We Shall Overwhelm
Money as overwhelmingly loud speech. A fascinating history of the genesis and strategies of movements to detax the rich in the United States.
Money as overwhelmingly loud speech. A fascinating history of the genesis and strategies of movements to detax the rich in the United States.
Eulogies to Nokia have been everywhere over the last few days. However, the most nuanced, informative and engaging account I know of is still the remarkable extended piece published in Helsingen Sanomat in October 2010.
The four part piece is structured through a series of anonymous interviews with a diverse range of Nokia employees. Many possible roots of Nokia’s diminution are traced, however, the reason identified as fundamental - and one I’ve not seen mentioned in this week’s rushed obituaries - is a decision taken by former CEO and Chairman Jorma Ollila in 2003 to radically transform Nokia’s internal structure. Instead of its original two units, Mobile Phones and Network Equipment, Ollila created three separate mobile phone divisions, Mobile Phones, Multimedia, and Enterprise Solutions. The aim was to achieve success by planting Darwinian competition within the company, instead, the results — according to the Nokia executives interviewed — appear to have been disastrous.
What emerged was a leadership vacuum. Right there and then the seed of gradual internal decay was planted. The various units began to compete tooth and nail with each other for the same resources and the same markets. And above them there were not the necessary strong decision-making mechanisms for control of the product assortment. I mean the sort of leadership that would have looked at the big picture and held up a hand and said: ‘Hey, just a second now, it doesn’t make any sort of sense to manufacture overlapping products like this’…People tried simply to respond to the challenges and needs of all the different product lines. There was not enough time and money for work at the long-haul end. For example for things like updating and upgrading the operating system software
From being a unified firm with an eye on the future, Nokia appears to have broken down into series of antagonistic parts, fighting each other jealously to achieve short term results. It was left completely unprepared for the tumultuous arrival of the iPhone three years later. And, perhaps most shockingly, Nokia was simply *unable to respond meaningfully for years * afterwards.
Ps The nostalgic photoset above shows (anticlockwise from top left):
Goodbye Nokia, old friend.
A great review in the NYRB of Stung! On Jellyfish Blooms and the Future of the Ocean by Dr Lisa-ann Gershwin, which appears to be a fascinating account of jellyfish and their seemingly inexorable rise.
I was particularly struck by the argument that:
…it has taken a mighty effort by other living creatures to hold jellyfish down. An important part of that effort has involved the maintenance of complex ecosystems, with their abundant predators and competitors of jellyfish. It’s no accident that prodigious jellyfish blooms have occurred in areas like the Black Sea and off South Africa, where anchovies once swarmed. Overfishing anchovies, which compete with jellyfish for food, has doubtless helped them take over. That alone might not have been enough to allow the jellyfish to gain the march on us, but we’ve overfished virtually every resource in the oceans, causing the outright collapse of many ecosystems, thus opening vast new resources to the jellyfish.
In environments that have been strained by human activity, efficiently feeding and reproducing jellyfish have quickly expanded in number. In addition their ability to feed on the eggs and young of their predators suggests that the balance may not easily tip back. Gershwin suggests that the future may be one of ‘jellification’ as jellyfish are able to respond positively to ecosystems altered directly by humans as well as the effects of human-induced climate change, such as ocean acidification.
The underlying hypothesis, that depletion of more complex forms of life can tip ecosystems into favouring previously more abundant but simpler organisms, is intriguing. It would be interesting to see whether there are parallels in other environments in which humans have removed large numbers of organisms occupying higher trophic levels, and whether similar ‘tipping points’ have been observed.
However, the hypothesis reproduced above does raise questions. First, how helpful is it to portray the interaction between more complex lifeforms and jellyfish as a ‘mighty effort’? Is this simply an illusion from the fact that the fish we consume occupy higher trophic levels than jellyfish? And second, does their ability to reliably bloom provide a similar growth opportunity for predators of jellyfish? For example, this is being observed in the Black Sea, which is one of Dr Gershwin’s case studies.
In any case, from the perspective of human welfare, the jellyfish renaissance would appear to be almost wholly negative. Unless we quickly develop tastes for jellyfish alongside other unusual delicacies.
The world’s first car featuring an option for integrated 4G LTE connectivity, the Audi S3 Sportback, has just begun shipping to customers in Europe. Audi also hopes to enable the option for US customers next year.
As Kevin Fitchard of GigaOM notes:
What’s particularly interesting is how signing up for in-vehicle 4G connected car service will be different in the U.S. and Europe, mimicking the way those different regions buy phone service. In Europe, 4G Audi owners can use any carrier’s network. They just need to insert a data SIM card into the dash.
In the U.S. though, Audi connectivity will most likely be tied to a specific operator, Stertz said, just like most of our smartphones are optimized for specific carriers. That makes some sense, given the complexity of U.S. LTE networks.
While in Europe most carriers are deploying their networks on three common bands, in the U.S. every operator is using a different LTE frequency. Just like Apple and Samsung make separate versions of the iPhone and Galaxy for different carriers, Audi may need to supply different modules for those different networks.
In Europe, most operators are using the 800 MHz, 1800 MHz and/or 2.6 GHz bands for LTE, whereas in the US there are three (mutually incompatible) versions at 700 MHz, and further bands at 800 MHz, 850 MHz, 1700 MHz, 1900 MHz, 2.3 GHz and 2.5 GHz.
Audi’s decision, therefore, reflects the economics faced by LTE chipset manufacturers. It appears possible to cost-effectively design a radio that is able to address the three LTE frequencies that are used in Europe. However, it appears not to be possible to cost-effectively design a module that can address all the LTE frequencies being used in the US, a market of comparable scale. This will reduce the ability of consumers to switch between carriers, and is likely to soften competition and ultimately harm end users through higher prices.
This serves as a stark demonstration of the limited substitutability of bands of spectrum, even using a common technology platform like LTE, which I touched upon in an earlier post.
This is part of a series of posts that make the case for spectrum access regimes based on systems of rules rather than systems of licences.
You can find an introduction to the topic and an index of all posts here.
In 1959 Ronald Coase argued he saw “no reason that there should not be private property rights in spectrum” based around frequency ownership. According to Coase, this would allow spectrum to be traded to its highest valued and most valuable uses, maximising the overall economic gain.
The view of spectrum as best managed through property rights has become the prevailing wisdom in a number of countries. Auctioning of tradable licences was pioneered in New Zealand from the late 1980s and then taken up by the US, Australia and the UK in the 1990s. In 2005 the UK’s Ofcom stated its aim for 71% of frequencies to be tradable by 2010. An EU decision of 2002 proposed to introduce continent-wide tradable licence systems by 2010.
These policy moves were justified by the benefits that tradable licences would bring. For example, Ofcom claimed that trading would result in:
lower prices for the most profitable and popular wireless services as wider availability of spectrum increases competition and supply; greater choice as alternative suppliers enter the market by acquiring rights to use spectrum; and innovation as entrepreneurs acquire rights to use spectrum and offer new services.
However, in practice, tradable spectrum licences have not nearly met these expectations.
First, activity has been slow and volumes low:
In 2006, Europe’s first pure spectrum trade happened in the UK, fourteen months after trading was allowed .
In 2007, ACMA, the Australian regulator, noted that “Spectrum trading has occurred in low volumes in Australia and New Zealand”.
In 2011, Peter Stanforth of Spectrum Bridge – one of the US’s biggest spectrum exchanges – identified lack of education, fear of interference, lack of incentives against hoarding, and high transactions costs as the culprits in a presentation entitled “Why Haven’t Secondary Markets Been Successful?”
In 2012, the economic consultancy Oxera wrote “while spectrum trading has been possible in a number of countries for some time (such as the UK, New Zealand and Australia), there have been few substantial trades” .
Second, and most disappointingly, tradable spectrum rights have not resulted in much, if any, meaningful innovation. Most transfers have been due to changes in company ownership and the few that have been ‘pure’ spectrum trades have often been due to larger firms buying up licences from smaller ones, quite the opposite of Ofcom’s hopes for new and innovative entry.
In 2006, Professor Martin Cave, one of the chief architects of the UK’s move to a regime of tradable licences, suggested that the success of the regime should be judged over a 3-5 year time frame. In 2013, we can largely conclude that the experiment in tradable spectrum property rights has failed.
In his 1959 paper, Coase drew close analogies between spectrum and land. However, he overlooked the importance of a key dissimilarity, degrees of substitutability. In the case of land, two adjacent parcels will have similar attributes for the building of houses, workplaces or farms. This high degree of substitutability is a characteristic found in the majority of well functioning markets1. However, although two adjacent frequency bands of spectrum are likely to have similar physical characteristics, they are only brought into use through highly specific infrastructure, designed to operate on specific frequencies. Advancing technology has compounded Coase’s error by increasing the cost of the required infrastructure, and so effectively decreasing the substitutability between the vast majority of frequency bands2.
Coase’s analogy between spectrum and land has had far-reaching consequences in spectrum policy that go beyond the experiments with tradable property rights. I’ll come back to explore some of these in future posts.
Even in the market for fine art where we could accept that each piece is a unique expression there is substitutability — I might be equally content with a work by Turner or by Klimt hanging on my wall. ↩
Those few bands that are accessible can be regarded as substitutable. For example, channels within the narrow range of an FM tuner might all have a similar value and be accessible using the same equipment. ↩
This is the first in a series of posts that make the case for spectrum access regimes based on systems of rules rather than systems of licences. These posts will use some material from my response to a consultation by the EU’s Radio Spectrum Policy Group (RSPG) on the issue of Licensed Shared Access (LSA).
This post provides an overview of the arguments. Later posts will examine specific issues — such as Quality of Service (QoS), innovation and the economics of spectrum access, making markets in spectrum and others — in greater detail
Wireless devices have become indispensable, and as their importance has grown so have concerns about a ‘spectrum shortage’ and an impending ‘spectrum crunch’. However, multiple pieces of research make clear that the vast majority of frequencies in most places and at most times are unused. The spectrum is full of ‘white spaces’. In fact, the perceived ‘spectrum shortage’ is an artefact of the prevailing mode of spectrum management, of granting exclusive-use licences to particular frequencies.
Dynamic Spectrum Access can revolutionise spectrum access but the system of governance chosen will be crucial
Dynamic Spectrum Access (DSA) technologies, such as white space databases and spectrum sensing, can allow new uses to make use of the unused white spaces, without interfering with existing users nor necessitating costly and time-consuming spectrum clearance. However, the potential impact of DSA will depend critically on the type of access regime chosen by regulators to govern its use. Each specific implementation will be one of two fundamental types:
Licence-based regime — in which a DSA user will need to obtain a licence through a regulatory award or market-based negotiation; or,
Rule-based regime — in which the satisfaction of certain conditions — such as limited transmit power levels, checking with an online database and/or payment of an access fee — will allow a DSA user access
Licence-based DSA regimes will create exclusive rights to spectrum capacity, in much the same way as the existing licensed bands supporting applications like mobile telephony and broadcast television and radio. For example, this is at the core of the concept of LSA, in which regulators will identify ‘beneficial sharing opportunities’ and award rights to use DSA technologies in these bands to a limited number of entities. Increasingly, regulators have allowed the trading of spectrum licences. This idea is supported by the influential, but dated, economic characterisation of spectrum as a simple resource, rival in consumption, whose efficient allocation amongst users is best assured through systems of tradable property rights.
Rule-based DSA regimes will permit access to spectrum bands in a similar way to the current licence-exempt(unlicensed) and light-licensed regimes that support technologies such as Wi-Fi, Bluetooth and NFC. Rule-based DSA is already being used in the US and Finland to allow access to the TV White Spaces (TVWS) spectrum. The utility of this mode of access is underpinned by newer economic theories that posit productive activities using certain partially-rival resources, or infrastructures, can display increasing returns to participation. Therefore, rule-based systems — which provide a guarantee of access — can be used to encourage greater participation and thus maximise the total economic benefits.
The experiment in tradable property rights has largely failed. The experiment in licence-exempt access has been an unexpected triumph
In practice, systems of tradable licences have not nearly fulfilled the hopes of their proponents. Activity in these markets has been slow and volumes low, with the large majority of ‘trades’ simply reflecting changes in company ownership. Perhaps most disappointingly, tradable spectrum rights have not led to the innovation that was promised by their proponents. LSA is an extension of the idea of exclusive-use property rights to spectrum and it is unclear how it would avoid these same failings.
By contrast, the successes of the narrow licence-exempt bands demonstrate the power of rule-based access to spectrum. Within twenty years of widespread availability licence-exempt technologies now account for:
The majority of innovation in wireless communications — licence-exempt connectivity is near ubiquitous in smartphones, tablets and PCs and is rapidly growing across a number of categories of consumer and industrial goods. The licence-exempt bands are also home to a diverse range of open standards that act as a platform for further innovation.
The majority of wireless devices — in 2013 fewer than 2.5 billion devices will be sold that are enabled with licensed connectivity. However, over 5 billion devices1 will be produced that use licence-exempt communication technologies2. This disparity is set to increase.
The majority of Internet data traffic delivered to end-users — Wi-Fi carries 69% of the total traffic generated by smartphones and tablets and 57% of total traffic generated by PCs and laptops3.
Indeed the success of the LE bands has been the most surprising and consequential regulatory action in the previous 15 years of spectrum management. The economic benefits from licence-exempt technologies are substantial, widely dispersed, and may exceed $270 billion per annum globally4.
The two great spectrum challenges of the future are best suited to spectrum that enables rule-based access
The future of spectrum usage will be marked by an increasing diversity of uses — especially due to the emergence of the Internet of Things — with demand growing most strongly on licence-exempt networks. Specifically:
Of the 50 billion wireless devices connected to the Internet of Things in 2020, the overwhelming majority — over 95% — are likely to use licence-exempt technologies. The remainder — less than 5% — will be served by cellular systems. Even using the highly conservative assumption that each licence-exempt node generates only a tenth of the value of each licensed node, over 65% of the value of the Internet of Things would come from licence-exempt devices — supporting up to $6.5 — 9.8 trillion of global GDP by 20305.
The traffic over licensed networks is likely to grow by 10.5 EB per month between 2013 and 2017. The volume over licence-exempt networks is likely to grow by over 40 EB per month in the same period6.
In such a scenario the opportunity costs of dedicating exclusive spectrum rights are likely to become progressively more difficult to justify — especially as advances in technology permit greater and more reliable spectrum sharing. Licence-based DSA, such as LSA, harks back to an era of few users of spectrum and few networks, and is likely to simply increase the exclusive holdings of existing operators. In addition, its overuse risks repeating the mistakes of exclusive allocation that have led to today’s artificial spectrum scarcity. Instead, regulators should grasp the possibilities offered by making rule-based access the default mode of implementing DSA. This can create bands of spectrum that can respond quickly to market conditions, allow for continuous technology upgrades, enable networks with finely tailored speeds, capacities and Qualities of Service (QoS) and effectively multiplex the countless wireless applications of the future.
The licence-exempt total does not include simple RFID circuits ↩
On Friday, I submitted this response to the European Union’s Radio Spectrum Policy Group (RSPG) consultation on the use of Licensed Shared Access (LSA).
This is an important process that will affect the future of the radio spectrum and wireless communication in Europe, and influence the direction in other parts of the world. The decisions taken could mean that access to the spectrum increasingly becomes the preserve of a few large companies, or, conversely, is opened up, permitting access to the whole range of innovators and users.
Over the coming days I will expand upon material in my response to explain the importance of this issue and why LSA is a fundamentally regressive idea - better suited to the past of few wireless uses and users. Instead, alternatives that encourage widespread access will be better able to generate the innovation and competition needed to meet the challenges of the future.