Seven Years of 'Decarbonization' Discourse — Discontinuous Diffusion of Climate Language in Japanese Local Assemblies (2018–2024)

This article draws on the minutes dataset constructed by the Machikarte Lab — containing more than one million assembly speeches as of 2024 — to aggregate speeches that include climate-related terms such as "decarbonization," "carbon neutrality," "renewable energy," "global warming," and "greenhouse gas" by year and prefecture for 2018–2024. The aim is to read the temporal structure and regional distribution of climate-change debate in local assemblies. Evaluating individual assemblies or individual assembly members is outside the article's purpose. Observation is confined to national trends and prefectural distribution; no top-or-bottom ranking by municipality name is published here.

Over the seven-year window from 2018 to 2024, mentions of climate-related terms in Japanese local assemblies exhibit a two-layer structure.

The first layer consists of established terms: "renewable energy," "global warming," and "greenhouse gas." These terms already had a stable annual mention count in 2018, and show a gradual upward trend across the seven years. There is no sharp discontinuity — a pattern consistent with the gradual institutional embedding of policy vocabulary.

The second layer consists of newer terms: "decarbonization" and "carbon neutrality." Before 2020 these terms appeared only rarely in local assembly proceedings, but in 2021 they surged to several times their previous-year count and by 2024 had entered the most-cited group among all climate-related terms. This discontinuity corresponds to the October 2020 policy declaration within a period of less than one year.

Comparing 2024 discussion density (climate-related term mentions per municipality) across prefectures, there is a several-fold gap between the top and bottom groups, indicating substantial regional variation in discussion intensity.

Timeline of the Act on Promotion of Global Warming Countermeasures

The Act on Promotion of Global Warming Countermeasures (Act No. 117 of Heisei 10, 1998) was enacted in 1998 and has undergone substantial revision since Japan ratified the Paris Agreement in 2016. The 2021 revision inscribed the 2050 carbon-neutrality target as a foundational principle and made it mandatory for local governments to strengthen their local action plans.

Two institutional milestones in 2018–2024 had the greatest effect on local assembly debate. The October 2020 declaration of the 2050 carbon-neutrality goal — delivered as a Prime Minister's address — put national-level targets squarely before local assemblies, which then faced pressure to articulate their own response strategies. Municipalities were required to draft or revise their Local Government Action Plans (regional-measures edition), and the resulting surge of "decarbonization" and "carbon neutrality" mentions in local assemblies during 2021 is structurally consistent with this sequence.

The Decarbonization Leading Area scheme, launched in April 2022, invites municipalities to submit proposals that are then selected by the Ministry of the Environment. Preparatory deliberations and assembly reporting ahead of selection plausibly kept mention rates elevated, and the continued high levels from 2022 onward also reflect the broader series of GX (Green Transformation) promotion policies.

The political function of "decarbonization" as a term

Among climate-related terms, the sharp rise of "decarbonization" and "carbon neutrality" in 2021 reflects the political function these words carry. "Global warming" and "greenhouse gas" entered the lexicon as direct translations of scientific concepts, becoming shared vocabulary in both academic and administrative domains. "Decarbonization," by contrast, circulated top-down as an administrative policy term.

This difference may signal distinct propagation routes. Contrasting a term that took more than a decade to become established after the Feed-in Tariff Act (Act for the Promotion of Utilization of Non-Fossil Energy Sources and Effective Use of Fossil Fuel Resources by Energy Suppliers) took effect in 2012 with a term that surged within a year of a government declaration highlights the difference in propagation mechanisms for policy vocabulary. This hypothesis will be examined in a subsequent version by analyzing prefectural variation in propagation speed.

Interpreting "high discussion volume"

Several factors may account for the prefectural variation in discussion density. Candidate explanations include high solar and wind energy potential in certain regions driving debate, concentration of municipalities targeting Decarbonization Leading Area status, framing of forestry and blue-carbon topics in agricultural and fishing-sector municipalities, and debates over factory recruitment tied to decarbonization requirements. Cross-referencing industrial energy consumption data with renewable energy installation volumes is a task for the next phase.

Speed of penetration of declaration-type policy into local assemblies

The clearest structural finding from this aggregation is the temporal pattern: "decarbonization" and "carbon neutrality" flooded simultaneously into local assemblies in 2021, within a year of the October 2020 Prime Minister's declaration. An observation of less than one year as the penetration speed of a declaration-type policy into local assemblies constitutes a reference point for comparison with other policy terms.

That said, interpreting this one-year figure requires caution. Whether the 2021 increase in mentions reflects reporting-type statements — "our municipality is committed to decarbonization" — or problem-raising statements — "we are concerned about the fiscal burden" — cannot be distinguished without contextual classification. Separating penetration speed by position for and against may reveal a different temporal structure.

The dual structure of established and new terms

The contrast between "renewable energy" and "decarbonization" illustrates the difference in the rate at which policy vocabulary accumulates in local assemblies. Juxtaposing a term that took more than ten years to become established after the Feed-in Tariff Act with a term that surged within a year of a declaration is an interesting vantage point from which to conduct comparative research on policy-term propagation mechanisms.

As a connection to other articles in this lab, combining with the distribution of "under consideration" expressions in responses (see case-sakiokuri-rate) allows measurement of how often "We will consider this" reservations appear in climate-change debates. The intersection with fiscal consolidation debate (see trends-fiscal-health-2018-2024) also connects structurally as the tension between decarbonization investment and fiscal constraints.

Outlook for 2025 and beyond

As of 2024, the total mention count for climate-related terms has reached its highest level in seven years. From 2025 onward, the formulation of the third Strategic Energy Plan and the concrete implementation of GX promotion policies are scheduled, and assembly debate may develop further. Continued data collection and year-on-year comparison will serve as an observational check on this forecast.

• No semantic classification applied: Contextual classification of speeches containing climate-related terms (advocacy for decarbonization / concerns about fiscal burden / explanation of national schemes / reporting on municipal plans, etc.) has not been implemented; the analysis remains at the level of surface-form aggregation
• Term co-occurrence in single speeches: When "decarbonization" and "carbon neutrality" appear in the same speech, both are counted in each term's total
• Speaker type not separated: Mentions by assembly members raising issues and by administrative respondents in answers are not distinguished
• Coverage variation: Year-on-year changes in the number of municipalities included in the dataset affect absolute counts
• No cross-tabulation by sector or energy source: Cross-aggregation with individual technology terms (solar / wind / hydrogen, etc.) is a task for the next phase
• No individual municipality named: This article is confined to structural analysis; no top-or-bottom ranking by municipality or assembly member name is presented

The query specification used for aggregation (spec_version v1-climate-carbon-2026-07) is stated explicitly at the end of this article, and the BigQuery aggregation queries are also placed in the machikarte GitHub repository, so that third parties can independently reproduce the results.

Aggregation queries (spec_version v1-climate-carbon-2026-07)

Annual trend (2018–2024):

SELECT
  year,
  COUNT(*) AS climate_mentions,
  COUNT(DISTINCT municipality_code) AS municipalities
FROM `correlate-workspace.isvd_machikarte.speeches`
WHERE year BETWEEN 2018 AND 2024
  AND REGEXP_CONTAINS(body, r'(脱炭素|カーボンニュートラル|再生可能エネルギー|地球温暖化|温室効果ガス|ゼロカーボン)')
GROUP BY year
ORDER BY year

Term-by-term breakdown (year 2021):

SELECT
  CASE
    WHEN REGEXP_CONTAINS(body, r'(脱炭素|カーボンニュートラル|ゼロカーボン)') THEN 'New terms (decarbonization group)'
    WHEN REGEXP_CONTAINS(body, r'再生可能エネルギー') THEN 'Established terms (renewable energy)'
    ELSE 'Established terms (warming group)'
  END AS term_group,
  COUNT(*) AS mentions
FROM `correlate-workspace.isvd_machikarte.speeches`
WHERE year = 2021
  AND REGEXP_CONTAINS(body, r'(脱炭素|カーボンニュートラル|再生可能エネルギー|地球温暖化|温室効果ガス|ゼロカーボン)')
GROUP BY term_group

machikarte — National Local Assembly Speech Search Platform (beta) — Institute for Social Vision Design (ISVD). ISVD

machikarte (GitHub) — schema, aggregation queries, license (MIT + CC BY 4.0) — Institute for Social Vision Design (ISVD). GitHub

Act on Promotion of Global Warming Countermeasures (Act No. 117 of Heisei 10, 1998) — Ministry of the Environment. e-Gov Law Search

Global Warming Countermeasures and Decarbonization Policy (GX, Decarbonization Leading Areas, etc.) — Ministry of Economy, Trade and Industry. Ministry of Economy, Trade and Industry

Verification of a Method for Role Classification of Parliamentary Minutes Speeches Using a BERT-Based Classifier — Yutaro Miyaki and Yuzu Uchida. Journal of Japan Society for Fuzzy Theory and Intelligent Informatics, Vol. 37, No. 1, pp. 530–534