OSAKA – When times are good, there is less political pressure at the local level anywhere to be economically efficient or carefully scrutinize predictions that a new public works project or expensive industrial or tourism promotion scheme will lead to prosperity in 20 or 30 years.
But with their rapidly aging and declining populations and shrinking tax bases, local governments now face a daunting task in formulating political, economic, social and environmental policies that will most likely benefit the greatest number of people decades from now. In contrast to the carefree public works spending of the bubble economy of three decades ago, often based on proposals that seemed little thought out, the demand for data-driven, evidence-based projections for various policy measures among local governments has grown, lest a wrong decision lead to local economic disaster, and voter anger.
Earlier this year, Nagano Prefecture announced it would rely more on computer modeling and scenarios for local policy decisions. The decision came after the prefecture cooperated with Kyoto University’s Kokoro Research Center, Hitachi Ltd. and Mitsubishi UFJ Research and Consulting to create two different models using artificial intelligence. Those were put to use in research on the best policy to realize a sustainable society and how to best take advantage of the opportunities, especially related to local tourism, that might come from the planned opening of a maglev shinkansen station in the prefecture as early as 2027.
Nagano Gov. Shuichi Abe and Kyoto University’s Yoshinori Hiroi, a professor at the Kokoro Research Center in charge of the project, explained the system at an April news conference.
The research flow, Hiroi said, basically consisted of three stages. During the first, people collected information, decided the two issues to pursue and created the computer models to address both.
Data based on the current situation and projected economic and demographic changes was then fed into the computer, which created different scenarios of what a sustainable future might look like and what a future maglev might mean. The numbers were crunched and different possible future scenarios based on those results were analyzed by researchers and put into a policy proposal.
“It’s kind of like a sandwich, with humans forming the first and last parts of the stage and AI in the middle serving as a tool,” he said.
Further work critiqued each possible policy based on how it would impact various areas: local industry/income, employment, tourism, the environment, population, health, education and the local community.
“We came up with six scenarios for obtaining a sustainable society in these areas, including one that was judged by the computer to be the most balanced,” Hiroi said.
The computer modeling also ran simulations and produced six future scenarios for a maglev shinkansen passing through Nagano, taking into account local industry, farming, and younger generations, as well as international tourism spending projections, the local economy, population and transportation infrastructure. The result was a recommended policy that balances these different concerns and, it is hoped, prevents the project from being a huge white elephant locally.
“Naturally, we have to promote policies that strengthen the local economic base and local communities so that the result (of a policy to increase the effect of the maglev’s opening) is one that doesn’t invite their hollowing out,” Hiroi said.
Nagano’s experiment with AI in the policymaking process claims to be the first of its kind. But other local governments have expressed interest in AI-influenced policymaking. Last year, the city of Maniwa in Okayama Prefecture announced it would team up with Hiroi’s research lab to do similar predictions for what its future might look like. Ogaki in Gifu Prefecture is also looking into greater adoption of AI to address policy issues. The city of Otsu in Shiga Prefecture earlier this year began using AI to analyze bullying cases reported in the past to help teachers detect signs of serious school bullying.
The idea for utilizing AI to help local governments formulate more effective sustainable development plans for the future was outlined in a joint proposal by Hitachi and Kyoto University in September 2017. They noted Japan faces an aging society and declining birthrate, and problems including social security, efficient use of environmental resources and eliminating socioeconomic gaps.
The Hitachi-Kyoto University collaboration generated some 20,000 possible scenarios for Japan over a 35-year period between 2018 and 2052 — the scenarios were later the basis of further experiments and refinements that produced the six specific scenarios for Nagano’s future. In 2017, prior to Nagano’s decision to give AI policymaking a try, Hitachi and Kyoto University produced a set of general policy recommendations for the nation as a whole.
The 2017 recommendation is based on one of two population models for Japan in 2052. In the first, most industry and technological innovation is concentrated in a few large cities, to which the population has shifted, leaving the regions abandoned. It also predicts individual health and happiness has decreased, and central government expenditures are concentrated on large cities.
In the second model, regionalization has taken place, the birthrate has recovered and human health and happiness have increased. However, both government finances and greenhouse gas emissions have worsened.
The recommendation adds that in eight to 10 years, changes in the demographic makeup mean Japan will reach a turning point and will have to choose either the urban concentration model or the regionalization model. Realizing what is considered to be the more sustainable model — the regionalization model — means effective policies for promotion of local communities and culture, strengthening local transportation and making better use of renewable energy, which will help local economies.
Thus, in order to have a sustainable, regionalized Japan by 2052, the proposal says, effective policies in these areas need to be enacted within the next 17 to 20 years.
Relying more heavily on AI and computer models to evaluate and decide public policies for the long-term future naturally invites questions and concern. One of the first has to do with the actual uptake of AI technology by local governments.
A survey of AI released in May by the internal affairs ministry showed only 36 percent of prefectures had introduced or were experimenting with AI, 23.4 percent were discussing the possibility but had no plans to introduce it, and 8.5 percent had no plans to discuss or introduce AI.
While 60 percent of major cities had introduced AI in some form, 70 percent of midsize cities, towns and villages had no plans to either discuss or introduce AI anytime soon.
Even if local governments without AI tried to adopt it quickly, they would face questions about whether it was worth it. Skeptics might point out computer models are only as good as the data entered. The vagaries of forecasting, especially for natural disasters or events beyond the control of the local government, mean even the best-laid plans of bureaucrats, politicians and computer programmers are apt to go astray and waste local tax money.
In addition, there is the political problem local politicians could face in dealing with voters who distrust the idea that a computer program can suggest policies that will effectively, and fairly, govern their locality better than those suggested by humans.
Finally, there is the fear and suspicion among those who worry AI-led policy decision-making is the first step toward an Orwellian state or some science fiction dystopia where computers, not humans, make political decisions.
Abe believes there is great potential for the use of AI in policymaking. But he was quick to address questions about public trust in AI at the news conference, saying the purpose of relying on it more was to further improve the policymaking process as a whole, but that humans were still very much in charge.
“We’re now at the research stage. But I feel, in relation to actual policy that, with the current population decline and the fact we’re in the age of artificial intelligence and the internet of things, the environment has changed. In the midst of such changes, and compared to that, much human thought and human systems are lagging in the past,” Abe said.
Hiroi agreed with Abe.
“AI has seen a bit of a boom recently. There are some extreme discussions that are overly optimistic about AI, or ridiculous suggestions that we should leave everything to AI. But AI is a tool. No matter what, people create the basic model, analyze the results from running the data, suggest proposals and decide policies,” he said.