Konosuke Odaka : How many countries does it cover?
Kuroda : 13 countries. America, Britain, France, Germany, China, Taiwan, South Korea, the Philippines, Malaysia, Singapore, Indonesia, Thailand, plus Japan. It partially overlaps your project.
Odaka : Are you making time series tables?
Kuroda : We're conducting horizontal comparisons of the I-O tables for three years, 1975, 80, and 90.
Odaka : There has to be a standard sectoral classification, right? How detailed are the classifications you are using?
Kuroda : The common classification tables are 183 x 183. We use Japan's basic I-O classification. First we make a 2-country bilateral table, then we expand it into a multi-country comparison table. But, when making a 13-country comparison, it is reduced to a 24 x 24 classification.
Odaka : Can you use your I-O tables for policy evaluation?
Kuroda : Making I-O tables in real terms has been the most important immediate goal. As soon as we have obtained this objective, we will be able to conduct policy analysis, too. We are straining ourselves to make a multi-country production deflator (a type of PPP: Purchasing Power Parity) in order to compile a real table. It will include energy statistics as well.
Odaka : You're very ambitious.
Kuroda : We would like to link up with the Asian Historical Statistics Project, at least partially. How do you plan to release the results of your project?
Odaka : We plan to put out publications and use the internet.
Kuroda : Japan hasn't thought seriously about copyrights, and that's a problem.
Odaka : Has that already caused some sort of problem?
Kuroda : The I-O table for Japan made by the KEO (Keio Economic Observatory) Project has been posted on the internet, and it gets about 30 inquiries a day. They receive different types of orders from all areas.
The data is provided free of charge to researchers whose goals are clear, but if it gets misused, it kills the motivation of the people who compile it.
Odaka : What kind of things are you worried about?
Kuroda : For example, using data incorrectly and getting strange results, or using the data to try to prove the presence of perfect competition even though the assumption of perfect competition was already made in constructing the data.
Odaka : Such problems are, I think, the responsibility of the users to resolve. To some extent, it's inevitable that things like that happen.
Kuroda : But if the data are blamed when there are strange results, and people complain that the data are unreliable and so on, the compilers are seriously discouraged and lose motivation.
Odaka : There is also the problem of collecting fees.
Kuroda : There are problems no matter how you collect fees, but it is possible to have someone take care of that. If you charge a fee, the users might be more likely to appreciate the significance of the data.
Odaka : The Ministry of Education's opinion is that even a national university can collect fees. However, the university's business affairs would get more complicated.
Kuroda : What's more, it's difficult to come up with an estimate of just what degree of value added there is in our data. Also, it is not always clear what parts of the data are copyrighted.
Odaka : If the original sources materials are public government statistics, then the results are in the public domain.
Kuroda : Even if you open it to the public, just how far do you open it? In our case, should we put out the final I-O tables only, or should we also go so far as to make available the intermediate estimation process or the raw data?
Odaka : For starters, it's enough just to make available the final results, I think.
Kuroda : I agree.
Odaka : Still, you want to provide as much explanation as possible about the methods for compiling estimates and the nature of the estimated values.
Kuroda : We just published a new manual, KEO Database: Production and Capital-Labor Input Estimates,1 which is supposed to serve just that purpose. Incidentally, it is possible to assemble the documentary information and put it on the internet.
Odaka : A researcher wouldn't be satisfied with just the final result.
Kuroda : In that case, how about having prospective users file applications for the use of the data, giving their reasons for needing the data and their objectives, and when whoever examines their application "OK's" it, they could go ahead.
Odaka : Good idea. But it's quite a bit of work for whoever does the examining.
Kuroda : As with the issue of fees that you talked about before, I think it would be all right to entrust the supervision and use of the completed database to a private think tank or a third sector organization.
Odaka : That might be a good way to get around the problem of fee collection and other related administrative matters. How would the database be maintained?
Kuroda : That's one problem. Unlike governments or consulting organizations, researchers don't usually continue working on the same topic forever.
Odaka : If we had an intermediary organization be our gateway to the world, then we could raise funds to cover the maintenance and other operating costs for the database. At the same time, we could also use the funds to pay for ongoing projects which further the development of the original database. But what would become of our relationship with the Documentation Center?2
Kuroda : There are a lot of problems that need to be dealt with, but I think we'll manage it.
This meeting took place at the Department of Commerce, Keio University, in Tokyo, on April 1, 1997.
Masahiro Kuroda is a professor of economic and econometrics at the Department of Commerce, Keio University, and serves on the Advisory Committee for the Historical Statistics Project.
(1) Kuroda Masahiro et al., KEO deetabeesu - sanshutsu oyobi shihon-rodo tonyu no sokutei (Keio Gijuku Daigaku Sangyo Kenkyujo (Keio University Economic Observatory), February 1997). This book is not for sale commercially.
(2) The official title of the organization is the National Center for Science Information Systems (Gakujutsu Joho Sentaa). It is affiliated to, and directly supervised by, the Ministry of Education.