There was a time when the study of economics had a huge impact on society. In those days, economists were revolutionaries struggling against vested interests in order to win rights and secure livelihoods for the people.
Economists' ideas bore fruit in the form of the creation of experimental communities and states, including the Soviet Union, in the 19th century through the 20th century. They saved lives but also caused, albeit indirectly, deaths through revolutions that they inspired with their ideas, as all revolutionaries did. In addition, new vested interests were created on an enormous scale.
What are today's economists doing? They write papers read by only a dozen or so people and trade compliments and criticisms within small communities of researchers. Or they spend time talking at government advisory panels and on television programs and in other media. This state of affairs clearly marks the decline of economists.
The evidence of their decline is apparent in the textbook-style guides to economics displayed on the shelves of bookshops. The approach of those books is to analyze the economy as something mechanical and static. Economists are imitating the approach of the natural sciences, so to speak.
However, unlike the world of natural sciences, which is governed by unchangeable laws and principles, the economy is subject to changeable laws and principles. For example, Japan experienced runaway inflation immediately followed by a real estate bubble at one time, but later, the country was trapped in three decades of deflation or zero inflation despite massive printing of money by the central bank.
When we focus our attention on the fickleness of the economy, we have to ask and answer questions such as "How should we create the next change?" and "How should we create an economy and a market." If we take this approach, it becomes clear that economics has a role as an engineering science that helps to create a change and envision and design an economy and a market.
With this lofty aspiration and a sudden burst of despair at remaining within the ivory tower of academia to write papers read by few people, I have recently pursued increasing opportunities to work with companies, hoping for a breath of fresh air. I am working with around 10 companies, including CyberAgent, ZOZO, and Mercari, on various projects, including both large and small ones. These are projects and companies that form the basis for creating small but visible changes.
Spreading Fruits of Original Japanese R&D
The foothold for this activity is Hanjuku Kaso Inc., a company for which I serve as the representative. The strange company name (Hanjuku Kaso literally means "half-boiled virtual") is derived from our objective of envisioning a new society based on nascent technologies that have yet to mature. It is also an homage to "Hanjuku Eiyu" (Immature Heroes), a videogame masterpiece created by Square Enix, a great Japanese science and technology company, and is a pun on "hanjitsu kaso" (counterfactual reasoning), which is one of our research areas.
Overcoming time, space and language constraints, several team members are working across different time zones and the communication barrier between the Japanese- and English-speaking worlds. For my part, I divide my time between daytime hours, which are devoted to business projects and writing engagements in Japan, and nighttime hours, which are spent remotely performing my research and education duties at Yale University in the United States. I am exploring a new way of working that takes advantage of the global trend of digitalization accelerated by the COVID-19 pandemic.
Hanjuku Kaso Inc.'s business projects are underway in the following three stages. The first stage is carrying out original, cutting-edge Japanese research and development (R&D) projects and making the results open to the rest of the world. We are developing and putting into practical use technology and techniques for designing new projects and policies through a zero-based approach and forecasting what may happen in the new world following their introduction. To use technical terms, we are targeting a group of multidisciplinary techniques integrating methodologies of economics, such as market design, causal inference and counterfactual reasoning, with computer science.
In the fields of data and algorithmic sciences and technology, a small number of U.S. and Chinese mega-companies are strengthening their control over data and technology. Defying this trend, we aim to contribute to the world by making original Japanese technology that can meet any global standard openly available.
One example is our joint project with ZOZO, Inc. In this project, we first revised the recommendation algorithm that is used by ZOZOTOWN, one of the largest fashion e-commerce (EC) malls in Japan. As a result of applying the revised algorithm to some features of the home page of the ZOZOTOWN website, the click rate and the purchase rate have recorded double-digit improvements compared with the previous algorithm.
Moreover, aiming to bring benefits not only to our company but also to society, we have made our data and software openly available to the public. First, we made public more than 28 million sets of fashion-related recommendation data obtained through the practical use of our recommendation algorithm on the ZOZOTOWN website.
We also made public our software development infrastructure capable of forecasting the performance of the algorithm in an actual service environment and verifying the accuracy of the forecast. The act of making public software and data used in a large-scale actual service is almost unprecedented not only in Japan but around the world.
Traditional Industries and Public Policies
The second stage is using the developed technology to revive Japan's traditional industries and analog industries. Until now, the benefits brought by data and algorithms have been overly concentrated within a very small part of the economy, including the web-based and manufacturing industries. We would like to break this status quo and circulate the fruits of data technology through wider society.
From that perspective, we are implementing joint projects with, for example, Kubara Honke Group, which is handling a group of local food brands, such as Kayanoya, which is popular for its additive-free dashi (soup stock) and seasoning products, and with Gakken Group, which is a leading group in the field of education and publishing. The goal is to introduce and widely promote the approach of making data-based decisions, just as we may do when designing web services, in order to figure out what kind of foods and education services should be developed and when and at which prices the developed foods and services should be provided and to whom.
We have also concluded a business partnership with CyberAgent, Inc., which shares our perspective, and started to support retail businesses and public organizations in digitalization and data utilization efforts. We have also recently started to provide a solution service called Kakaku No Kagaku (Science of Pricing), which supports pricing by retail businesses.
Activities to address challenges related to everyday necessities (eating, clothing and shelter), retail trade, education, and healthcare take on the characteristics of social projects and policies as a natural course of events. We would like to contribute to the resolution of wide-ranging social and policy challenges while involving for-profit companies, and non-profit and public organizations in those activities.
For example, what effects can be achieved by the various policies taken against the novel coronavirus pandemic? We are implementing a project to publicly disclose forecasts of the situation of infection and economic conditions, together with the data and programs used to produce the forecasts, under counterfactual scenarios of implementing and not implementing those policies, and conduct real-time verification, among other projects. We are exploring ways of realizing a 21st century dream of digital automation of public policy development.
The third and final stage is drawing up a vision of society for the 22nd century. For example, Satoshi Nakamoto, who invented Bitcoin, has created a new economy worth tens of trillions of yen based on a research paper of just nine pages. With an information authentication algorithm integrating economics, cryptography, and computer science, he has realized the dream of an economy free from central government control. His approach is similar to the classic social sciences approach of the 19th century that is epitomized by Marx's reference to the role of economics in "changing," rather than "interpreting," the world.
A trend like this is not limited to the realm of virtual currency. Data and algorithms change the values and institutional infrastructure that underlie our society, including political and economic systems, religions and mass media. How can we rejuvenate existing political systems (e.g., the designs of elections, legislative, and administrative processes) and economic systems (e.g., capital market mechanisms), which appear to have become obsolete legacies of the past as they were built upon the norms and technology of centuries ago, into modern, 21st-century systems? Hanjuku Kaso is starting to announce plans and experiments aimed at solving that question.
The references in this article are as follows: The Worldly Philosophers (1999, Robert L. Heilbrone, for the opening statement.); https://arxiv.org/abs/2008.0714 (for examples of research and development); https://www.cyberagent.co.jp/news/detail/id=25506as (for an example of business implementation); https://forbesjapan.com/articles/detail/39053 (for the policy implementation example); https://globe.asahi.com/article/13857057 (for examples of social concepts).
>> Original text in Japanese
* Translated by RIETI.
January 23, 2021 Weekly Toyo Keizai