Columns that reveal the world
- Getting up close and personal with the researchers -

Demography, which I study, is an academic field that aims to solve population problems occurring in Japan and around the world by observing population trends not only at birth and death but also at each life event, such as continuing education, finding employment, marriage, retirement, and the associated movements. In particular, I specialize in regional demography. I use official statistics such as the national census to analyze population distribution and movements, and develop mathematical models to calculate estimated values.

Demography is related to all of the global issues raised in the SDGs, such as poverty, food, resources and energy, and the natural environment. It is humans who carry out economic activities and cause environmental problems. It is no exaggeration to say that demography, which observes humans in groups, is a field that is directly linked to solving the problems of the SDGs.

My interest in demography and geography dates back to my childhood. At that time, there were no computers or the Internet, and I was fascinated by looking at the world and Japanese population statistics published in yearbook supplements and atlases. I was a boy who could be called a "population geek" who would rank Japanese cities by population and write them down on paper, wondering why the population of each city was increasing so much. When I became a junior high school student, I also became interested in mathematics, and began to think about becoming a mathematician or a geographer. At first glance, the two fields of study seem to be unrelated, but in geography, mathematics is used for statistics and data analysis. I studied statistics in college, and in graduate school, I majored in population geography, which is a field of geography that often deals with mathematics. So I ended up majoring in the field of demography that I was passionate about as a boy. My current specialty, regional demography, overlaps with population geography in many ways, and in any case, I believe that a research job dealing with regional demographic statistics was the right job for me. I feel that I was extremely fortunate to have had the opportunity to pursue my calling at Aoyama Gakuin University, and I am deeply grateful to them for giving me this opportunity.

In recent years, I have been working on research into estimating future populations by small area (town and village) for the next several decades, and building a web system for this purpose. I began this research because I realized that when the national and local governments formulate policies to deal with the declining birthrate and aging population, they need population estimates by small area. Until now, the smallest division for official regional future population estimates was limited to the city, ward, town, and village unit, and as areas have become larger through mergers of cities, towns, and villages, there has been a growing need for estimates at smaller units. However, no such estimates have existed anywhere in the world until now. The reason for this is technical challenges. Generally, the cohort change rate method (a method of estimating future population based on the population change over a certain period of a group of people born in a certain year = cohort) is used for future population estimates. The biggest problem with this method is that the smaller the area being estimated, the more unstable the statistical indicators become, and the estimates can fluctuate due to minor factors. Furthermore, the longer the period covered by the estimate, the greater the deviation; in some cases, the estimated population 50 years from now can be 10 to 100 times the current figure.

To solve this problem, demographers and statisticians have proposed various smoothing methods. For example, the spatial average method smoothes by taking the average value of the entire surrounding area including the target area as the value of the target area. However, the drawback of this method is that it does not take into account the distance between the target area and the surrounding areas.

Therefore, I proposed a new method that employs the idea of "population potential," a classic theory in population geography. Population potential defines the sum of the influence of surrounding regions on region A, based on the principle that "the demographic influence of region B on region A is proportional to the population of region B, and inversely proportional to the distance between regions A and B." In this way, I was able to create an equation in which the closer the region is to the target region, the stronger the influence, and the farther away the influence, the weaker the influence, making it possible to estimate future populations by small region.

The smoothing method that utilizes population potential was fully presented at the 8th International Congress of Population Geographers held in Brisbane, Australia in 2015. Although it was difficult to present in English, the idea was highly evaluated and he won the poster award. Furthermore, with the cooperation of ESRI Japan Co., Ltd., the estimation results were made available free of charge on the web as the "National Subregional Future Population Projection System," which not only drew a positive response from researchers but also inquiries from local governments and infrastructure-related companies. Figure 1 shows an example image of the system.

Figure 1. Changes in the aging rate (proportion of the population aged 65 and over) in the Tokyo metropolitan area (upper row: 2020, lower row: 2065)

Note: Created from screenshots of the National Small Region Future Population Projection System ver. 3.0

*As of 2020, light blue (10-20%) and yellow-green (20-30%) areas were prominent, but by 2065, yellow (30-40%) and orange (40-50%) areas will be more prevalent, indicating that the metropolitan area will experience rapid aging.

The "National Small Region Future Population Projection System" can display population increases and decreases by age and sex for small areas, so it is being used in a wider range than I had imagined. For example, such small region future population projections are very useful for planning the layout of water and sewerage facilities, and it will also be useful for building commercial facilities such as department stores, as it will show in detail how the population of the surrounding commercial area will change. I think it can also be used when making urban plans for the construction of elementary schools and Kindergarten Building. I myself am conducting several applied research projects, such as examining the risk of future uninhabited areas for each small region and predicting the aging of areas expected to be inundated by flooding (Figure 2).

Figure 2. Changes in the aging rate across Japan by region expected to be inundated and non-inundated due to flooding

(2010–2060)

Source: Excerpt from Figure 9-4 on page 198 of Inoue and Wada (2021)

*Created from data uploaded to the "National Small Region Future Population Projection System." The numbers on the horizontal axis represent years (last two digits in the 2000s). A chart like this cannot be created without population estimates for each small region. This chart shows that the aging rate for the whole country will exceed that for areas expected to be flooded compared to areas expected not to be flooded from 2045 onwards.

We are currently working on building a web system that will target countries and regions other than Japan, and have already released similar systems for Washington State in the United States and Taiwan. (Figure 3: An example of the Taiwanese version screen) In the future, we plan to first develop a system targeting Pacific Rim countries, and then share the results with each other, aiming to apply it worldwide. There are many issues that need to be resolved, especially in developing countries, where detailed population data like Japan is not available, but we will continue to work hard to develop new methods and build a system that will be so effective that people will say, "When it comes to small-region population estimates, it's Inoue."

Figure 3. Initial screen of Taiwan's small-area future population projections web mapping system

Note: Created from screenshots of version 1.0 of the same system

*All information is in English. As with the Japanese version, it can display aging rates, population density, average annual population growth rates, etc. by small region.

This research has received a lot of attention, but in a highly specialized academic field, it is not easy to leave a social impact. Basic research such as developing mathematical models is a process where small discoveries are gradually accumulated, leading to big results that can contribute to the future development of humanity. In my everyday research life, I look at my subjects from various perspectives and value my discoveries.

When creating new methods, it is common to hit a wall and not make progress. At times like these, instead of concentrating at your desk, always think about the problem in the back of your mind from the time you wake up until you go to bed, and ideas will come to you at random times, such as when you are taking a bath or spending time with your family. The most satisfying part of research is when you announce a new population indicator that no one has thought of before.

I have pursued various academic fields that interest me with the aim of becoming a researcher. I encourage all students to dive fearlessly into the "ocean of knowledge" that is university. Aoyama Gakuin University has experts in a variety of fields, so your intellectual curiosity is sure to be satisfied. If you can set out on the ocean of knowledge, you will be able to reach the depths of the academic world. Just take the first step. (Published October 2021)