[Note: below is Chapter 20 from Great Wall of Numbers]
Any company wanting to conduct international business on the mainland, irrespective of whether it is local or foreign, must invariably factor in the additional costs of communicating electronically beyond the mainland. Beginning in 1996 and launching in 1999, the Ministry of Public Security and other governmental organizations began implementing and enforcing a series of regulations involving data filtering that ultimately matured into what is commonly referred to as the “Great Firewall” (GFW). Collectively, the types of filtering techniques employed by the Ministry through Tier 1 ISPs (all SOEs as noted in Chapter 15) include: IP blocking, DNS filtering, URL filtering, packet filtering and connection resets. Simultaneously, the Ministry maintains a continuously updated “black list” of websites that mainland users are unable to access through this dynamic filtering and blocking mechanism.
This presents an opportunity to virtual private network (VPN) providers overseas. A VPN is a type of technology that effectively creates a secure tunnel from one computer to another, isolating its data stream from the surrounding traffic. This can be done by means of encryption and as a consequence many banks, financial institutions and national security centers – irrespective of the region or hemisphere – typically use some form of VPN to securely communicate with outside parties (e.g., for wiring money, discussing trade secrets, or diplomacy).
There is no shortage of VPN providers in Western countries and there are in fact, Chinese-based VPN providers as well – the efficacy and reliability of which is debatable. In my own anecdotal experience, even with a paid service based in a foreign country, data can still be throttled and your connection reset. One reason is that the GFW is not a passive system – it is continually tweaked and changed. In an interview in 2011, Fang Binxing, the Father of the Great Firewall explained that he himself has “six VPNs on my home computer.” He uses them to “test which side wins: the GFW or the VPN.” And in his opinion, “[s]o far, the GFW is lagging behind and still needs improvement.”
One world, two internets
As I mentioned in Chapter 12, while there are any number of domestically made and managed counterparts and clones of foreign social media services (e.g., Sina Weibo is the equivalent to Twitter), there is still niche demand for foreign-based web services. For example, as I mentioned in Chapter 9, there are now about a million Chinese students studying overseas each year; more than 190,000 Chinese students studied in the US this past year alone. In addition, 1.36 million Chinese tourists visited the US in 2011. What this means is that as I mentioned in Chapter 3, Chinese consumers are increasingly exposed to Western and in particular, American tastes and services.
Yet to temper the optimism that a VPN provider could immediately sell several million service packages to individual mainlanders, consider this rough facsimile: while we may never know the real number, Facebook’s own indirect estimate of mainland usage of its social network is close to 600,000. While there are a number of other niche services in demand, especially from financial service firms, this 600,000 number can be used as a proxy to estimate the general demand for VPNs.
It should also be noted that not all foreigners use Facebook in China nor do all foreigners want to pay for a VPN to access it. Furthermore, based on my own anecdotal experience at various institutions, the average Chinese user does not currently have access to a VPN or other fan qiang (“Wall Climbing”) software such as UltraSurf or Hotspot. In their mind, why should they have to pay to access foreign services when there is a similar Chinese version available for free? This is not to say that they could not gain access to the services if they were motivated and inclined to do so. In my own anecdotal experience virtually none of the several thousand students I have had at various cities on the mainland over the past four years have had active Facebook accounts. A few however have had VPN accounts so they could play online games like World of Warcraft on servers outside the mainland (e.g., “gold farmers,” see Chapter 14).
Assuming the number is around 600,000, how much can a foreign-based service provider expect to generate? Currently, the average monthly rates at PandaPow, Astril and most others are roughly $10 a month. And because a large portion of a VPN package is based on software that is open-source and free, the initial setup costs are minimal. Yet bandwidth charges, hardware purchases, customer support labor and utilities charges all vary and will depend on how large you plan to scale your company to. Thus before investing in this segment, do your due diligence.
Entrepreneurs should also consider this: Bill Bishop has cogently noted numerous times over the past several years that while mainland users are effectively prevented from using some foreign web services, the mainland equivalents are not only more easily accessible and relatively comparable (e.g. same features), but the interconnectivity issues (e.g., latency, bandwidth) with them are relatively muted. In other words, assuming you have access to a VPN, it is still more convenient for mainland based users to stream videos on Youku than it would be to stream from Youtube because of the increased bandwidth throughput and reduced lag due to closer proximities to the content delivery networks (CDNs) for Youku-like providers such as ChinaCache. David Wolf, a partner at Allison+Partners (a consulting firm) echoed similar reasoning recently in an interview with The Wall Street Journal, noting that “What they [national government] prefer is that Chinese users decide it is just too much trouble and by default use onshore sites, or sites that are mirrored onshore.” As a consequence, because of the sheer size of the Chinese-based internet (see Chapter 12 and Chapter 13), there is now “one world, two internets.”
Climbing the wall
I spoke with security expert David Veksler (see also Chapter 13), CEO of CryptAByte, who has given security workshops and seminars about these issues. In his view, “the GFW presents a fundamental problem for domestic researchers. Because significantly large portions of foreign-based information are blocked and denied, only researchers with VPNs are able to keep up-to-date with their foreign counterparts. Those without VPNs are left trying to use Google which is frequently blocked and misdirected or Baidu, which outputs few useful or useable results. Thus they become discouraged, often times quitting and are ultimately unable to do the necessary research – idea investigations – for innovation.”
How does this create opportunities? According to Veksler, this ties into another Catch-22 that domestic firms find themselves in, this endless cycle of benchmarking and cloning. That irrespective as to whether or not they want to innovate, they are in a prisoner’s dilemma, “every competitor on the mainland expects to have the lowest costs. Yet if they increase their research and development – creating higher quality products – consumers do not believe them, because consumers also expect that domestic companies are cutting corners, so why pay extra for a product that is probably just the same as the rest?”
He also likens this dilemma to a game theory scenario: the first domestic company to make that leap into quality is punished because consumers simply do not trust the product quality due to a history of scandals. Thus any firm that does it is unable to recoup the capital costs of the research and development. In contrast, foreign companies have spent decades building up their brands and reputations based on quality control programs (e.g., Six Sigma) and as a consequence are readily more trusted on the mainland. Yet he remains optimistic, “the first domestic company to make a concerted, long-term leap into quality will not only be monetarily successful, but will help end this never ending cycle of benchmarking and cloning.”
Thus Veksler thinks that foreign brand managers, experts like Matt Garner, will be able to find opportunities to work within the entire spectrum of industries as their participants build national and internationalization expansion plans.
Chicken and egg problem
It is hard to measure the impact that an apparatus like the GFW has on productivity and creativity which business start-ups should take into consideration. Consider Silicon Valley and Moore’s Law. Much like other projects and mian zi gong cheng, there have been several public initiatives to replicate Silicon Valley in China, such as Zhongguancun in Beijing. And yet for every successful start-up like iQiyi or venture capitalist like Kai-Fu Lee (see Chapter 12), large quantities of resources have been misallocated towards supercomputers that when installed – while capturing headlines for theoretical peak performances – are unable to be fully utilized because there are not enough trained software engineers to develop the sophisticated machines. Similarly despite 2 billion RMB ($320 million) in investment since 2010, Jike, a new search engine developed by People’s Daily (an SOE) has managed to capture a mere 0.0001% marketshare forcing the organization to lay off 20% of its staff.
Empirically speaking, if central planners were to be the creators of Silicon Valley, they would have created Silicon Valley. If central planners were to be the creators of Moore’s Law, they would have created Moore’s Law. For example, the Soviets spent decades and relatively large budgets to overtake the West in computing innovations, yet failed at every turn. In fact, it was not just one or two half-hearted attempts, it was a concerted effort directed from the top. Mikhail Gorbachev himself made advancements in microprocessor technology a cornerstone part of Perestroikain 1985 (encompassing the 14th Five Year Plan).
Just how much effort was put into their centrally planned machine industry? Consider what the USSR tech industry was like circa 1988:
Machine building is the sector of industry on which Gorbachev is relying to ensure the success of his [Perestroika] strategy. The hub of Soviet [computing] industry, this complex employs over 16 million workers at more than 9,000 research institutes, design bureaus, and production and enterprises, and is responsible for designing, developing, and producing over one-fourth of the country’s industrial output. Of the 17 industrial ministries that make up the machine-building complex (detailed in foldout at back of paper), nine — collectively referred to as the defense industry — specialize in military hardware. The other eight produce primarily consumer goods and equipment for investment in the civil sector.
Gorbachev recognized that “a high-investment, high-growth strategy must, at a minimum, continue through at least the first few years of the period to renew the sector’s capital stock.” Yet ultimately, the Soviets tried, consumed their capital base, and failed. Instead, hundreds of private companies, entrepreneurs, venture capitalists, designers, and one relatively free market created a semiconductor industry that accounts for the number one export of the United States. Furthermore, this is not to say that technological activity will not take place in China, nor that Chinese institutions and researchers will not produce usable technology. The question is rather, can it be cutting edge and innovative? And if your firm hopes to tap into the innovation potential of the mainland, how does this impact your firms’ investment?
Many of these artificial technology and science research parks conflate cause and effect. For example, during World War II, the Allies used Pacific islands as forward operating bases to protect their overseas supply routes. On many of the islands the Allied forces built airstrips, including one on Vanuatu. Following post-war demilitarization, most of these islands were vacated as the warring militaries returned home. On Vanuatu, many of the islanders wanted the supply ships to return and provide modern goods to their pre-industrial society. As a consequence, the islanders staged “drills” and “marches” with mock soldiers while others attempted to man the airstrips – all under the belief that it is these superficial motions and actions that originally brought the Western supplies. Richard Feynman dubbed this “cargo cultism” (e.g., a cult that dreamt of Western cargo).
In November 2012 I spoke with Mark Thornton, an economist at the Ludwig von Mises Institute and an expert in the boom-bust investment cycle. According to him, “Research parks are all about inventing technology for commercial and other purposes. Generally we are speaking of higher order goods, the types of goods associated with the boom phase of the business cycle. Therefore we would expect that research park projects tend to be established during booms when profits are high, the cost of capital is low, and where retained earnings are more than sufficient to support additional projects. If research parks are established at or near the peak in the business cycle then it would be wise to avoid contracting with research parks that have few tenants. Traditionally one of the main benefits of research parks is synergy. If your research park has no tenants then you do not have the type of synergies that successful research parks generate. New companies, new technologies and products, as well as successful research parks (e.g. Stanford Research Park and Research Triangle Park) tend to get their starts during bad economic times. During recessions land, labor, capital are cheaper and budding entrepreneurs are more abundant.” In economic terms, higher order goods are goods used to produce consumer goods (e.g., those which require a long-term investment such as building a factory which in turn creates consumer goods).
Similarly, many of these research parks and endeavors – not just in China – arguably exhibit patterns of modern-day cargo cultism. Thornton noted that, “The next Silicon Valley will not look like Silicon Valley. It will have some new features and not have all the same features as Silicon Valley. You cannot just build “it” and expect them to come. Silicon Valley is more than just Stanford Research Park and Stanford University. There are tangible and intangible factors that matter. They include things like the weather, demographics, culture, and relatively limited regulatory impact from the government. Even some factors we just do not know. Government can subsidize research parks but it takes a free market and entrepreneurs to actually weave the fibers of something extremely complex like Silicon Valley.”
In fact, in the US, nearly every state has erected several tech parks with the hopes of “creating” another Silicon Valley; there are dozens of research and technology centers across the country. This raises the question: if you build it, will they (the creative classes) come?
In February 2013 I spoke with Becky Wu a native of Jiangsu province and a project manager at Xi-Tong Scientific & Technology Industrial Park located in Nantong, Jiangsu province. The primary task of her job is attracting and relocating foreign firms so that they will build and setup operations in the industrial park. According to her, “we provide incentives and subsidies to attract firms from abroad. For example, if land prices were with 230,000 RMB per mu, depending on how promising the project is and what industry your firm is in we can lower the price to 200,000 RMB or even 150,000 per mu. This helps attract firms, enticing them to construct their new offices in the park. We will also provide free temporary offices for new companies for up to 6 months while their new office is being built. The utilities are also free of charge as well.” As noted earlier in Chapter 3, a mu is 1/6th of an acre.
Wu also explained that there are other rebates and training subsidies that firms can receive. She noted that, “we also offer new companies subsidies for research and to train personnel that can be allocated and spent without strings attached. For example, we can provide up to $1,000 a year per person, up to 10 people to help offset training and research costs. In terms of income taxes, we provide rebates to specific workers, typically managers and high-level executives for 3-5 years. The way this works is that if you have to pay 100 RMB in taxes, 60% goes to the central government, 8% goes to the provincial government, the remaining portion goes to Nantong, thus we at the park can reimburse the remaining 32% back to you.” Clients such as Caterpillar, BIC, Accuma and Kopron have taken advantage of these incentives over the past several years.
Does the return-on-investment pay for the capital expenditures which were originally expended? While it is impossible to say yes or no for all the cases, what can be said is that the GFW itself probably does not create innovation, foster creativity or act as an incentive to attracting outside talent. If it did, the Chinese computing industry would not be reliant on Western semiconductors, Western software and foreign know-how. And as a consequence, mainlanders conducting research are left using a virtual straw in order to access, view and communicate with the outside world.
How is this relevant and how does this affect your company? Without virtual openness to new ideas, the domestic, indigenous engineering industries – while not autarchic – will probably always be laggards due to what Veksler noted above (e.g., getting frustrated and quitting). To this point, last year the American Chamber of Commerce in Beijing conducted a survey of its members, “nearly three-quarters of about 300 businesses it surveyed said unstable Internet access impedes their efficiency. About 40% said China’s censorship efforts have a negative business impact.” Similarly, economist Arthur Kroeber, founder of Dragonomics research noted in March 2013 that one obstacle to growth is the GFW. In his view, innovation in the modern world today comes from “the sharing of knowledge and information across a variety of fields. Innovation comes when you take knowledge in one area and it migrates over to another area and someone comes up with a new way of using it. China seems to have a political system that mentally at its core is opposed to those networks ever becoming viable.” Thus, in addition to the issues raised in Chapter 15, this obstacle is another consideration that all firms looking to recruit talent must take account for.
While there are occasional opportunities and projects like “1,000 talents” (mentioned in Chapter 9 and Chapter 15) that provide monetary and other perks and incentives to relocate, these well-intentioned plans may be unable to offset the hurdles created by the GFW and as a consequence there has been a “brain drain” that all firms and HR departments should be aware of.
Yet to be even handed, Larry Chang mentions that he works within this system on purpose because it is “an untapped opportunity.” He only hires fresh mainland graduates with the sole purpose of building an indigenous software industry. And in his opinion, with more than 6 million college students graduating each year, there are bound to be creative, outside-the-box thinkers. Similarly, at the 2013 Unleashing Innovation conference recently held in Singapore, Ya-Qin Zhang, chairman of Microsoft’s Asia Pacific research and development group, noted that “Chinese engineers are well equipped to produce the kind of innovative work that their more illustrious American rivals are renowned for” and continued with, “[t]he scale of innovators and the scale of the market will converge and eventually make China a key [innovation] center in the region.” Thus it may just be a matter of time before the right combination of inputs brings about the transition up the value chain as described in Chapter 7.
Opportunities in the rough
Again, even with these seemingly insurmountable challenges there are also opportunities. For example, as I noted in Chapter 17, foreign architects are in high demand to help build and design buildings, bridges and even office parks. Perhaps your firm can find new revenue streams by helping to build out domestic content delivery networks (CDNs) and cloud computing initiatives that are part of these technology parks. As I mentioned in Chapter 13, according to IDC, $286 million was spent on cloud-computing infrastructure in China in 2011 and this is expected to increase to $1 billion by 2016. And this segment is quickly professionalizing, for example, ChinaCache is the largest CDN on the mainland with 53% of the marketshare. It was initially funded by the likes of Intel and is now listed on NASDAQ.
Another opportunity is with corporate VPNs. While the individual market may seem like a logical way to establish a steady revenue stream, according to David Veksler, corporate enterprises – both domestic and foreign – will eventually want and need to have VPNs to secure their communication with clients, vendors and essentially anyone. Irrespective of the GFW, Veksler’s own estimate is that there is an unlimited amount of potential growth for VPNs because very few domestic firms currently recognize the need to protect their assets. But Veskler suggests, “this attitude will probably change, due to the increasing security vulnerabilities publicly acknowledged by even the largest of enterprises.”
But there is also a challenge regarding foreign owned and run VPNs on the mainland, as the Global Times recently quoted Fang Binxing (father of the GFW as noted above) that, “[u]nregistered VPN service providers are not protected by Chinese laws, and any company running a VPN business should realize they have a responsibility to register.” More directly, an employee in the Ministry of Industry and Information Technology pointed out in the same report that, “only Chinese companies and Sino-foreign joint ventures can apply to establish a VPN business.” This is not to say that is illegal to connect to a VPN outside of the mainland. Currently there are no laws which prohibit users in China from connecting to an overseas VPN.
In December 2012 I spoke with an American executive at a large IT company that provides dedicated internet connections to enterprises and institutions primarily in Tier 1 cities. According to him, “no foreign IT company and few domestic companies advertise their VPN services yet many of them will bundle it as part of a package to corporate clients. Furthermore, Chinese regulators typically permit VPNs so as long as it is privately – not publicly – accessible as well as the stipulation that consumers use leased-lines. A typical dedicated leased-line will cost over 3,000 RMB a month for 1 mb/s, this scales linearly (e.g., if you need 4 mb/s you are charged around 12,000 RMB), thus this option is typically out of reach by most consumers outside of the corporate and foreign communities. In addition, you can find a number of local firms that will provide point-to-point VPN services within the mainland. So if you are an expat that works for a foreign company that operates a VPN network elsewhere, then you will be able to securely connect from your local VPN to their secure environment overseas.”
Similarly, as an entrepreneur you can utilize these tech parks in China since they are not going to disappear overnight, if ever. For example, Larry Chang merged all company divisions under one roof in a research park located on a campus of a local college in Changning, Shanghai. His firm was provided incentives such as reduced rental rates for doing so. Similarly, Richard Qi mentioned that a new area in Shanghai called Cloud City – a tech park – provides perks and benefits to foreign software, engineering and IT firms. For example, Cloud City provides discounted office property, assistance in communicating with governmental organizations, stipends form the government and as the name-sake suggests, access to cloud services. Prior to relocating to this tech park, Qi mentioned that it was often difficult as a foreign service provider to issue invoices because of unclear laws (e.g., Shanghai and other municipalities are currently transitioning from a business tax to a VAT) and it was hard to find the government contacts needed to settle these transactions. In addition, perhaps your software or semiconductor firm can also take advantage of these inducements created by the 2011 policy which provides a tax holiday for several years, reduces the subsequent tax rates and provides exemptions on profits.
Takeaway: Due to a variety of regulations and policies on the mainland, certain telecommunication restrictions have germinated into a formidable barrier called the GFW. And with several million technologically-inclined consumers familiar with Western tastes and styles, there exists a potentially new customer base for VPN service providers. Yet just because there is potential for growth does not necessarily mean that the potential customers will purchase your goods and services (e.g., “if you build it, will they come?”).