mere threat of a municipal installation causes the existing local private provider to invest in improving its network, in order to fend off a municipal installation.
Which of these effects is stronger determines whether municipal broadband boosts or hinders private investment. Singer: How do you estimate your crowding out and preemption effects? Wilson: Before I can estimate those effects, I need to estimate some values that govern consumers’ preferences and firms’ costs.
I look at how households’ broadband purchase decisions differ in areas with high and low prices, high and low incomes, etc. This gives me a sense of what types of areas are likely to value high-speed Internet the most and where providing it is likely to be profitable. Singer: I’m with you.
Go on. Wilson: Then, I estimate a model of firms’ and municipalities’ decisions to enter markets and invest in fiber. Since I already have a sense of how profitable markets are, the decisions we observe firms and municipalities making reveal the costs of entering and deploying fiber.
To the extent that municipalities act differently from private firms, this reveals whether they are driven by consumer welfare or just profits. I find that they value consumer welfare about 22 percent as much as they value profits. Singer: And the crowding out and preemption effects?
I’m dying here. Wilson: Now we’re ready. The last step is to take this same model but remove municipalities from it in order to imitate a ban on municipal provision.
I then simulate from the model what actions private firms would take in the absence of municipal provision (or threat thereof). Seamans: Kyle, did you have any ex ante expectation about how much municipalities would value consumer welfare versus profits? I would have expected higher value of consumer welfare than 22 percent.
Ford: I think that’s a high value. There’s no revenue from consumer surplus, suggesting the cities do place some concern on financial considerations (which I think they do, just not enough). Wilson: My prior was that the weight on consumer welfare would be higher as well.
I suspect they end up placing a high weight on profits because they need to justify the network in terms of breaking even. Ford: It seems certain that they behave differently, doesn’t it. They enter in markets and do things private firms do not.
I think we could interpret consumer surplus here as the degree of the uneconomic nature of muni entry. Wilson: Agreed, it does seem certain they behave differently. And the less profitable of a market they are willing to enter, the greater the value they apparently place on consumer surplus.
Seamans: Kyle, also is there any way to benchmark the 22 percent? Are there other studies that look at the consumer welfare vs profit tradeoff of public entry? Wilson: I know of one other study that does something similar. Christopher Timmins has a paper that looks at municipal water administrators and he actually finds that the city places more weight on consumers than it does on profits–very different from what I find.
Singer: How do you know what is a preemptive upgrade by an ISP versus just a normal upgrade? Or an upgrade for a non-residential purpose, as to increase backhaul capability or serve major institution? Wilson: I run the simulation I described both with and without the possibility of municipal provision.
I then take note of markets in which a private provider upgrades when facing the possibility of municipal provision. If that provider doesn’t upgrade when municipalities are banned, then I conclude that it was acting preemptively because of the threat of municipal provision. Singer: Well that’s a fancy answer!
Let’s get to your major findings. You conclude that government spending on broadband crowds out some private investment (net of preemption), but that the decline in private investment is more than offset by the increase in public investment, so that net investment increases. In particular, you find that placing a ban on new public provision decreases the share of markets with public provision of fiber by 3.1 percentage points but increases the share of markets with private provision of fiber by 0.76 percentage points, relative to the case where no ban is in place.
Did I get that right? And those numbers seem like small effects. Wilson: That’s right.
If these sound like small numbers, that’s because municipal networks are still quite rare in the grand scheme of things. So my finding is that if we had banned municipal provision in the states that currently don’t restrict it in 2010, we would lose fiber availability in 3.1 percent of markets that currently have it provided through their local government. But, we would gain back some private investment that had previously been crowded out, in 0.76 percent of markets.
This gain in private investment isn’t enough to offset the loss in public investment, so the ban does more harm than good. Put in relative terms, I find that, roughly, for every four municipal installations that occur, one private firm is crowded out from making an investment in fiber. Seamans: Kyle, just trying to interpret the result a bit.
Why is there relatively little crowding out? Is it because muni entry is mostly in areas that are not served, or underserved by private companies? Wilson: I’d wager there is little crowding out because cities are installing networks in areas where private firms were unlikely to make investments in improving quality.
The crowding out can only occur if the private firms would have otherwise invested. Ford: Kyle, is there any extension of crowding out through a threat in other markets? Put differently, is the reduced fiber investment by private firms only in the markets where the public firm entered?
Wilson: The crowding out I measure is isolated to the markets where the public firm entered. Do you imagine that municipal entry in one city might reduce private investment in a different city? Seamans: Or it might induce more investment to preempt muni entry, as I show in my paper!
Ford: Your paper is a little different, Rob, focusing on cable systems. But think about a private fiber company, such as WOW in Alabama. They may alter their decision to expand due to the fiber system in Opelika.
You see a lot of smaller providers complain about muni systems. They foreclose opportunities. Kyle, did you have a dummy for municipal electric utilities?
That’s where the threat is. Wilson: No, I didn’t have a dummy for municipal electric utilities. If possible, I think it would be worth considering municipal electric utilities separately since, as Rob has mentioned, they may have a lower cost of entry than a city without municipal electric.
Ford: A lower cost of entry and a ready source of subsidy dollars from captive electric ratepayers. The latter being more important, I think. Singer: Kyle, you find that a hypothetical ban on public provision in 30 states (that currently have no restrictions on muni broadband) increases private firms’ profits by £3 billion, decreases municipalities’ profits by £21 billion, and decreases consumer surplus by £1 billion.
We all hate when consumer welfare falls, and we will get to that in a bit, but why should we care about a change in muni profits? Wilson: Any profits the muni might earn could be reinvested into the city in other ways. I think there’s value in that.
Ford: No muni system has ever turned a profit without subsidies. And, by the model, that loss to munis is revenues and not profits (they are assumed to be equal). Singer: Those are important critiques, George.
But before we start criticizing Kyle’s paper, I have one last question on the nuts and bolts: Is the £1 billion loss in consumer welfare an annual loss to consumers in the 30 states, or is it the loss over 20 years? I’m trying to figure out if the loss is economically significant. £1 billion spread over 20 years across 30 states would be very small. Seamans: LOL about Hal saying £1 billion is small.
I guess if we are increasing deficit by £1 trillion over 10 years, what’s an extra £1 billion? Wilson: That £1 billion consumer welfare loss is indeed the loss over 20 years. It’s a small number because we’re only talking about roughly two percent of all markets.
And even then, the gains to consumers come in the form of having access to a new technology, fiber internet. Most of these consumers would still have broadband access regardless, just not as high of quality. Ford: This “having access” issue is a big deal.
The muni debate conflates the importance of the difference between “nothing and something,” on the one hand, and between “something and something better,” on the other. Singer: I was expecting the consumer welfare harms to be much bigger. After all, with muni entry, we are moving from one to two high-speed broadband providers in an underserved area (ignoring momentarily the impact of mobile broadband).
I would expect a big price effect, resulting in a big welfare lift. I was surprised by the size. Wilson: That’s a fair point, Hal.
I don’t include pricing decisions in the model, so I explicitly hold price fixed regardless of competition. This does seem consistent with the way private firms set prices (at a national level), but it’s possible I’m understating the gains to consumer surplus. Singer: George, I understand you to have a problem with the way Kyle estimated the change in social welfare.
What’s your beef? Ford: To be clear, I really liked the paper. I view it as more of technical exercise than a policy-relevant paper, however.
Three assumptions are problematic: (1) The assumption that the muni’s revenue is equal to its profit (or producer surplus); (2) The model assumes the market is for a standalone broadband product, which may not exist; and as noted above, (3) No muni system has ever turned a profit, of which I am aware. Wilson: Thanks, George. It does get technical, but I hope that it can be a bit of both.
Ford: It would be interesting if the simulation could be modified to match a more plausible scenario. Say, impose the constraint that the market is in private equilibrium so that no firm can enter profitably. Perhaps the constraint has the bonus of permitting the simulation of marginal cost.
In this scenario, the muni may enter by assigning positive weight to consumer surplus, which private firms do not. The muni system may receive a transfer to cover the inevitable losses, and these subsidies are of course costly to collect. What is the welfare result in this situation?
I’m willing to bet it will be negative, and also believable. Wilson: The estimation actually does exactly what you’ve described in terms of imposing that the market be in an equilibrium where no one else finds it optimal to enter. In the simulation where munis are banned, profitability of private firms forms the basis for the equilibrium.
When munis are able to enter, the equilibrium condition is that they only enter if their profits plus (some of) consumer surplus outweigh the costs. Ford: I don’t see how welfare can rise. Consumer surplus is smaller than the loss to private providers.
So the welfare gain is just profits to the muni provider. But it seems the muni must lose money if it requires consumer surplus to enter. Singer: George, can you elaborate on how the first assumption (revenue equals profit) could cause Kyle’s model to overstate the social harm of a ban on muni broadband?
Ford: Revenue equal to profit will overstate the benefits of entry (though it will also reduce the private profit effect). The going-in presumption has to be that private entry isn’t feasible, otherwise it would occur (with perhaps some random variation). Thus, the welfare effects are likely to be negative.
It seems to me the model is a bit too “free” is its treatment of muni entry. Wilson: I do assume that revenue equals profits on a per-period basis. That is, that there are no ongoing costs.
But I also calculate profits net of the large upfront cost of installing the network. Singer: Also, George, explain to our readers how your second point–assuming a single revenue stream from standalone broadband–might bias the results. By not allowing for other revenue streams (video and voice), how would you expect Kyle’s model to turn out differently?
Ford: Comcast’s average revenue is £150 per sub per month. Standalone broadband is £60 per sub per month. Comcast isn’t in the standalone broadband business.
The whole model is based on that market definition. Singer: Fair enough, but how would that bias the results? Ford: It’s so far from reality.
How that biases the results is unknown. Looking at network availability is a different question, and I suspect the broadband prices Kyle uses are correlated with triple-play prices and that may be helpful statistically. Wilson: It’s true that the prices I used are for standalone broadband, not triple play, and this may not reflect the prices that many consumers are actually considering.
Ford: That’s important, and why I say it’s more of a technical exercise. A few caveats in the paper might keep the results from being used in an improper way, which has probably already occurred. Singer: Let me drill down on this revenue-stream assumption.
Kyle, your model assumes that telcos and munis are looking at the same revenue streams from fiber deployment. But munis look at basic transport, which has razor thin margins, and most don’t want to get into programming for TV. They get some cost savings by self-providing as opposed to buying their transport from an operator.
So the muni has the cost of build and operations, reduced by the revenue from the transport and the cost savings as its calculus. In contrast, telcos (especially those aligned with wireless), look at transport, wireless backhaul and reduced maintenance (fiber versus copper), and video programming to build their business case. They spread the cost over all those revenue streams.
Thus, if there isn’t a compelling business case for telcos, isn’t it reasonable that there likely isn’t a sustainable business case for the muni with its fewer revenue streams and zero starting market share? Wilson: That’s right. I do (perhaps unrealistically) assume that both private firms and munis stand to earn the same revenues.
But interestingly, the zero-starting market share you point out provides a reason why municipalities might still have a viable business case despite fewer revenue sources than private providers. For the municipality, every customer represents a new revenue stream that they can weigh against the costs of fiber deployment. But for an existing private telco, most of its would-be fiber subscribers are also its existing DSL internet customers.
Because the private firm is cannibalizing its existing market share, investment in fiber may not generate as much new revenue. Singer: As mentioned above, Kyle seems to limit his crowding out effect to the area where the muni network is deployed. But allowing muni broadband in one area of a state potentially depresses private investment in other areas, as the potential of a state-backed actor includes areas where the muni does not deploy.
For example, where a muni in such a state actually deploys, there could be a “bandwagon effect” where other munis are encouraged by their residents, and by the competition to have businesses locate in their jurisdiction to deploy. Is that a fair concern, and if so, can the crowding out effect be bigger than you estimated? Wilson: It’s definitely possible that this “bandwagon effect” exists.
Whether or not this leads to more crowding out, I think, hinges on whether private providers were likely to make investments in those areas but were deterred by the deployment of the “bandwagon municipalities.” In any case, I wouldn’t say that this means that I’m understating the extent of crowd out. But rather, it raises the question of whether we should attribute the crowd out to the first market or the “bandwagon” market. Singer: There’s another muni-broadband study making headlines this week–sorry for the competition, Kyle–and it’s from Harvard’s Berkman Klein Center.
The Berkman study concludes that muni broadband was cheaper than a comparable offering from the private sector in the same geographic market. Before offering our nits, I’ll note as an interesting aside that while Kyle finds that “a noteworthy feature of this [pricing] data is that for a given plan, a firm prices it identically in all markets,” the Berkman study suggests that ISPs are engaging in rampant price discrimination across markets for the same product. What could explain that difference?
Wilson: The Berkman authors only find evidence of price discrimination across markets for Comcast, and only in their introductory rates and contract terms. This feature doesn’t show up in my data because I only have data on providers’ regular rates, not their teaser rates. The fact that price discrimination across markets by ISPs isn’t more common is honestly very surprising to me.
Ford: The discrimination point is interesting. Charter charges the same price pretty much everywhere. The munis it competes against vary their prices around Charter’s uniform prices.
So you can make the case that muni undercuts a uniform price across markets. Seamans: The price dispersion is an interesting finding from the Berkman study. I do like the authors’ suggestion that more could be done (perhaps by FCC) to make prices more transparent to consumers.
Ford: The FCC won’t do it. They don’t want to aid collusion. Seamans: Agreed that is the danger of making price more transparent.
There always are tradeoffs! But it would be great if things were more transparent for consumers. Ford: The Berkman data is also two years old now.
I’ve looked at a number of markets this week and the prices are almost identical, as you would expect under the law of one price. Singer: Let’s talk about the straight up comparisons in public to private prices. The Berkman study potentially ignored whether that lower muni price was sustainable, and ignored the tax-cost implications of shortfall.
Just because muni prices are lower doesn’t mean they are sustainable, and if they aren’t, it’s the taxpayers who wind up paying. Taxpayers also pay taxes to cover the muni’s increased debt and the higher interest rate associated with that higher (and riskier) debt. Query whether these considerations should be added to the muni’s price before making any comparisons.
Seamans: There is lots of speculation in what you just wrote. Ford: Muni prices aren’t lower. How could they be?
How could a private broadband provider survive charging £1,200 for something the muni sells for £600? That’s what the Berkman study concludes. In Lafayette, the actual price difference between muni and private sector is about £0.30 per month.
I called the company listed as charging £600 more per year in Lafayette. It doesn’t even offer service in Lafayette. Seamans: But the real question is whether the low muni price is sustainable.
The chair of the National Broadband Network (NBN), Bill Morrow, said in October that the NBN was losing money on its average broadband connection, and that the NBN “might need to be shielded from competition from ultra-fast 5G mobile networks if it was ever going to make a profit, perhaps through a levy on wireless internet.” Closer to home, the mayor of Seattle concluded, after studying the economics of muni broadband, that financing a muni network “would result in one of the largest tax increases in the city.” And a recent study by Christopher Yoo at Penn listed U.S. cities that defaulted on the debt they issued to fund muni networks, including Burlington, Vermont, and Monticello, Minnesota.” Ford: More generally, muni broadband systems, on average, lose money and are heavily cross subsidized, which further complicate price comparisons. For instance, the Berkman study says that the utility in Opelika-Alabama charges £139 per month less than the cable system (for an unlike service by the way).
But the utility increased its electric rates in 2015 by £66 a year per account to cover losses from the broadband network, which is about a £200 per year subsidy per broadband line since only one-third of Opelikans buy communications services from the city. In all, the Opelika system loses about £900 per year, per communications customer. A £3 million loss on just over 3,000 customers.
Singer: Rob or Kyle? Any thoughts on whether these straight up comparisons are economically valid in light of the subsidy or taxpayer risk factors? Ford: I was looking a Clarkesville’s finances this week.
No attempt to hide the cross subsidy. The debt was put on electric side and the city increased electric rates 5 percent to cover the losses. Wilson: I think both comparisons are fair game.
If the question is: who offers cheaper internet service?, then the authors have answered that. But we might also want to ask: are low prices the end all be all? Or are there other burdens that make those low prices possible?
Seamans: I’d like to see a more in depth study that factors these in. But the straight-up comparisons are useful too. Ford: The law of one price.
If the prices are different, then the services are different. Seamans: Or the consumer is getting screwed. Ford: Here’s another nit: The Berkman authors set out to find the lowest priced offering that satisfies the FCC’s definition of broadband.
But there is no market for such a good. These firms are not in the business of selling the product being analyzed. In fact, whether muni or private, it’s a challenge to find an “Internet only” service on the provider’s websites.
Let’s look at Comcast, for instance. The report lists Comcast’s average price at about £60 per month for broadband. The company’s average revenue per customer is about £150 per month.
Very few of Comcast’s customers buy broadband only. Most are triple play. Thus, Comcast is not in the business the report presumes it is.
Broadband providers sells bundles of data, voice, and video. So is the report even focusing on a relevant price? Wilson: I disagree that there’s no market for standalone broadband.
More and more households, especially younger demographics, are moving toward internet only subscriptions as video streaming and mobile phones are replacing cable TV and home phones. Private broadband providers may not want to be in the business of selling internet-only service, but perhaps they should be. Seamans: I buy standalone broadband!
Singer: You academics just aren’t getting paid enough! Seamans: Just you old guys who buy everything … Ford: Academic economists are, after all, the mean.
Seamans: We are mean. Ford: So true. And not even charming when being mean.
I think there’s a future market for standalone broadband. But that wasn’t the case in December 2015, the year for which Berkman assembled its pricing data. Wilson: I do think, though, that Berkman’s price comparisons are a bit misleading.
The authors compare the lowest-speed plan that exceeds the FCC’s broadband definition of 25 Mbps. But oftentimes a municipal provider offers an inexpensive plan with exactly 25 Mbps while the private provider offers a more expensive plan with 40 or 60 Mbps. It’s true that the municipal price is cheaper, but when normalized by speed and measured in, say, dollars per Mbps, their prices seem to be very similar.
Singer: Let’s conclude with some policy implications. Before we get to the issue of government provisioning, isn’t there some intermediate steps we could take to spur private broadband investment? For example, we could expand demand-based user subsidies to support poor families who can’t afford high-speed broadband.
Or we could initiate supplier-based subsidies to make uneconomic areas profitable to service–for example, we could hold a reverse auction and give telcos a subsidy for purchasing and deploying fiber. Why are those intermediate interventions not getting any attention? Seamans: This happened under the Obama Administration when the FCC modernized Lifeline to allow those customer subsidies to be used for broadband.
It would be great to see a study of the effectiveness of this change. This would be a great study for the FCC’s proposed new Office of Economic Analysis to undertake. Ford: Before it is disbanded under the next Democratic chairman.
Or Republican one. I’m an equal opportunity skeptic on the relevance of economics in modern public policy. Seamans: Also on the supply side, as I wrote about recently in Forbes, the Trump Administration is undertaking some steps to spur rural broadband investment.
Wilson: And the Connect America Fund does aim to subsidize entry into unserved areas and bring them up to a speed of 10 Mbps. But as far as I know, there’s no systematic subsidy program that aims to encourage cutting-edge speeds. As far as using targeted subsidies to encourage private firms to deploy fiber, I think this could be a viable alternative to municipal broadband, but it does come with its own challenges.
You only want to subsidize private investment where it would otherwise not occur. And even then, you only want to subsidize costs by just enough to induce investment. How do you avoid ISPs holding out for subsidies in all markets?
How do you avoid over-paying in markets you do want to subsidize? Ford: Targeted subsidies is one answer. It has theoretical support.
Singer: We’ve been talking about the costs and benefits of muni broadband. But the policy question of the day is whether states should ban munis from engaging in broadband activity. I see the net benefits of muni broadband and the state-issued bans as two different but related concepts.
Any concluding thoughts on the wisdom of a statewide ban? Seamans: I don’t like the bans. I think they benefit incumbent companies, but not consumers.
I think the threat of entry from a municipal electric utility provides a nice competitive spark to incumbents. Ford: Letting a city use its tax base or captive ratepayers to subsidize public entry is not good. It reduces welfare.
The threat of uneconomic entry does not lead to efficient decisions, as Rob’s paper confirms. Subsidized entry in markets served by private actors is anticompetitive and welfare reducing. Wilson: I’m with Rob and against bans.
Municipal provision might not always make sense, but I think bans make things worse. Seamans: And I’m with Kyle that municipal provision doesn’t always make sense. But again the threat of entry is a useful tool.
Ford: The threat of uneconomic entry does not lead to efficient decisions. Private firms are only subsidized to be the sole entrant. Subsidies (no longer) subsidize multiple entrants.
In fact, the FCC was explicit that subsidizing multiple entrants is bad policy. Singer: Even George would tolerate muni broadband in unserved areas. So if the bans barred that activity even in those areas, then I suspect George would be against a statewide ban?
Ford: It’s the truly unserved markets where the bans are problematic, if poorly drafted. Cross subsidy should be banned only when there is private provision. You have to cross subsidize in unserved markets.
Singer: Let me try a final motivation for the ban. Cities compete for football teams or large firms (like Amazon) and are so desperate at times, they overpay, leaving taxpayers on the hook. Because the state and county need to step in if the muni fails (like Detroit), those entities have a real stake in not seeing their cities do risky things.
Can fiscal responsibility be an alternative hypothesis (to cable lobbying) for the “no muni broadband” laws? Seamans: I don’t like the NFL analogy at all. The NFL has strategically kept a limit on the number of teams.
There are fewer teams than there are cities that want an NFL team, which shifts the bargaining power entirely to the NFL. For the analogy to work, you need there to be a limited number of broadband systems in the US, and for cities to compete over ability to get one of these rare systems. That’s not the case, there’s no limit on availability of a broadband systems for a city.
Ford: Fiscal responsibility is a justification for the laws, and a valid one. Singer: We’re going to have to end it there. Thanks so much for joining.
And best to luck to Kyle in getting his paper published!
- ^ announce an executive action (www.forbes.com)
- ^ bipartisan legislation to “dig once” (www.broadcastingcable.com)
- ^ introduced legislation supporting municipal-provided broadband (thehill.com)
- ^ 21 states (www.baller.com)
- ^ a federal court has once again ruled (www.opn.ca6.uscourts.gov)
- ^ Robert Seamans (www.stern.nyu.edu)
- ^ Fighting City Hall: Entry Deterrence and Technology Upgrades in Cable TV Markets (pubsonline.informs.org)
- ^ George Ford (www.phoenix-center.org)
- ^ The Impact of Government-Owned Broadband Networks on Private Investment and Consumer Welfare (sglf.org)
- ^ Does Public Competition Crowd Out Private Investment?
Evidence from Municipal Provision of Internet Access.(papers.ssrn.com)
- ^ Hal Singer (haljsinger.wordpress.com)
- ^ one lawsuit away (www.bna.com)
- ^ National Broadband Map (www.broadbandmap.gov)
- ^ State Broadband Initiative (www2.ntia.doc.gov)
- ^ Christopher Timmins has a paper (dukespace.lib.duke.edu)
- ^ as I show in my paper (pubsonline.informs.org)
- ^ muni-broadband study (cyber.harvard.edu)
- ^ 1997 RAND paper (www.jstor.org)
- ^ Australia’s experience (www.smh.com.au)
- ^ concluded (www.seattletimes.com)
- ^ study (www.law.upenn.edu)
- ^ utility increased its electric rates in 2015 by £66 (phoenix-center.org)
- ^ moving toward internet only subscriptions (techcrunch.com)
- ^ FCC modernized Lifeline (apps.fcc.gov)
- ^ FCC’s proposed new Office of Economic Analysis (www.forbes.com)
- ^ wrote about recently in Forbes (www.forbes.com)
If you missed out on BT’s last round of great deals on broadband packages, never fear – the firm has just announced its next wave of money-saving offers. This new wave of bundles follows in the footsteps of the last by offering competitive monthly prices, with bonuses in the form of low set-up costs and cashback in the form of BT Rewards Cards – prepaid Mastercards of varying values depending on which package you go for, but that you will have to claim yourself once your broadband is up and running. Once again, these are limited-time offers that only run until the end of the day on 25 January, so best switch or upgrade in the next week if you see something that catches your eye here.
All packages here also include free weekend calls within the UK from your landline, all are 18-month contracts, and they all require a GBP9.99 delivery charge in order to ship your hardware – a small price to pay considering BT’s Smart Hub is one of the very best provided by any major supplier. First up is the regular Unlimited Broadband option, which offers a connection good for speeds of up to 17Mbps. If you don’t make heavy enough use of your online services for fibre to be a consideration, this standard package will serve your needs just fine, and comes in at just GBP23.99 a month.
As part of the current promotion, you’ll also get a GBP50 reward card to spend on whatever you fancy.
Add BT TV to your package from GBP3.50 a month
Moving on, we make the leap to fibre and the noticeably faster speeds these connections can offer. BT’s Infinity 1 package is the first of two here, boasting top speeds of 52Mbps and perfect for households with many connected devices and where streaming and downloading are everyday activities. This one’s GBP29.99 a month and while there is a GBP20 connection fee, that’s offset nicely by an GBP80 reward card.
Infinity 2 is BT’s quickest widely available service, with scorching speeds up to 76Mbps. As well as the fastest speeds you’re going to get from standard OpenReach broadband, this one also comes with the most generous reward card – a cool GBP100. The offer price here is GBP39.99, perhaps not the cheapest monthly price for a 76Mbps line but still pretty competitive when you factor in the reward card, great hardware, and truly unlimited service with no traffic management.
If you do want to go even faster, BT’s new ultrafast options might be what you’re looking for, although coverage isn’t great at this time. It’ll roll out to more locations over time, of course, but right now, Infinity 2 is probably as fast as most people will be able to go. But it’s fine – that’s plenty fast enough for all but the most hardcore data users.
If you want to add a little more to your bundle, there are a number of options. Unlimited evening calls can be added to the phone package for just GBP4 a month, or those who love a good natter can get unlimited anytime calls for the reduced price of GBP7.50 a month. You can add TV packages, too – a basic TV bundle with BT Sport is a mere GBP3.50 a month extra (plus a small activation fee), the larger Entertainment package will run you up an extra tenner a month, and TV nuts can get the works with the Max package for GBP15.
Oh, and don’t forget that BT broadband customers get GBP5 off BT Mobile monthly prices too, so you could save even more if you’re looking for a new mobile tariff as well.
Why do we need your postcode?
Once you enter your postcode, Cable.co.uk will perform a live lookup and check all the available providers in your area. This ensures you receive accurate information on the availability of providers and packages in your area. Your information is safe with us.
We won’t share your postcode with anyone.