Below are the provisional minutes of Forum #11, prepared by Secretariat of ESRI Economic Policy Forum (for details, see the whole record of the discussion in Japanese on our website).

   At the beginning of the discussion, Professors Murota and Miyagawa made the keynote speeches on the attached materials. A panel discussion followed, including questions from the general public and opinion exchange.
 

1.  Panel discussion
 

(Mizuno)

• In his report, Professor Murota mentioned Professor Jensen's hypothesis on why industry shifts do not succeed. I think this hypothesis perfectly delineates the situation in Japan for the past 10 years. Declining industries are saddled with excess capacity, but disposal of it does not go forward. As was pointed out by Professor Miyagawa, Japan's financial market lacks liquidity.
   
• A comparison of the Japanese and American stock prices with a 10-year lag points toward an interesting fact: recent stock prices in the United States are similar to Japanese ones in 1992. However, the market signaling is different. In the United States, productivity bended and began to rise in 1995. Judging solely from the data on non-manufacturing industries, productivity in these industries has been rising recently. The fact that productivity grows irrespective of the present recession indicates substantial changes in the real economy. On the other hand, the productivity of the Japanese manufacturing industries is closely connected with their output and, thus, it only fluctuates with the business cycle.
   
• In the United States, the aggregate market value of listed stocks reached the peak of 1.9 times GDP, and then shrank to the present level of 1.1 times GDP. In Japan the aggregate market value was 1.4 times GDP during the bubble period, and is now 0.5. The present situation in Japan is similar to that of the United States before the shift to the New Economy before 1995. Besides, the present aggregate value is the same as the average value in the United States in the 20th century. The United States made a breakthrough when it stepped toward the New Economy. In Japan, however, the stock market seems to determine that this did not occur here.
   
• Judging from the data on IT investment in Japan, the order volume to the manufacturers of IT is big, while the order volume from IT users remains low. The difference is exported. IT does not contribute to changes in the society or the productivity of the Japanese economy.
   
• While in Japan employment is not expanding, in the United States both employment and productivity are growing. Especially, employment in the computer-related and data processing sectors has been increasing at the annual rate of 12 percent since 1995. For seven years number of workplaces increased by16 million, 7 percent of which were created in IT-related sectors. The Dow Jones Industrial Average stands at approximately 8,000 today. It is necessary to note that the productivity growth caused by IT greatly contributed to the increase in Dow Jones Average from 3700 to today's level.
   
• I agree with Professor Murota that business cycle fluctuations will become sharper because of advances in IT. In the United States, the investment based on the estimated middle-term demand exceeding the potential growth continued for 5-6 years and, thus, led to excessive production capacity. However, further productivity growth may entail an increase in profit ratio.
   

(Nishimura)

• The productivity increase in the United States is usually attributed to technological progress. However, it is necessary to keep in mind not only the fact that the real economy has changed, but also the fact that data compilation has been altered. The IT-related deflator has been changed significantly. Comparing Japan and the United States is difficult due to the difference in data compilation. I think that Professor Miyagawa's estimates follow the golden mean, but it is necessary to be careful as there are extreme cases in the U.S. methods.
   
• Software can be classified into three types: in-house developed, custom developed, and package software. In Japan, however, only statistics on custom software are included in capital investment. In the United States all three types are included. Although there are properly compiled statistics on package software in Japan as well, their treatment is overly limited as compared to the United States. In-house software is estimated on the basis of costs, but I think that either Professor Miyagawa's estimates or the U.S. estimates are overvalued.
   
• In regard to statistical usage, it is essential to pay attention to how these statistics are compiled. Otherwise, it is easy to make a mistake in the following two points. The first is how to interpret the actual situation, and the second is how to interpret the response of the market in respect to the incorrect statistics.
   
• Computer software or hardware has overcapacity or excessive quality. Nevertheless, the price indices are compiled on the premise of no idle capacity. In reality, two-year-old computers function quite well in most cases, so I doubt whether the leading-edge capacity is necessary. Since the neo-classical school assumes away the existence of idle capacity, U.S. economists such as Jorgenson or Stiroh believe that new software will be put to practical use of the full capacity some day in the future. It is necessary to bear in mind that the data are compiled in conjunction with this way of thinking.
   
• Cautious optimism concerning the present U.S. economy prevails. It is the first time that the above-mentioned problem of excessive quality has been raised. Many unknowns remain. Hewlett-Packard has shifted one of its software development centers to India. Maybe this is an indication of hollowing-out of U.S. IT-related industries, beginning with software development.
   
• Moreover, since in the United States the period of high growth and the period of active investment in IT were concurrent, the seeming correlation may be interpreted as "technological progress" in the growth accounting. While in the United States there was an increase in the productivity of non-manufacturing industries such as financial and insurance sectors or wholesale and retail businesses, such productivity growth is not seen in Japan. It is difficult to interpret this result. Certainly, specific Japanese problems may account for this situation, but the growth of the U.S. non-manufacturing industry is also highly likely to be cyclical as pointed by Professor Gordon. It is rather in Japan where there is significant progress in the utilization of IT such as usage of POS data, but the productivity increases in the wholesale and retail trade sectors have not been observed.
   
• I have a pessimistic outlook on Japan because the current technological innovations have a negative bias with respect to traditional Japanese advantages. The Japanese-style architecture integrated from the upstream to the downstream suffered a devastating blow from modularization that became even more integral along with IT progress and globalization. At present, it is only the automobile industry that is on the competitive edge. The main reason is in its integrated architecture. IBM's mainframe computers used to be the representative of integrated architecture in the electronic machinery industry. Nevertheless, as soon as the Japanese producers got a lead on IBM, the central technology shifted into the fields of personal computers and distributed computing.
   
• The new networks using IT made conventional networks such as human relations and prior negotiations outdated. Along with this, human capital specific within the companies lost its value. The market value of old capital equipments became zero due to the IT progress, and this gave rise to idle facilities. At the same time, human capital became obsolete. This resulted in unemployment. For this reason, even if there is additional demand, employment does not grow and the economy is at underemployment equilibrium.
   
• What is alarming is that the Japanese economy has continued to show signs of fading of the market over the past 3-4 years. Till the first half of the 1990s the companies with low TFP (total factor productivity) growth left the market. In 1997-98, the opposite happened: companies with high TFP growth rates left the market. Such a strange phenomenon is taking place in the Japanese economy. Something parallel to the advances in IT is occurring behind the curtain.
   
• The project "e-Japan" paints a rosy picture of Japanese society because of the advances in IT, but I don't agree. Government's officials advocating the Project 'e-Japan' do not use e-mail. It seems doubtful that Japan will be able to keep pace with IT progress. In fact, if we fail to create the framework under which the people actually using IT will change the bias of technological innovations toward a favorable direction for Japan-or at least away from a detrimental direction-Japan will find itself in dire circumstances indeed.
   

(Murota)

• It does not mean that everything about the pricing of software is known in the United States either. Researches have been conducted step by step, and the results are utilized. The measurement problems are not restricted to software. For example, the output of bank services has long been discussed, but no conclusions have yet been reached.
   
• Overcapacity is a reality, but it is precisely the key of future development from the supply-side viewpoint of software. If computers are treated as one of commodities with prices declining to the level of 100,000 yen, much more people will be able to use computers. It is possible that there are geniuses such as Linus Torvalds or Bill Gates among them, who can use 100 percent of the computer's capacity and may succeed in innovation. Therefore, if even a small number can make full use of the computer, it does not matter if there is excess capacity for the rest of the users.
   

(Miyagawa)

• It was said that the pattern of U.S. stock prices resembles Japan's in 1992. If appropriate policy had been implemented then, the situation would have been different. The future development in U.S. economic policy is a matter of deep interest in this sense, while its economic policy team has changed.
   
• In regard with the statistical problems, I agree with Professor Nishimura. The U. S. method is not necessarily good. According to OECD papers, the prices of IT products of the member countries are so different that there was an attempt to adjust them. It is necessary to make a decision after recognizing the limitations of the data.
   
• The present situation in Japan, except deflation, is similar to the situation in the United States in the beginning of the 1980s, which was characterized by high unemployment and a high rate of bankruptcies. That time the biggest task was promoting productivity growth. In the second half of the 1980s, General Motors adopted the just-in-time system developed by Toyota. Today, Japan needs to incorporate business models that focus on IT as a tool. Even more important is the direction of Japan concerning what industries will lead the economy. It may be not to follow the United States but to create new products and new style of production.
   
• Professor Prescott raised the question of why the 20-30 times gap in GDP per capita exists, notwithstanding the world wide spread of technologies. I think this situation can be explained whether new technologies are adopted or not. During the Meiji Period (1868-1912), cotton manufacturing industry expanded in Japan but it did not grow so much in India, though the economic situations were similar. This difference is mainly attributed to the fact that in Japan young female operatives quickly adapted to the new technology, in contrast with India, where adult male operatives were engaged in and delayed the adoption of new machineries. Likewise, today Japan falls behind in acceptance of new technologies. The majority of the customers of e-business are in their 30s and 40s. But most financial assets are held by people in 60s. If we do not overcome this difference, we will not be able to hope for a recovery like the one experienced in the United States.
   

(Mizuno)

• It is pointed out that Japanese money-losing enterprise do not withdraw from the market smoothly. In this connection, the estimates of the growth rate of value-added of non-manufacturing industries show that input volume in real assets by the service sector has been big since the beginning of the 1990s. Under normal circumstances, advances in IT are supposed to reduce the necessity of real assets such as stores. However, in reality the non-manufacturing industries have followed the same pattern of development as the manufacturing industries during the high growth period in the 1960s. This is a problem of IT using industries, but this also provides a key for Japan to come out of the period of low growth.
   
• It was said that the networks of Japanese companies became outdated. In this conjunction, the drop in sales is concentrated on the wholesale and construction industries. The wholesale industry is a network industry, tying manufacturers and retailers with information. Employment problems are serious in these industries. It is necessary to take the plunge and change even the points that have been advantages in the past.
   

(Nishimura)

• Construction is a network industry, with contractors and subcontractors, like the automobile industry. However, the value of this network became zero, and the impact has been immense. The automobile industry was able to change substantially under the pressure of market competition. But the construction industry is protected by the Land, Infrastructure and Transportation Ministry and does not make reforms. There were times when Toyota was a small business and Honda did not exist, but the large-scale construction corporations have not changed. If such an industry enters the market even partially, its fragility will become apparent.
   
• The software industry often conducts outsourcing, which involves some kind of modularization. This is supposed to raise productivity, but in practice the opposite sometimes takes place. For example, assume that a software company wins contract from an administrative institution with a bid price of one yen. It is usually believed that if it leads to the subsequent receipt of orders, the business will be profitable. But there are cases that the company has to accept endless requests from the administrative institution, which leads to an increase in costs and, as a result, losses. It is essential to define clearly the scope of what can and cannot be carried out by modularization, and to have customers to pay for the latter case. IBM is doing that in the solution business, but it seems to be difficult for Japanese companies. The top managers who give orders do not have enough knowledge about modularization, and that is why they give strange instructions. The contractor also tends to accept it. Further modularization leads to the concentration of ultimate authority in the hands of top management. In such a case, it is fatal if the top managers can't even use e-mail.
   
• The main reason for the low productivity of the financial sector and other service businesses is excessively expensive software. Most companies didn't think of modularization and developed almost the same software, which incurred high costs. If the project "e-Japan" follows the same pattern, the situation will become even worse.
   
• I think there will be demand, but the problem is whether this makes a business model work out, in other words, whether payment accompanies the demand. Considering P2P, it is clear that as long as we do not make tighter regulations, information will become charge-free on the Internet, and this would not come into existence as a business model. In that case, employment will not increase. I think it is difficult to create demand like in the United States, where profits through initial public offerings are common. What is rather promising is the demand creation within the framework of an NGO-style corporate model.
   

2.  Open discussion with audience
 

(Audience)

• I think usually there is a trade-off between productivity and employment. Nevertheless, according to an OECD survey, both productivity and employment are growing only in the United States compared to Japan and Europe. This may be a kind of paradox. How much has IT contributed to demand creation? Has IT given rise to new demand or new business in the service sector? I think that consideration of these issues is useful for future policy-making in Japan.
   
• I would like to ask a question in connection with the pessimism which was expressed concerning the future prospects of Japan. In the 1990s the United States swept the board as a front runner. I think it's possible that other countries will catch up and threaten the position of the United States in the years ahead. How do you think about that?
   

(Audience)

• Let me ask a question to Professor Miyagawa. While decreasing in-house and custom-developed software, companies may shift to package software, which will decrease the costs and increase productivity. Don't you think so?
   
• My second question is for Professor Nishimura. If IT industry is an increasing returns industry, this is good for society, but companies will not be able to make profits. What do you think about this issue?
   
• A question for Professor Murota. You said that Japan is lagging behind. But Japan is good at catching-up. So isn't it possible to expect that Japan will do better in the future?
   

(Miyagawa)

• Software, including that developed in-house, reached a peak in 1992. And as I said earlier, this is a strange phenomenon as compared to other countries. I do not think that software investments were made effectively in Japan. As a result, it has not contributed to productivity growth. If we had clearly defined the scope of what can and cannot be achieved by modularization, it would have contributed to productivity growth. In the 1990s there was a movement toward a change in redundant investments, but it has not been enough.
   
• Concerning the first question, the United States has been creating new businesses such as booking and fund transfers through the Internet and so on. If the demand side accepts the new technology, it will match the supply side and this, in turn, will give rise to new demand. In Japan demand and supply do not match now. This mismatching is not a specific Japanese problem: in the 1980s the American automobile industry failed to respond to consumer demands for fuel-efficient cars.
   

(Nishimura)

• The increasing returns to scale are supposed to generate monopolies as you have pointed out, but in practice companies face sharp decreasing returns. Information can be almost perfectly copied. The reproduction costs are zero when interacting through the Internet. In this case, the market price becomes zero and a business model will not be formed. Every company desperately seeks places to add value. What can never been done by IT is time-related things. On-demand service is valuable: it supplies what you want just when you want it. This is a part of decreasing returns. Apart from this, a business model will not be formed as long as there are no restrictions. There has been significant progress in copy-protection technology, but this protection is usually cracked within about six months of its introduction. Even in that case, it is possible to make founder's profits. Such a rat race is going on in the open backbone Internet. Even if it is impossible to form a business model, it is good to push NGO-style business. From the standpoint of employment or social welfare, IT is extremely useful and beneficial.
   
• Concerning the first question, the reason for the U.S. economic recovery is the boom of the following three demands. First, a large amount of IT-related investment was implemented. Second, new financial instruments such as MBS were developed and the financial revolution has been successful. Finally, the SUV, developed on the basis of the architecture of a modularized truck, became a big hit. IT played an important role in the 1990s as investment demand, but recently the impact has become much smaller.
   

(Murota)

• Japan can hardly even catch up. In contrast with automobiles, new types of cellular telephones are developed every two weeks. Nokia can do that, but the Japanese producers are not able to keep pace with these rapid changes. It is significant that Japan could not take the global standard like GSM. Besides, the business structures are not such that it is possible to keep abreast of technical development. Even if the software industries of India or China can catch up with the United States, Japan's industry's catch-up would be too little and too late due to the way of business.
   
• It was pointed out that business models in connection with IT are difficult to set up. In this regard, there is a hypothesis that it is consumers who receive the full benefit of IT advances.
   

(Mizuno)

• In 1995, the productivity of the U. S. economy showed a sharp rise. At the same time, demand exceeding the potential growth rate emerged. According to the data at that time, personal consumption increased by approximately 4 percent and the saving ratio turned out to be zero or negative. This became possible because of a shift toward the New Economy as well as the mechanism that prevented interest rate rises by utilization of foreign savings, even in the case of a huge current account deficit. Besides, consumption was also raised by the asset effect. In Japan a negative asset effect, combined with uncertainty about future income, restrains consumption. To spur consumption it is important to allay anxieties about the future.
   

(Audience)

• IT investment is believed to increase productivity in the ordinary sense. In addition, IT can be considered to increase the effectiveness of research and development through CAD, simulations, and so on. The former effect disappears after the investment is over, while the latter effect seems eternal. To what extent is the latter effect important? In which field does Japan lag behind the United States?
   

(Murota)

• Concerning this point, I would like you to look at page 17 of my handout quoting Cummings and Violante (2002). It seems difficult to give a definite conclusion.
   

(Nishimura)

• From a macroeconomic perspective, such an effect will appear positive if the estimate includes IT-producing industries, but I think this effect varies among companies to a large degree.
   

(Miyagawa)

• I have not tried to treat research and development as output and IT as input, but this idea is extremely interesting.
   

(Moderator)

• Finally, I would like each panelist to summarize what we should do in the future.
   

(Murota)

• It is important to facilitate an exit from the market. The government should not turn its hand to IT-related matters but should, for example, purchase and dispose the excessive capacity of the other ineffective sectors.
   

(Miyagawa)

• I have written some conclusions at the end of my handout. I would like to add that it is important to revitalize the economy using IT. It is necessary to change working patterns using IT, to increase the speed, and to entrust the job to the young professional instead of making decisions at the top of hierarchical organizations.
   

(Mizuno)

• To create an environment where savings can shift to consumption and investment centered on IT-related goods, it is necessary to eliminate the anxiety people have about the future of pensions through the institutional reform. If people understand the maximum level of risk, they can decide the necessary amount of savings and use the rest. In addition, it is important to determine the rules of the exit from the market and to facilitate the smooth shift of resources.
   

(Nishimura)

• The negative legacy of the 1990s amounts some hundred trillion yen. Despite the fact that there is no perspective on its solution, discussions are conducted here and there as if there is no such problem. If the market hits the lowest point, it moves to recovery, but that point has not yet been observed. Because the market does not function properly, the government should not leave the problem to the market but play the role toward its revitalization.
   
• Judging from the 10-year experience, a new business model on the basis of IT has not been formed. The case of Microsoft is a traditional model of monopoly. It is important to establish a new business model, a corporate model. This will lead to the reactivation of human capital which has already disappeared. The government should play the role in strengthening the functions of the market.
   

(End)