Framing the Historical Statistics Project: Questions and Answers with Professor Konosuke Odaka, Project Director

Compiling the Asian Historical Statistics Database

1. Setting up the 'Stage'

Q Why are you conducting yet another macroeconomic research project?

A Because no one has yet prepared long-term time-series statistics which would allow us to draw systematic comparisons among various countries, although we have the basic data for assessing their conditions. Compiling macroeconomic statistics will help us to know the real conditions of the whole of Asia, and also help us in evaluating the economic management of each government. Moreover, these materials will prove invaluable in letting us apply historical insights to the resolution of future problems.

Q What kind of 'data' are you talking about?

A We aim at presenting historical economic statistics and their background materials as systematically and mutually comparably as possible in order to understand the real nature of the national economies. The 'data' we are talking about is the statistical information suitable for the aim.

Q Could you tell me more concretely how the project works?

A First, we collect and order time-series statistics on macro performance of the national economies in the least arbitrary and most 'value-neutral' manner possible. Next, we statistically process and integrate the information we get according to a frame of reference common to the countries. We call our macroeconomic time-series data our 'databank' and the processed and integrated information our 'database.'

Q What is the common frame of reference?

A The main items are (1) gross domestic product (GDP), (2) gross domestic expenditure (GDE), (3) prices, wages and interest, and labor forces, and (4) money supply.

Q There are various ways of estimating GDP or GDE, aren't there?

A Yes. To arrange the necessary concepts, we use the SNA (System of National Accounts), which has been adopted by many countries. We plan to finally compile the databank and the database in the workstation and present the information both in printed books and on the Internet.

2. The Image of 'Production'

Q I'm beginning to understand what you're trying to do.

A To grasp what our research project is about, please consider population statistics, for example. Suppose we want to know, for a certain period and place, how dense the population was, whether it was increasing or decreasing, what the age constitution was, how many people in the productive age group participated in the work force, what was produced, and what the people's mode of living was like. Population statistics give us the basic information to know about such points.

Q So population statistics really convey a lot of important information.

A However, even for such important statistics, long-term series can't be easily prepared.

Q Can't you at least easily estimate the population in whatever country?

A Absolutely not. Without stable and strong political authority, we can't conduct reliable population censuses, the most basic and comprehensive population statistics. Then, it is natural that we did not have reliable nationwide population censuses in premodern times. And censuses are very expensive to conduct. Even if there are censuses, they are conducted only once every ten years or so; then, regarding the periods not covered by the censuses, we can only make estimates utilizing other statistics. So, we try to do our best in utilizing various related research on health, household registrations, births and deaths, police surveys, conscriptions, education, and so on.

Q Is it the same also in Japan?

A Of course, yes. The first population census in Japan was conducted in 1920, 15 years after the first census in Taiwan. If we want to know the population in pre-modern times, before modern-style population censuses were conducted, we can only make estimates using microlevel materials, called shuman arateme (religious registries), scattered around the country; the outcome of such estimates, which can be made only with the help of computers, is nowadays remarkable.

Q I see.

A However, in the case of population statistics, as the counting unit (one person) is universal and unchangeable, if the reliability of the data can be confirmed, there is little confusion or misunderstanding regarding the basic concepts.

Q I have thought for a long time that it does not seem practical to count large and small persons the same as 'one person.' For example, it doesn't seem fair to charge the same plane fares to a sumo wrestler and a marathon runner.

A That's right. For instance, for comparative analysis of nutrition intake, it is not enough just to think of the total number of the population. The differences of weight, age, and sex also should be considered.

Q Is it the same for GDP statistics also?

A Regarding GDP statistics, the situation is much more complicated than for population statistics. To compare various kinds of goods and services, a common value standard must be used. As the common unit, currency units, necessary labor units, or standard foodstuff units can be adopted.

Q The most usual one is the currency unit, isn't it? But there are many kinds of currencies.

A Yes. Different currency units must be converted according to the exchange rates. However, we should note in the beginning that using a common standard is meaningful only when various goods and services can be exchanged with each other.

Q Do you mean to say that such kind of analysis presupposes conditions in which the market system is working?

A Yes. We must be very much careful when we try to apply the concept of GDP to analyze societies in which the market system is not fully operational.

Q Even in a given market economy, new products appear, old products disappear, and even when the products stay the same, they are constantly improved.

A That is why it is natural if the comparative value of goods or services changes from time to time.

Q That is as if a measure keeps lengthening and shortening, isn't it?

A That's right. For that reason, the meaning of macro aggregate economic data may change every year. It is not an easy task to express the conditions of ever-changing economies and societies with simple statistical data, but I think that only assures us of the importance of substantive analysis.

A I'm really looking forward to seeing the results of the project.

Q Thank you very much.