[This post is a translation of this other post of mine]
As we saw in a previous post, one of the main examples used by Mariana Mazzucato to validate her arguments is the case of the iPhone. Let us remember, Mazzucato’s thesis is that most of modern technology is due to directed governmental research efforts, not only in basic science, but also in technological commercialisation, direct investment, loans, etc. These kinds of policies make up what she calls The Entrepreneurial State. We must make clear the difference between this concept from another, more reasonable one, that is of a State that merely invests in basic science to correct market failures. Such thing is what is usually accepted in the economic mainstream, where Mazzucato is not, given she is precisely arguing against entrenched conceptions of the role of the State in innovation.
In the present article, I shall criticise her arguments, and question the evidence put forward to defend them. Like in previous articles, her own bibliography allows us to criticise her own points. We shall begin from a study of every single bibliographical reference she cites in the chapter to support every single one of her paragraphs, and complete this with additional references wherever it is necessary.
This critique limits itself to the relevant chapter of the book, in this case to chapter five, where the iPhone example is discussed. Some elements that are mentioned in the article are excluded: the case of computers and semiconductors in general, and GPS, as these deserve to be discussed separately. Mazzucato also says these also come mostly from the State, so I’m not going to criticise her arguments twice.
The chapter begins explaining that Apple’s history, and the Jobsian “stay hungry, stay foolish” philosophy, are somewhat mystified, as according to Mazzucato
Individual genius, attention to design, a love for play, and foolishness were no doubt important characteristics. But without the massive amount of public investment behind the computer and Internet revolutions, such attributes might have led only to the invention of a new toy – not to cutting-edge revolutionary products like the iPad and iPhone which have changed the way that people work and communicate.
While the products owe their beautiful design and slick integration to the genius of Jobs and his large team, nearly every state-of-the-art technology found in the iPod, iPhone and iPad is an often overlooked and ignored achievement of the research efforts and funding support of the government and military.
A discussion of Apple’s R&D expenditures follow. It is mentioned that is has been growing in quantity (it multiplied by 25 since 1999, and by more than 4 in the last 5 years), it has not represented a constant percentage of sales revenue: As time passes, Apple’s R&D spending represents an ever shrinking fraction of sales revenues, according to Mazzucato’s book. We can see that this tendency seems to have reverted in the graph linked in the previous link, however. The economist gives two possible interpretations to this: One, that maybe Apple uses her funds very efficiently. Or second, more plausibly according to her, citing analyst Horace Schmidt, that Apple is not really a company that develops new technologies, but is more of a company that integrates new technologies. To Schmidt’s analysis, Mazzucato adds: “technologies, in many cases finances by the State”, using this to link his section to the one that follows, where what the State has done for Apple is detailed.
These interventions are of three types:
- Direct investment in the initial funding and growth of Apple
- Access to technologies resulting from state research programmes, military initiatives, or carried out by public research institutions
- Fiscal, commercial, and/or technological policy to support Apple in hard times.
With respect to point one, Mazzucato says that Apple had many private investors: Don Valentine, Arthur Rock, Venrock, Mike Markkula… but that it also received money before its IPO in 1980 from Continental Illinois Venture Corp, which was, according to Mazzucato, a “Small Business Investment Company (SBIC), authorized by the Small Business Administration to invest in small businesses”. For the list of investors, the economist cites Rao and Scaruffi 2011 and for the state investment data, Slater 1983 and Audretsch 1995. I wasn’t able to access any of these sources, so I will use others. It is known that Markkula invested 250,000$ in 1977 and in Moritz 2009, we find that in 1978, Apple managed as much as 517,500$ in investments, all of that being before CIVC’s entry.
This CIVC, say Lazonick and Mazzucato 2013, would be an arm of Continental Illinois Bank who had gotten guaranteed loans from the US Government, and that it had invested 500,000$ in Apple in 1978. In this report from the San Francisco Federal Reserve, we can see that a SBIC is a private venture capital fund, that can have -but that does not necessarily have- some sort of State backing. They also have lessened regulatory requirements. Sources from that time confirm CIVC’s investment, but seem pessimistic about the SBIC concept.
A more plausible interpretation of this, given how SBICs work, is that this was not an action taken by an Entrepreneurial State, but that of a group of private investors who managed to get state benefits through the SBIC programme (the people from the Continental Illinois Bank). I am not the first in reaching this conclusion. Taking into account the existence of private investors, and the sales revenue and valuation of Apple back then, the importance of State guarantees for the loans to a private entity that in turn invested in Apple is framed into its historical context, and has its importance relativized. That is, Apple was steaming ahead before the involvement of the CIVC and given what we know, it is most reasonable to assume that it would have continued to do so hadn’t there been government involvement. Therefore, you can certainly say that Apple had government support in its initial stages, but not that it was crucial, or that we would have had no Apple without it.
What technologies, according to Mazzucato, are the ones that set the iPhone, iPod, and Ipad apart from others in the market, or that allow them to work?
As central technologies, microprocessors, RAM, hard drive storage, LCD screens, Li-pol and Li-ion batteries, digital signal processing (DSP) based in the Fast Fourier Transform (FFT), the internet, HTTP and HTML protocols and cellular networks.
As additional technologies that give these devices a touch of distinction: GPS, click wheel navigation, multitouch screens, and an speech based artificial intelligence assistant: SIRI.
Hard Drives (GMR)
Mazzucato begins with hard drives, that according to her enabled the iPod to dethrone Sony’s Walkman and Discman. These new hard drives were made possible thanks to the discovery that earned Albert Fert and Peter Grünberg the Physics Nobel Prize in 2007: Giant Magnetoresistence (GMR). The discovery was parallel, in 1988. Both Fert and Grünberg discovered GMR independently, that is.
The problem here is that the history of hard drives doesn’t begin or end with the discovery of GMR. The first disks that apply GMR were commercialised in 1997, and certainly the Toshiba disk that the first iPod had used GMR heads. But the step from discovery to manufacturing wasn’t trivial. Fert and Grünberg discovered on one side the GMR effect, but then IBM Research takes that research and applies it to hard drives as early as 1991, and from there to Toshiba it was another step, probably due to the sale of IBM and Hitachi technology to Toshiba, as mentioned in the previous link. For the case of the discovery itself, the structure is similar to what other discoveries follow: They begin from existing scientific results and take them to the next level. GMR was discovered after the decade long effort by thousands of scientists, it wasn’t just a happy idea that popped into existence. (Grünberg et al. 1989, Fert 2007)
Mazzucato tries to argue, via McCray 2009a, that the history of GMR and hard drives illustrates the role of the State not only in research (Most economists were already saying this), but also in the work necessary to convert scientific ideas into manufactured, commercially viable products. She also says that, even when the most important scientific development leading to the discovery of GMR was European, the US government also played an important role in commercialising the technology, arguing that Grünberg’s lab was affiliated with Argonne National Laboratory, and that he received Department of Energy grants. From there, companies developed the technology. This evidence does not seem to support the idea that the State helped in the transition from science to product. As we saw before, it was IBM who played that role.
According to McCray, the origin of the field of magnetoresistence can be traced back to the 1857 research of William “Lord Kelvin” Thomson, and that the modern understanding of the electron spin (also base of GMR) can be traced back to the 1920 work of Pauli, Dirac, and Goudsmit. Next, scientists in Bell Labs and IBM experimented with diverse techniques to produce artificial materials, developing technologies such as molecular beam epitaxy (MBE), to be able to, in turn, build nanostructures. MBE-based apparatuses were crucial in Grünberg and Fert’s research, according to McCray.
On the role of the State in commercialization, which is why Mazzucato was citing McCray, he says that GMR was first applied to weak magnetic field detection (by James M. Daughton, electrical engineer from IBM Research, who later founded his own company, Nonvolatile Electronics, in 1989. It was in IBM Research where high temperature superconductivity was also discovered, in 1986)
McCray confirms what we were saying: it was IBM the one who made the crucial step of going from research to commercial development:
Companies were eager to apply the nascent form of spintronics to broad er and more lucrative markets. Following the discovery of GMR, scientists at IBM’s Almaden laboratory near San Jose, California searched for other ways to apply it to commercial products. Unlike IBM’s Thomas J. Watson Research Center in Yorktown Heights, New York, which had a long history of research on semiconductor logic devices, the older Almaden lab had traditionally focused on magnetic information-storage technologies.
Unlike Fert and Grünberg, who built their samples with the more precise but slower molecular beam epitaxy technique, Parkin’s [IBM researcher] group adopted sputter-deposition equipment. The focus of the Almaden group wasn’t on basic science per se, but on making devices that could be readily manufactured. Parkin’s use of the quicker and cheaper sputtering technique made sense for IBM, a company with extensive experience in fabricating sputter deposited magnetic-storage media on an industrial scale. As one observer of Parkin’s research later recalled, the British scientist and his colleagues “simply engineered the [expletive deleted]” out of the underlying GMR discovery as they made and characterized over 30,000 different multilayer combinations
IBM eventually used the spintronics research of Parkin and his colleagues to redesign and improve a basic element in the company’s hard disk drives. The Wall Street Journal revealed the company’s innovation with a front-page story in November 1997. Based on the Almaden group’s exploitation of GMR, IBM’s new drives featured exquisitely sensitive magnetic-read heads. Able to store eight times more data than competitors’ equipment while remaining smaller in size, the redesigned read heads set the stage for a subsequent explosion in computer memory that, in turn, helped make it possible for music lovers to store gigabytes of music and other files on iPods and similar handheld gadgets. IBM licensed its GMR-based technology to other companies and, with in a few years, practically every computer hard drive included a read head based on IBM’s innovation. Firms like Seagate Technology and, for awhile, IBM reaped tremendous profits. IBM’s rapid commercialisation of Fert and Grünberg’s basic discovery introduced spintronics into a market worth billions of dollars annually. And while these ubiquitous tools were not explicitly marketed as nanotechnology, IBM researchers could boast that, be cause of their research, “everybody has [a] spintronics device on their desktop.”
Once products were already in the market, the State began to invest in them:
As products exploiting the GMR effect appeared on the market and more scientists began to do research in what would become known as spintronics, science managers from military laboratories and funding agencies began to take notice. At the same time, major newspapers presented spinbased electronics as a potential new industry paradigm. Journalists struck a tone that comported with a prevailing tendency to predict the end of Moore’s Law by hyping the new electronics technology as the “next big thing” for the industry.
DARPA invested in GMR in 1992,
One of those monitoring nascent GMR-based spintronics was Stuart Wolf. Trained as a physicist at Rutgers University, Wolf received his doctorate in 1969 and initially researched low-temperature superconductivity in metals.44 After three years at Case Western Reserve University, Wolf became a staff scientist at the Naval Research Laboratory, where his work on superconductors held interest for a number of defence applications. In 1993, Wolf accepted a new post as program manager at the Defense Advanced Research Projects Agency. DARPA, of course, was no stranger to supporting research for computer and semiconductor applications; for decades, it had funded major cutting-edge programs in areas such as molecular electronics, integrated-circuit design, and supercomputing.
In late 1992, as part of the transition to the new geopolitical environment, the Department of Defense initiated the Technology Reinvestment Project (TRP). Managed by DARPA, the TRP’s purpose was to build stronger links between the commercial and military sectors and help the United States reap a greater share of the anticipated “peace dividend.” The TRP also emerged at a time when established companies like IBM and Bell Labs were cutting support of basic research. The TRP directed over $800 million (with commercial firms providing additional funds) to scores of dual-use technology projects before the program ended in 1996.
One of the modest efforts in the TRP portfolio involved trying to stimulate commercial innovations based on the GMR phenomenon. In 1995, with an initial $5 million from DARPA and matching industry funds, Wolf initiated the GMR Consortium Project. In this program, large companies like Honeywell and newer firms like Nonvolatile Electronics partnered with Wolf’s former colleagues at the Naval Research Laboratory to develop a new form of computer memory
How did Wolf manage to get his GMR Initiative accepted by the State? Plopping a 40 pound satellite memory in DARPA’s director’s desk, and saying he could do the same with a fifty cent chip. It wasn’t a DARPA visionary effort, they just saw a man confident in the success of his idea.
Wolf sold his program by pulling an old memory component from a satellite system and taking it into the DARPA offices. “I plopped it on the director’s desk,” Wolf recalled. “It weighed forty pounds and cost a quarter of a million dollars. I said, ‘I’m going to replace this with a fifty-cent chip.’” Wolf’s modest program soon expanded into what DARPA initially called the Magnetic Materials and Devices Project. Unsatisfied with this moniker, Wolf suggested a new name—SPin TRansport electrONICS—which he shortened to “SPINTRONICS.” DARPA eventually provided some $100 million to support Wolf ’s SPINTRONICS program. During its six-year lifetime, four companies (Honey well, Motorola, IBM, and Nonvolatile Electronics) took part in it.
The project as such was not basic science, but applied engineering, with special application to Defence.
It is important to note that applied physics and engineering, not fundamental research, was the program’s primary focus. Through his SPINTRONICS initiative, Wolf helped broker a partnership between commercial and defence interests, an outcome that meshed well with the TRP’s goals.
The rest of the article, about the SPINTRONICS and SPINS programmes are basically the story of how Wolf and other scientists proposed successive programmes to DARPA, and got financed. In this case, according to McCray, it seems plausible to assume this financing was effective:
The general pattern of increase in publications and patents throughout the 1990s coincided with the growth of DARPA funding and the development of new magnetic data-storage technologies. While direct causality cannot be assumed, it is reasonable to conclude that the funding Wolf and his colleagues helped make available played a not-insignificant role in fostering the growth of the spintronics re – search community.
Finally, he does say that
The case of spintronics and federal support for nanotechnology in general are suggestive examples of what sociologist Fred Block calls the “hidden developmental state” through which the U.S. government has supported the private sector’s commercialisation of new technologies
But from the reading of the full article, what we can infer is that it refers mostly to the research side of the technologies (Scientific support), not the conversion of research into saleable artefacts. Let us remember that the first IBM disks were being sold before SPINTRONICS even began.
IN 2001, DARPA stopped financing this sector, but that didn’t cause researchers to stop investigating: they went on to ask companies for funds, and they got them.
Finally, the article also criticises (as Mazzucato also does in another chapter of the book, to be fair) the linear model of innovation, the idea that there is a causal arrow from scientific discoveries and the application of them, and that’s it.
To a first order of approximation, the case of spintronics appears to lend credence to the traditional linear model, which posits science as a prime mover for technological applications. As members of the 2007 Nobel committee saw it, an unexpected laboratory discovery inspired IBM’s industrial research and successful exploitation of the phenomenon and consequently billions of computers and iPods followed. The full story, of course, was much more complex, revealing the interplay among basic science, instrumentation, federal policy, industrial research, and commercial goals. One cannot help but conclude that the “simple” linear model, when examined closely enough, is anything but
Finally, a summarised and more readable version of the previous article can be found in McCray 2009b.
Click-wheels are not components of the iPhone as such, but of the iPod. One could argue that without iPod there wouldn’t have been no iPhone and therefore click-wheels were necessary for the latter. In any case, our interest in this section is to analyse the underlying technology.
Mazzucato begins by saying that the click-wheel benefited from capacitive technology, invented by E.A. Johnson, who published his first studies on the topic in 1960, while he was working in the Royal Radar Establishment, a British government agency dedicated to defence technology research. With respect to touchscreens, an important development was Bent Stumpe and Frank Beck’s 1973 work which we will discuss later on, and Samuel Hurst’s, who invented tactile resistive screens just after he left Oak Ridge National Laboratory to teach for two years at the University of Kentucky,and when he came back to Oak Ridge, he launched a company in 1971 to commercialise the technology, producing the first operating version in 1983.
The articles cited here are basically Buxton 2012 and Brown et al. n.d.
In Brown et al. we can find that the reason Hurst invented resistive touchscreens wasn’t that he was carrying out related research at Oak Ridge, but that he just had some trouble reading certain charts, task that even with the help of two students, would take months:
He was on leave from the Oak Ridge National Laboratory to teach at the University of Kentucky for two years, where he was faced with a need to read a huge stack of strip chart data. It would have taken two graduate students approximately two months to do the task. He started thinking of a way to read the charts, and during the process, the “Elograph” (electronic graphics) coordinate measuring system and Elographics the company were born. The University of Kentucky Research Foundation applied for and was granted a patent on the Elograph. The Foundation granted an exclusive license to Elographics.
His company began with a very Silicon Valley style model, in a garage and with scant resources:
In 1971, after returning to Oak Ridge, Dr. Hurst gathered nine friends from various areas of expertise to start a company to refine, manufacture and sell this new product. At this point Elographics truly began as a basement business. All work was done from three different basements; sensors in one, electronics in another, and cabinets in still another. The office was located in the home where the sensors were being built before moving to Four Oaks Center in February of 1972. The parts of the product were still being produced in basements at night and on weekends and brought to Four Oaks where they were assembled and shipped.
Was it necessary for the government to acquire huge amounts of these screens to make them viable? It seems not:
Ted Wilmart, one of the first sales representatives, sold over 60 Elographs at a price of $8000 per unit to a carpet company in Dalton, Georgia, who used the product to measure contaminated particle sizes in carpet.
During the time from March, 1971 until December, 1971, 25 units of Model E-100 Elographs were built, and Elographics started a very minor sales campaign. Out of this there were a few sales to end users. The first purchase order for one Elographics Model E-100 measuring system was taken on August 15, 1971 for delivery on November 30, 1971 to Massachusetts Institute of Technology for $995.00. Talk about extended lead-time!
On E.A. Johnson, I don’t question Mazzucato’s account. For Stumpe and Beck, apparently his screen was inspired by the work of the former in a TV factory, and this screen doesn’t seem to be the result of a long investigation:
In the early 1972, the Danish engineer (who had an experience from a television factory in the early 1960s), working in CERN—Bent Stumpe (see the nearby photo), was asked by Frank Beck, who was in charge of the central control hub in the SPS (Super Proton Synchrotron) control room, to build the hardware for an intelligent system which, in just three console units, would replace all those conventional buttons, switches, etc. In just a few days, in March, 1972, Stumpe came up with a proposal to build a touch screen with a fixed number of programmable buttons, a tracker ball to be used as computer-controlled pointing device (something like mouse) and a programmable knob. The first touchscreens were installed in 1973 and remained in operation until 2008. (History of computers)
A final detail is that a device with a touchscreen existed before E.A. Johnson’s: the musical instrument Electronic Sackbut, only that here technology wasn’t used to select options in a panel with images, but to generate sounds.
If modern smartphones are known for something, that would be gesture controlled multi-touch screens. Mazzucato says this technology was invented by Wayne Westerman and John Elias, from the University of Delaware (according to Westerman 1999). Westerman was a PhD student under John Elias, who studied neuromorphic systems in the university, as part of the post-doctoral joint programme NSF-CIA/DCI. His research were commercialised by a company they founded, Fingerworks, and they began selling their first product, the iGesture Numpad. This company was acquired by Apple in 2005.
But Elias and Westerman’s device was born out of serendipity:
Prof. John Elias and I began our Multitouch research in 1996 as a serendipitous outgrowth of neuromorphic systemsexperiments. We were interested in training the neuromorphs to memorize multi-point paths, like the patterns sensed by a patch of skin when fingers skim across. With no multi-touch tablets on the market to capture such patterns, we soon got ‘side-tracked’ building our own, drawing on the same analog VLSI and capacitive circuitry expertise with which we designed silicon neuromorphs.
And the first prototype, financed by the head of the department they worked in,
Special thanks to former ECE Dept. Chair Neal Gallagher for funding our first prototype from departmental coffers.
Multitouch research at UD was also supported by grants from the National Science Foundation and the CIA/DCI Post-Doctoral Fellow Program.
Multi-touch technology, however, comes from long before. Coming back to Buxton, whom we cited earlier, one can read there that he attributes the invention of the first multi-touch screen to Bob Boie, from Bell Labs, even though there was quite a lot of earlier work already there: as early as 1982 there existed a multi-touch system designed by Nimish Mehta from the University of Toronto (it wasn’t a screen, however). So it cannot be said that Westerman (And the relation with NSF and CIA/DCI programme) was crucial in the invention of multi-touch. But what about gestures?
According to Buxton, not the case either: Westerman’s thesis cites works where gestures are already mentioned, such as Myron Krueger’s VideoPlace, a digital artist (video).
Once again, reality is more complex than Mazzucato’s account, even if we are just reading her same sources.
SIRI began as a petition from DARPA to the Stanford Research Institute (SRI, a private nonprofit. That receives plenty of public funding). They wanted the to lead a project to develop an “virtual office assistant” to assist military personnel. This project was initially called CALO, Cognitive Assistant that Learns and Organizes, and it brought together 20 universities from all over the US. Afterwards, SRI launches a startup and commercialises SIRI, which ends up being acquired by Apple in 2010.
Here I don’t question Mazzucato’s interpretation, but I will point out that similar technology, developed separately, exists: Cortana from Microsoft, and Google Now, from Google. Also, people don’t really use SIRI. Without SIRI, the iPhone would be mostly the same device. And without DARPA, there would have been SIRI surrogates.
According to Mazzucato, the key development for LCD screens happens in the 70s, when the thin film transistor (TFT) is created in Westinghouse under Peter Brody’s direction. This was almost completely financed by the US Army according to Hart and Borrus 1992. Westinghouse executives ended the investigation at some point, and Brody tried to ask other companies to finance his research. He asked Xerox, 3M, IBM, DEC and Compaq, and none wanted to buy screens from him because his they doubted he would be able to manufacture them at a reasonable cost in relative to his Japanese competitors, according to Florida and Browdy 1991. In 1988, after securing a 7.8 million dollar contract from DARPA, he sets up Magnascreen to develop the TFT-LCD technology. Florida and Browdy see the LCD episode as a problem in the American innovation system: both big corporations and venture capitalists didn’t develop the product, even though they had made possible the semiconductor and personal computer industries (Which would contradict other things Mazzucato says about semiconductors and computers, but that is not the present article’s topic).
To retain leadership in TFT-LCD and keep manufacturing in the country, the main manufacturers launch a consortium, ADMARC, with support and funds from the National Institute of Standards and Technology, antidumping tariffs against the Japanese, and also contracts with many public agencies, both civilian and military, to buy these screens.
My lecture of this episode is basically that a group of companies successfully lobbies the US government to resist the superior Japanese technology. Having a better and cheaper alternative, it is not irrational to reject investing in TFT-LCD. Maybe the time for that would come later, and the project should have been carried out by long term research channels (research institutes and universities).
Let’s now read the evidence cited by Mazzucato. First, Hart and Borrus 1992.
The primary alternatives to CRTs for television displays are liquid crystal displays (LCDs). LCDs were developed initially in the Pittsburgh laboratories of Westinghouse and at the David Sarnoff Research Center (then owned by RCA) in Princeton, New Jersey. George Heilmeier and Richard Williams of the Sarnoff lab first put forward the idea that liquid crystals could be used for displays in 1963. Jim Fergason at Westinghouse began to pursue his own approach to LCDs in 1964. Fergason left Westinghouse in 1970 to found a new company called Liquid Xtal, which later failed due to delays in obtaining patents. Peter Brody, also at Westinghouse, pioneered “active matrix” addressing of LCDs. Brody left Westinghouse in 1979 and two years later founded his own firm, Panelvision, which was sold to Litton Industries in 1985.
The paragraph taken to support the book’s affirmation does fit within the explanation given:
In 1972-77, Davies managed the laboratory at Westinghouse which was working on active matrix flat panels. Peter Brody — the originator of the idea of active matrix — was one of Davies’ engineers. The lab had about 25 people funded mostly by money from the U.S. Army. Almost no Westinghouse funds were used. Army Electronics Command wanted electroluminescent (EL) displays because of their utility for night-time viewing. LCD displays were not considered appropriate for military applications. In addition, Westinghouse engineers in the Western Electric Tube Division in Elmira, New York, blocked commercial development of flat panel displays within the company. The money from the government was too little in the wrong place to develop active matrix TFT-LCDs.
It seems that screens had a relatively independent development, but that the step to LCD-TFT was totally financed by the Army. However, that technology was invented in Radio Corporation of America by Bernard J. Lechner (Kawamoto 2012) and implemented by his team in 1968, in a 2×18 matrix. Brody and his colleagues join the fray in 1974, and finally in 1988, Sharp produces a 14 inch full colour screen, that convinces the industry that there is future in the LCD-TFT technology.
Brody actually founded two companies. Before Magnavision he founded Panelvision and that one was successful. There it seems he started working in technology he would later implement for Magnavision without success, given he required high manufacturing volumes to make it profitable.
Let’s now check Florida and Browdy 1991
There we find that originally, Westinghouse wasn’t developing screens alone:
Westinghouse was not alone in these early days of fiat-panel display development. RCA had large scale efforts in both thin-film technology and flatpanel displays. Other companies-including General Electric, Hughes Aircraft, Raytheon, Zenith, Burroughs, Owens-Illinois, and IBM-were also active in the field. But most of them abandoned their efforts when they failed to come up with a way to produce inexpensive, manufacturable flat-panel displays. By the early 1970s, Westinghouse had the field almost to itself.
They also indicate that their work wasn’t entirely dependent on military contracts:
After making the rounds of Westinghouse divisions, Brody got several to sign on in support of the thin-film transistor research. (In the interim, he received military contracts to keep his work going.) At Westinghouse, his biggest supporters were the consumer-electronics division and the electron-tube division. Consumer Electronics was a large and powerful organization with a long history in radios, televisions, and home appliances: The division saw flat-panel displays as away for Westinghouse to gain ground on KCA and others in the television business, where Westinghouse was losing market share.
And against Mazzucato, Brody did manage investments after leaving Westinghouse after founding Panelvision, the company the economist does not mention,
Brody then got the attention of Wall Street venture capitalist Bruce K. Anderson of the venture-capital fund of Welsh, Carson, Anderson and Stowe. Anderson suggested that one of his major limited partners might be willing to fund the outfit. That limited partner turned out to be 3M. Even after being told about the previous turndown, Anderson still decided to proceed. In the brief interim since rejecting Brody’s earlier proposal, 3M had restructured. A new vicepresident now headed technology development, and the venture-capital firm’s proposal became his first opportunity to launch a visible new project. As a major producer of overhead projectors, 3M wanted to use active-matrix technology to make LCD overhead-projector screens. The board of directors took only three weeks to approve an investment of $1.5- million. In November 1980, the new company, called Panelvision, was launched. Panelvision bought equipment from Westinghouse’s old thin-film transistor labs. By the summer of 1981, the firm had rented a building in a Pittsburgh suburb near the Westinghouse R&D center, and begun developing a process for manufacturing active-matrix display products. Seemingly on the verge of pilot production, the company got an additional infusion of venture capital, bringing total investment to almost $4-million. […]
the company became reasonably successful. Between 1979 and 1984, it raised roughly $13-million in six or seven rounds of financing from heavyweight venture capitalists such as Welsh, Carson, Anderson, and Stowe; Drexel Burnham; First Chicago’s venture arm; and several Boston-area concerns. More significant, Panelvision became the first company to bring active-matrix display screens to market. In 1984, the firm began selling experimental products and lab prototypes. They soon had 80 customers in 12 industry segments.
Now, the goal was to increase production to make the company profitable, but with the Japanese competition, that seemed difficult
But it was impossible to break even, much less turn a profit, selling on such small scale. The company needed to develop a real manufacturing process and high-volume production capability-and this required more capital. After squabbles between the board and management over how to do this, the board hired Panelvision’s third president in three years, Tim DeSilva. Armed with a new business plan, the company aimed to raise $5-million and move into larger-volume production.
By this time the Japanese had entered the picture. Seiko introduced a color pocket television in the United States, infringing on the original Westinghouse patents for active-matrix displays, to which Panelvision held exclusive rights. The International Trade Commission encouraged Panelvision to bring suit. The company started this process in motion, alerting Seiko of a potential lawsuit. Japan’s entry sounded the death knell for Panelvision. Investors had already been hesitant about moving from R&D into volume production. Now they thought it utterly foolish to try to compete with the Japanese on their strong suit of manufacturing efficiency. The board of directors decided to recoup its investment by putting the firm up for sale. A team from 3M evaluated the firm and recommended taking it over, but top management declined. In 1985, Panelvision was sold to Litton Industries, which wanted to use the active-matrix technology in aircraft-cockpit displays.
Panelvision ended up being bought by Litton, who instead of producing screens for the consumer market, investing in technological improvements, chose to dedicate to the defense market:
Litton-Panelvision began to produce display products for its own defense avionics systems but never ventured into the commercial markets. And while Litton made some significant improvements, it was not in the business of advancing the technology.
It is now when Brody tries again with a new company, but it seems Body wasn’t able to promise attractive prices compared to the Japaneses’
Several major US computer makers were excited by the possibilities offered by flat-panel displays. Apple, IBM, DEC, and Compaq each indicated that they would place big orders, but shrank from becoming involved in the extremely expensive undertaking of building a factory that could produce large volumes of flat-panel displays. They believed it was not the job of computer firms to create their own supplier base, especially since they could buy flatpanel displays from the Japanese. The most receptive company was Apple, which was planning its Macintosh portable. Enthusiastic about active-matrix displays, Apple told Brody to bring back a proposal for a factory capable of producing 50,000 units a month. But Apple balked at the price tag. The company ultimately decided to buy screens for its portable Mac from a Japanese supplier.
But he went on and in 1987 he founded Magnascreen, with the objective of manufacturing screens for the defense sector
Brody decided to rethink his strategy. The Japanese had been concentrating on small displays (10 to 14 inches across), for laptop computers. Brody decided to develop larger (20- to 40-inch) displays for use in military command-and-control systems and in corporate teleconferencing. Brody saw large displays as the key to the next big frontier-high-definition television. The idea was to create a large screen out of smaller active-matrix “tiles.”[…]
The new company, called Magnascreen, attracted significant funding from individual investors close to Wiesner and Leghorn-including John Sculley of Apple and a former chairman of Xerox-and from Ven-West, the venture-capital arm of Westinghouse. All told, the company raised $2.3-million in start-up capital. Leghorn alone eventually put up more than $1-million.
And confirming what I mentioned earlier, for a different case, the DARPA contract seems to be the end result of a lobbying process (and not entrepreneurial state activity)
Despite difficulty raising money, Brody launched Magnascreen in 1988. The company bought Panelvision’s original Pittsburgh facility from Litton, and Brody rehired his old collaborator Tom Maloney. Magnascreen sought funding from the Defense Advanced Research Products Agency to develop a 45- inch color display. At the time, DARPA was headed by Craig Fields, who strongly supported industrial policy-the idea that government should channel money to develop technologies key to the nation’s competitiveness. Brody, naturally, became an ardent proponent of industrial policy, lobbying in Washington and writing letters to the popular press. Although Fields’s outspokenness on industrial policy got him fired by the Bush administration, Brody’s efforts paid off for Magnascreen. DARPA awarded the company a $7.8-million contract, of which it has so far provided $2-million
US manufacturers of LCD screens were complaining of the Japanese competition, and even when no evidence of dumping was found, small importation duties were imposed (An antidumping tariff is usually above 10%, reaching even 40%. It is doubtful that the small tariffs we are discussing here would have had any significant effect)
The situation is so serious that US computer makers are siding with the Japanese against the US display makers. Last July, a coalition of seven US flat-screen producers accused 12 Japanese companies of “dumping” flat-panel displays in the United States at prices well below those in Japan. But at a preliminary hearing before the International Trade Commission, IBM, Apple, Compaq, and Tandy testified against the US display companies. The computer makers insist that they have no choice but to turn to Japanese vendors because domestic companies are unable to produce large volumes of displays. Even the Semiconductor Industry Association (SIA)-which has aggressively challenged “unfair” Japanese trade, and favours an industrial policy to rebuild US consumer electronics- refused to get involved, perhaps fearing retaliation from US computer firms. In its initial ruling in February, the Commerce Department found no evidence of dumping by Hoshiden or Matsushita. It imposed small tariffs of 1.46 percent for Toshiba, 4.6 percent for Sharp, and 2.33 percent for the rest. The department was to make a final ruling in July following on-site investigations in Japan.
While significantly higher tariffs are unlikely, even modest increases may force more US manufacturers of laptop computers to move production to Japan, or convince Japanese display makers to move more production to this country. Sharp is already building a $30-million plant in Camas, Wash., where it expects to produce up to half-a-million portable computer displays a year.
Some companies, trying to launch a recovery of their LCD manufacturing sector in the US, set up ADMARC, that Mazzucato mentions, with about 1.25 million dollars from NIST:
Last year, a group of 10 small companies-Cherry Electrical Products, Coloray Display Corporation, Electro-plasma, Magnascreen, Optical Imaging Systems, Photonics, Planar Systems, Plasmaco, Standish Industries, and Tektronix-banded together to form the Advanced Display Manufacturers of America Research Consortium (ADMARC) to develop flat-screen technology. In March, ADMARC received a $1.25-million grant under the National Institute of Standards and Technology’s new Advanced Technology program.
But the problem was not research and development, but manufacturing!
While such consortia are a move in the right direction, they are not the answer for the US display. industry. Research consortia, by their nature, focus on high-end R&D or advanced development work-so-called generic or precompetitive technology. They have not had great success in manufacturing, where US industry is weakest. Indeed, we may well see a repeat of the computer-memory-chip story.
The article, with somewhat of a mercantilist tone, ends with a recommendation for American corporations: to apply Continual Improvement (aka Kaizen). The problem wasn’t technologica, says the article, it was organizational:
US investments in manufacturing must be coupled with deep organizational and management changes. The (mainly Japanese) companies that have succeeded in active-matrix technology have applied a basic formula: continuous process improvements on the factory floor. In these companies, R&D scientists and engineers work alongside factory workers to make sure the manufacturing process works. The factory is a center for innovation, change, and constant refinement. Such perseverance has, more than any other single factor, spelled success for the Japanese in active-matrix technology. This is where we failed and continue to fail.
After this series of interventions, as of today, the US has not managed to lead LCD screen manufacturing, yet we have ever better and cheaper screen of many different sizes. If one looks at the world from a mercantilist point of view, it certainly is a defeat for the US and its supposed mercantilist policy (and its Entrepreneurial State). But if one sees it from the point of view of consumers all over the world, they have won. Mazzucato criticises investors and American companies for avoiding battling the Japanese, and that that’s a failure on their side. I would criticise her for seeing that as a problem.
As an additional note: If tariffs and the like had been successful, would Mazzucato score a goal for the US Entrepreneurial State, saying that it wasn’t possible for Japanese to develop LCD screens instead?
Next example are lithium batteries. John B. Goodenough was the researcher who opened the path in that secor, receiving funding from the Department of Energy and the National Science Foundation in the late eighties, according to Henderson 2004 and OSTI 2009. Later on, scientific developments in the University of Texas were commercialised by Sony in 1991. In a NIST conference in 2005, Ralph J. Brodd identified problems similar to those we saw in the case of LCD panels, and blames corporate and investor short termism for them. Says Brodd that the federal government helped small companies through a variety of agencies and rogrammes to strenghten their productive capacity.
Let’s read. In Henderson 2004 and OSTI 2009 we find that Goodenough started developing his technology in Oxford University, before going to the University of Texas, even though it was there where he improved his designs. His international fame, however, did came from his work at Oxford: the invention of lithium-cobalt oxide batteries. In general, the history of batteries shows the usual pattern: many development centers, in many parts of the world.
Brodd 2005 confirms what we said about Goodenough and Oxford. With regards to state support of small businesses, it is there, but it doesn’t seem to have helped much except to develop batteries for some niches (defence and healthcare). We may wonder what was the cost of the programme.
In this Reuters chart, we can see who manufactures Li-ion batteries today, and they are still the Japanese, plus Koreans. ame as in the case of LCD screens: Batteries are being manufactured and improved. Does it matter where they are manufactured?
Some could argue that Sony’s battery is an outcome fostered by Japan’s Entrepreneurial State. According to Sony, the history of their battery is this. Trying to argue that is left as an exercise for the reader.
In a previous post, we saw this chart:
We already criticised her argument with repect to LCD screens, Li-ion batteries, click wheels, multitouch screens, SIRI, and hard drives. We now have to criticise DRAM, signal compression, and celullar technology, because as I said before, I won’t discuss technologies related to the internet (Internet, HTTP/HTMP, TCP/IP), GPS and semiconductors.
Before continuing, it is of interest to comment that this figure comes from a diagram that one can find in a report from the Office of Science and Technology Policy (2006). The report does not have citations to back the chart, which can be found in a simplified manner in page 11, with evidence. The caption says that research financed by various public agencies contributed to the development of portable MP3 players
Even simple hand-held calculators were rare and expensive at that time. Research funded by the Department of Defense, the National Science Foundation, the National Institutes of Health, the Department of Energy, and the National Institute of Standards and Technology contributed to the breakthrough technologies of magnetic storage drives, lithium-ion batteries, and the liquid crystal display, which came together in the development of MP3 devices. The device itself is innovative, but it built upon a broad platform of component technologies, each derived from fundamental studies in physical science, mathematics, and engineering
What precedes the figure, though, is more dubious, given what
we have argued,
And in every case, research funding at NSF, DoE SC, or NIST core (consisting of NIST lab research and construction accounts), has been essential to proceed to the point at which the private sector recognizes a potentially marketable product and invests in its development.
As much, you could argue in favour of basic science research, but not much more. We have seen that who have really have done the effective commercialisation, and technological transfer from the lab to the market have been companies. This doesn’t look much like an Entrepreneurial State.
Fast and Furious: The Fast Fourier Transform
The Office of Science and Technology Policy (2006, 8) also documented the role of State support in the digital signal processing (DSP) technology that came about following scientific advancements in the application of the fast Fourier transform (FDT) algorithm during the 1980s. This new signal processing approach enabled real-time processing of sound (such as during a two-way phone call) as well as real-time processing of large audio or multimedia files that can improve the quality of their playback.
Says the report that it was in 1965 when the FFT revolutionizes signal processing. Searching, I find that she means the rediscovery, by Cooley and Tukey, of Gauss’ algorithm to calculate the FFT (Heideman 19849, a century before Gauss. Note that since Gauss, advancements had been made in FFT, and that Cooley and Tukey represented just one step in that chain. Rediscovering Gauss with this paper, which yes, financed by the Army Research Office, doesn’t seem that important if you want to argue for the role of the ARO in the discovery, given that the authors worked for Princeton University (Cooley was also an IBM employee, and Tukey a Princeton professor). A mathematical advance doesn’t seem to require huge quantities of resources, and Princeton and IBM are and were financially sound entities, which were supporting whatever research Tukey and Cooley were doing at the moment. If even after this is said one still wants to argue that this event is due to ARO funding, with more reason you will have to say that we have this algorithm thanks to Princeton and IBM, and then with that same reason the algorithm is a private sector triumph. Now, this delves into concepts of causality, which we will addressed in future posts, so if you are not convinced by the above reasoning, I hope you will find it more plausible after I explain the implicit view of causality I’m taking here.
In this case, the chart points us to 1960-70 and the development of VLSI (Very Large Scale Integration) technology. It points us to basic science funding of IBM and DARPA. Mazzucato doesn’t devote much lines to talk about this and neither will I. You can read the history of DRAM here and see how much state financing it had (Which isn’t even mentioned). In Itoh (2008) I cannot find state influence either. We find IBM, Intel and Hitachi. It doesn’t seem plausible to affirm that DRAM was a Entrepreneurial State creation, given the evidence supplied by her and the additional evidence found by me. It could be argued that VLSI was a Entrepreneurial State creation, and via that, that there was an indirect influence in DRAM. But this belongs to the category of semiconductors, and along the points I’m not addressing here (Which aren’t that central for the iPhone), I’ll discuss them in a separate article.
In the chart it is said that the US Army supported this technology and the Breakthough Report (2010) is cited to support this affirmation. The report says
The early foundation of cellular communication lies in radiotelephony capabilities progressively advanced throughout the 20th century with support from the U.S. military
But doesn’t ground that in anything at all, nor it expands on that point. Using evidence this poor to affirm something as controversial in a written book is somewhat sloppy. Let’s be charitable and examine other evidence we could have. Let’s go beyond Mazzucato. Where does cellular communication come from?
In Gosh (ch. 1,2010) we find no state influence in technological development of cellular technology – It is present in the GSM protocol, but that came later, and certainly not from the US Army. Could it be the Mobile Subscriber Equipment (MSE) from General Dynamics, developed in 1985 for the US Army? According to Ghosh, in 1983 there were already commercial cellular services in Chicago, in 1979 in Japan, and Bell Labs had already fleshed out the concept in 1960-70. So it seems dubious.
Again, it is left as an exercise for the reader to try to find a link between the development of this technology and the State.
The third Mazzucatian point was that the American State had protected Apple’s intellectual property, and fought to open the international market for them. It is the government they go to when global conflicts occur. Says that Apple had problems accessing the Japanese market in the eighties, and that it was able to enter it thanks to US government support. Also, Apple receives a series of sical benefits, R&D incentives, and State purchase of their products via public schools.
But even accepting all of this, we would have to quantify it to see if it is really useful for Apple, or for society in general. Intellectual property is something companies who manufacture iPhone clones don’t seem to care about, and could be more detrimental than beneficial. Apple patent wars vs the world, patents in hand, are really useful to allow technological progress? Doubtful (Boldrin & Levine 2002)
On the other side, the best that can be done with this point, even if it were correct (that is, that it happened as Mazzucato says, and that it did help Apple) is to criticise it, and ask for an end to corporate welfare, that even from a simple efficiency point of view doesn’t seem much beneficial, at least until Mazzucato presents us with solid evidence.
A related point is that this whole problem is generated by the kind of policies Mazzucato seem to endorse, or at least don’t oppose. Maybe mercantilist policies such as tariffs and nascent industry protection can foster a national industry in that sector. Or just decrease overall welfare, depending on its success, and they usually aren’t succesful, as Park Chung-hee’s don’t abund. Even then, if we take a global perspective, we should care not about national industry, but about global consumer welfare, thus it doesn’t matter much where a technology is developed, but that it is developed, and that it can be exported and sold all over the world.
The chapter of the book ends like this
In sum, ‘finding what you love’ and doing it while also being ‘foolish’ is much easier in a country in which the State plays the pivotal serious role of taking on the development of high-risk technologies, making the early, large and high-risk investments, and then sustaining them until such time that the later-stage private actors can appear to ‘play around and have fun’. Thus, while ‘free market’ pundits continue to warn of the danger of government ‘picking winners’, it can be said that various US government policies laid the foundation that provided Apple with the tools to become a major industry player in one of the most dynamic high-tech industries of the twenty-first century so far. Without the frequent targeted investment and intervention of the US government it is likely that most would-be ‘Apples’ would be losers in the global race to dominate the computing and communications age. The company’s organizational success in integrating complex technologies into user-friendly and attractive devices supplemented with powerful software mediums should not be marginalized, however it is indisputable that most of Apple’s best technologies exist because of the prior collective and cumulative efforts driven by the State; which were made in the face of uncertainty and often in the name of, if not national security, then economic competitiveness.
Basically this final paragraph is false. State interventions in science that were conducive to the iPhone were neither crucial, nor entrepreneurial, not as numerous as Mazzucato tries to show.
Let us ask again what makes the iPhone the iPhone. Mazzucato could have tried to say that the State invented the mobile phone, or the smartphone, but she chose the iPhone (Because the controversy whe supposed it would generate, and did generate). But the truth is, most of the things we have talked about here are somewhat irrelevant to answer this question. If we take an iPhone and any other high end smartphone (especially from the time the iPhone was released), we won’t find the difference in the screen, internal memory or GPS. Most smartphones have that.
What makes the iPhone the iPhone, a product that managed to put Apple in the position it is, is design, its iOS operating system, a correct technological integration, and a careful quality assurance process. All this factors are essentially internal to Apple itself. Without these things we would have an smartphone that wouldn’t be an iPhone. That is what it gave Apple its comparative advantage, the base of its huge revenues. Mazzucato addresses in other chapter -we’ll come to that in other article- the question of whether the State deserves part of Apple’s benefits. To this, a triple criticism can be made: First, the one presented in this article. The role of the State is not the one Mazzucato talks about. Second, those extra profits, comparatively, are not to be attributed, in the case Mazzucato were right, to state reserach, but to internal company factors. And third, the ethical premises Mazzucato uses to argue (Any recommendation implies an ought) for Apple owing something to the State are dubious.
Technology in general, and the iPhone as a particular case is a complex story. Talking of who invented this, or who discovered that usually hides many nuances that we shouldn’t obviate when we study the history of technology. And of course, talking about an Entrepreneurial State to whom we owe the phone that launched the smarthphone revolution doesn’t have basis in available evidence. The State was one more actor, like IBM, Bell Labs, Sony, Goodenough, Brody, or Lechner.
Science and Technologies have heroes who make it go forward step by step, not Gods without which progress is impossible. We are right to discard the Great Man Theory of Innovation, where we owe individual geniuses for discoveries only them could have made, but for the same reason we should also discard the Great State Theory.
Audretsch, D. B. 1995. Innovation and Industry Evolution. Cambridge, MA: MIT Press.
Binasch, G., Grünberg, P., Saurenbach, F., & Zinn, W. (1989). Enhanced magnetoresistance in layered magnetic structures with antiferromagnetic interlayer exchange. Physical review B,39(7), 4828.
Boldrin, M., & Levine, D. K. (2002). The case against intellectual property.
Breakthrough Institute. 2010. Where Good Technologies Come From: Case Studies in American Innovation. Oakland, CA: Breakthrough Institute, December. Available online at: http://thebreakthrough.org/blog/Case%20Studies%20in%20American%20Innovation%20report.pdf (accessed 24 January 2013).
Brodd, R. J. 2005. ‘Factors Affecting U.S. Production Decisions: Why are there No Volume Lithium-Ion Battery Manufacturers in the United States?’ ATP Working Paper Series, Working Paper 05-01, prepared for the Economic Assessment Office Advanced Technology Program, National Institute of Standards and Technology (NIST), June.
Brown, D. et al., eds. n.d. ‘History of Elo’. Elo Touch Solutions. Available online at http://www.elotouch.com/AboutElo/History/default.asp (accessed 22 January 2013).
Buxton, B. 2012. ‘Multi-touch Systems That I Have Known and Loved’. Microsoft Research. Available online at http://www.billbuxton.com/multitouchOverview.html (accessed 3 October 2012).
Cooley, J. W., & Tukey, J. W. (1965). An algorithm for the machine calculation of complex Fourier series.Mathematics of computation, 19(90), 297-301.
Fert, A. (2007). The origin, development and future of spintronics. Nobel lecture.
Florida, R., & Browdy, D. (1991). The invention that got away. The Journal of Technology Transfer, 16(3), 19-28.
Ghosh, A., Zhang, J., Andrews, J. G., & Muhamed, R. (2010). Fundamentals of LTE. Pearson Education.
Hart, J. and M. Borrus. 1992. ‘Display’s the Thing: The Real Stakes in the Conflict over High Resolution Displays’. Berkeley Roundtable on the International Economy (BRIE) Working Paper no. 52. Available online at http://brie.berkeley.edu/publications/WP%2052.pdf (accessed 24 January 2013).
Heideman, M. T., Johnson, D. H., & Burrus, C. S. (1984). Gauss and the history of the fast Fourier transform. ASSP Magazine, IEEE, 1(4), 14-21.
Henderson, J. 2004. ‘UT Professor, 81, Is Mired in Patent Lawsuit’. Houston Chronicle, 5 June. Available online at http://www.chron.com/default/article/UT-professor-81-is-mired-in-patent-lawsuit-1662323.php (accessed 11 February 2013).
Itoh, K. (2008). The history of DRAM circuit designs–at the forefront of DRAM development–.Solid-State Circuits Society Newsletter, IEEE, 13(1), 27-31.
Kawamoto, H. (2012). The inventors of TFT active-matrix LCD receive the 2011 IEEE Nishizawa medal.Journal of Display Technology, 8(1), 3-4.
Lazonick, W., Mazzucato, M., & Tulum, Ö. (2013, December). Apple’s changing business model: What should the world’s richest company do with all those profits?. In Accounting Forum (Vol. 37, No. 4, pp. 249-267). Elsevier.
Mazzucato, M. (2013). The entrepreneurial state: Debunking public vs. private sector myths. Anthem Press.
McCray, W. P. 2009a. ‘From Lab to iPod: A Story of Discovery and Commercialization in the post-Cold War era’.Technology and Culture 50, no. 1 (January): 58–81.
McCray, W. P. (2009b). How spintronics went from the lab to the iPod. Nature nanotechnology,4(1), 2-4.
Moritz, M. (2009). Return to the Little Kingdom: Steve Jobs and the Creation of Apple. Penguin.
OSTI (Office of Science and Technical Information). 2009. ‘John B. Goodenough, Cathode Materials, and Rechargeable Lithium-ion Batteries. DoE R&D Accomplishments’. 17 September. Available online a t http://www.osti.gov/accomplishments/goodenough.html (accessed 14 February 2012).
OSTP (Office of Science and Technology Policy). 2006. ‘American Competitiveness Initiative: Leading the World in Innovation’. Domestic Policy Council, Office of Science and Technology Policy, February.
Rao, A., & Scaruffi, P. (2011). A History of Silicon Valley: The Largest Creation of Wealth in the History of the Planet: A Moral Tale. Omniware Group.
Slater, K. 1983. ‘Banks Seek Profit in Venture Capital Arena; High-Risk, High-Reward Business Is Attracting More Adherents’. American Banker, 31 October.
Westerman, W. 1999. ‘“About” Wayne Westerman’. Available online a thttp://www.eecis.udel.edu/~westerma/About_Wayne.html (accessed 22 January 2013).