The Blind Corner of e-Government

IZUMIDA Hirohiko
Visiting Fellow, RIETI

In March 2001, the IT Strategic Headquarters headed by Prime Minister Junichiro Koizumi determined an e-Japan Priority Policy Program that called for the "formation of the world's most advanced information and telecommunication networks" within five years, transforming Japan into the "world's most advanced information technology society." In line with the program, yearly e-Japan plans have been drawn up, with a certain portion of the annual government budget allocated for the implementation of relevant measures. For fiscal 2002, some ¥2 trillion has been allocated to promote the formation of an advanced information and telecommunication network society, up about 2 percent from the preceding fiscal year.
As of December 2000, despite huge amounts of disbursed or committed government funding, Japan ranked 14th in the world in Internet diffusion rate, slipping from 13th place the previous year, according to a survey conducted by the Ministry of Public Management, Home Affairs, Posts and Telecommunications. In another survey (conducted by Accenture in April 2002) on the progress toward electronic government, Japan was ranked 17th among 23 industrialized countries. Ideally, the government budget earmarked for measures to promote the establishment of an advanced information society would be spent on areas that provide indirect support for the innovation of spearhead technologies. In reality, massive amounts have been wasted simply to maintain and renew the conventional system, which heavily relies on a "mainframe" of well-guarded vested interests. The situation is akin to pouring water onto desert sand. The goal of realizing an IT-oriented state cannot be achieved simply by increasing budget allocations for policies and programs. The existing administrative system and governance structure must be overhauled to facilitate a transformation based on the latest technologies.

Enormous Electronic Networks Led by Government

Designed at a time when the now-obsolete value-added network (VAN) was in its heyday, enormous networks continue to serve Japan as mission-critical systems. As the name indicates, VAN is an information distribution and processing mechanism in which a network itself has added value. Composed of a single host computer and a number of terminals, the VAN system permits communications between joining personal computers, or terminals, through an intermediary, the mainframe. Under this mechanism, the "mainframe," a host computer that undertakes all information processing procedures, inevitably bears heavy loads at certain times. Ensuring adequate performance under such loads requires expensive facilities and special software, resulting in massive costs and time simply to maintain the network. Examples of VAN networks include the Magnetic Electronic Automatic Reservation System (MARS), a seat reservation system used at "Midori-no-Madoguchi (Green Ticket Counters)" at JR stations since 1965, when Japan Railway Group companies were state-owned Japan National Railways; the Nippon Automated Cargo Clearance System (NACCS) introduced in 1978; and the Bank of Japan Financial Network System (BOJ-NET) introduced in 1988. The Social Insurance Agency's Social Insurance Online System and General Automated System for Money (Postal Savings) are also VAN networks. In 1982, major revisions were made to the now-defunct Public Telecommunications Law, marking a departure from the principle of direct management of telecommunications services under state monopoly and allowing online networks to link to each other by private companies. However, due to historical inertia, gigantic electronic networks have continued to be built under government initiative. These networks remain a mainstream system in Japan.

Limitations of "Mainframes" that Support VAN Systems

Simply put, a "mainframe" is a large computer. The term refers originally to the cabinet into which a main central processing unit (CPU) was installed at a time when personal computers were nonexistent or considerably less powerful than they are today. Today, even insurance companies and banks - private sector entities that require extremely reliable computer systems - are beginning to shift to open platform systems based on UNIX or WINDOW operating systems, primarily due to the inferior cost performance of mainframe-type systems. The development of a mainframe-type system requires 10 times the money and time of open platform-type systems. The debacle resulting from Mizuho Financial Group's attempts to integrate its computer systems following the merger of three major banks has been attributed to the management's vacillating decision-making. But another major factor was the nature of the system integration required, which involved the complex linking of mainframes, despite the fact that the primary computational task executed by the systems was quite simple, data exchange. Currently, the government is proceeding with plans to convert complex port procedures for export and import, ranging from customs clearance and quarantine services to immigration control (reference chart), into electronic form, a mission that involves multiple government agencies. The move is meant to bridge computer systems operated separately by different government ministries and agencies in order to improve user convenience. Given ample time for systems development, it is unlikely that this integration will result in any system breakdowns. Nevertheless, many problems remain, since this effort also involves the linking of mainframe-type systems. Specifically, when relatively new systems such as the Justice Ministry's system for immigration control or the Port Electronic Data Interchange (EDI) system of the Ministry of Land, Infrastructure and Transport are linked with old-fashioned mainframe-type Nippon Automated Cargo Clearance System, the result is an awkward system capable of handling only fixed-length data. Cargo lists and crew lists exceeding the designated length must be split and sent in separate files. In terms of cost, some ¥90 billion has been spent on NACCS system alone, while some ¥2 billion will be disbursed to realize single-window port administrative services by linking various systems currently operated by different government agencies and ministries. These figures are comparable to the approximately ¥2 billion price tag of developing a similar single-window system called the Port Management Information System (PORT-MIS) in South Korea. Indisputably, Japan's cost performance is extremely poor. The global market for mainframe computers exceeded $12 billion in 1997, but has since rapidly contracted to $7.4 billion in 2001, roughly 60 percent of 1997 levels. More significantly, Japan's mainframe market increased 3 percent between 1998 and 2000, whereas the U.S. market plunged 42 percent (Dataquest) over the same period. In 2001, Japan and the U.S. switched positions, with the Japanese mainframe market reaching $2.3 billion and exceeding the $1.9 billion of the U.S. market. Why has Japan been so slow to shift from mainframes? It is often cited that Japan has far more mainframe vendors than other countries, and that these vendors have little incentive to withdraw from the mainframe business, which is quite lucrative when combined with related software business. But the biggest reason is that the largest user of mainframe computers is a cost-blind government. According to the Japan Electronics and Information Technology Industries Association (JEITA), overall shipments of mainframe computers in 2001 totaled ¥474.5 billion, of which ¥189.5 billion, or about 40 percent, were accounted for by government agencies and ministries. Here we see a mechanism that suppresses a shift to open platform systems, even when their merits are fully recognized.

Why Government is Slow to Shift to an Open Platform System

Let's list the reasons government agencies and ministries are slow to adopt open platform-type computer systems. First of all, government officials in charge of computer systems, who tend to prefer system stability to cost efficiency, are reluctant to make any changes, since they fear that any system breakdowns resulting from a shift to an open platform system might raise questions of individual responsibility. Second, the government has maintained "free contracts" with specific vendors - contracts with suppliers selected not through competitive bidding, but according to the procuring organization's preferences. This tendency is especially conspicuous with old systems initially developed when Nippon Telegraph and Telephone Corp. was state-owned. Under such circumstances, market mechanisms do not work properly. Third, since the VAN-type system model is based on the premise of charging information processing fees, various extra-governmental organizations have been established and remain in service for the purpose of receiving fee income from users. Fourth, some of these auxiliary organizations retain a number of supernumerary full-time employees of Japan Customs, and are thus inclined to avoid any changes that may provoke debates over the reform or abolishment of such organizations. As an example, let's take a look at NACCS, a core system for the aforementioned export and import procedures. The system, a closed network based on a design concept dating back more than 20 years, is operated by Nippon Automated Cargo Clearance System Operations Organization (NACCS Center). Since the center was given monopoly status under the Special Act of Customs Procedures through the Electronic Data System (NACCS Special Rules Act), market mechanisms do not apply when setting prices for user fees or in improving service quality. The NACCS Center receives large sums of state funds in the form of system use fees from Customs, from which the center accepts a number of personnel. Meanwhile, the center's system division - the core of the center - is provided office space inside a data center set up by a contracted company. Thus, there is no incentive for the NACCS Center to update its old-fashioned system. The result is the continuation of costly free contracts with the same contractors. Despite the likelihood of substantial cost savings, the presence of various vested interests and institutional disincentives has prevented a shift to an open platform system. Should the movement toward electronic conversion proceed without changes in the current system, users will end up paying fees that are far too high by international standards for administrative procedures previously free of charge when filed using paper forms. Chambers of commerce and industry, acting on behalf of private-sector companies and the U.S. government have filed a number of complaints, such as the following: 1) the current system is inconvenient, with users often having to reenter the same data many times; 2) user fees are exorbitant. (In fiscal 2001, the average cost per company amounted to some ¥7.8 million, with some companies paying more than ¥100 million.)

Direction of Reform

The world's preeminent e-government cannot be realized simply by increasing budget allocations. First of all, the government - the largest purchaser of information systems - must introduce market mechanisms into its procurement processes to improve system cost performance. Then, fee structures suitable for the era of the Internet, different from those in place in the era of VAN networks, should be implemented. Business models and administrative procedures suitable for the era of electronization need to be established and relevant institutional systems quickly established to benefit all of Japan. For example, in the above-mentioned case of NACCS, the core portion of customs clearance administrative procedures, currently handled by the government-authorized auxiliary organization, must be returned to the state. The rapid progress of information technology is now referred to as the "mouse year." To keep pace with such high speed changes, contractors for basic system designs and operations must be selected in competition to enable efficient execution of work procedures. Services that must be provided by the government should be separated from those that can be provided by the private sector. Services falling into the first category should be provided at the same cost as they were on a paper basis (free of charge), even after conversion to procedures handled electronically. Meanwhile, private-sector companies should be solicited to provide services falling into the second category to introduce market mechanisms. These services should be charged separately from government-to-private sector services. All of these users will significantly reduce or eliminate user fees, while a number of relevant entities, including small and midsize enterprises, can join the electronic networks. Promoting e-government and improving distribution efficiency would allow Japan to reinforce industrial competitiveness. The world is changing quickly. South Korea has already reformed (deregulated) institutional systems in line with the results of a business process reengineering survey that examined both business-to-government (B2G) and business-to-business (B2B) services, while companies are now required to submit all necessary documents to the government as electronic files. In this regard, South Korea has become more competitive than Japan. Japan faces the dire prospect of paying the cost, as an entire nation, of the snail's pace of improvement in work efficiency resulting from impediments created by vested interest groups, which hamper the introduction of the most advanced information systems. If the e-Japan initiative is to amount to anything more than a pie-in-the-sky proposal, Japan must thoroughly review its existing institutions and quickly improve operational efficiency, striving to reform the whole governance structure to promote the movement toward electronic administration and government.

October 15, 2002

October 15, 2002

Article(s) by this author