AAMAS 2025 Computational Economics Competition

Background

This competition, "Optimal Tax Policy with Heterogeneous Agent Models," is one of the challenges hosted by AAMAS 2025. It focuses on designing optimal tax policies in an economy with heterogeneous agents. In real life, taxation and public spending are two key government policies that affect economic growth and wealth distribution. A good government needs to adjust tax and spending policies based on the needs of society to support economic development, social fairness, and long-term stability. However, people’s economic behaviors are different (heterogeneous), making tax policy design a complex challenge.

game roles
render gif

Behavior of Heterogeneous Agents

Traditional economic models assume that people are rational decision-makers who maximize their own benefits. However, in reality, people are different in many ways, including:

  1. Income levels: Some people earn high salaries, while others rely on investments or low-income jobs.
  2. Consumption and savings: Some people save money for the future, while others spend a lot and sometimes go into debt.
  3. Work preferences: Some work long hours to earn more, while others prefer a better work-life balance.
  4. Investment strategies: Some people choose safe investments, while others take high risks to try to earn more.

These different behaviors together shape the overall economy, affecting wealth distribution, economic growth, employment, and social welfare.

Challenges of Government Tax Policy

For a government, designing an optimal tax policy in a society with different types of people is a difficult task. The government must balance several goals, such as:

  1. Income redistribution: How can taxes be fair, so that the rich pay more while keeping the economy strong?
  2. Work incentives: How can taxes encourage people to work without making them avoid paying taxes?
  3. Public welfare: How should tax money be spent to benefit everyone?
  4. Long-term growth: How can tax policies support both short-term stability and long-term development?

To study these problems, we introduce the TaxAI Simulator, a simulated economy where different economic actors make decisions, interact, and adjust to changing conditions. This simulator allows us to test how AI can make smart decisions in a complex economic system.

TaxAI Simulator: Economic Roles

The TaxAI Simulator includes four key roles: government, households, companies, and financial institutions. The government and households make decisions that change over time, while companies and financial institutions follow fixed rules as part of the environment.

Households

  1. The simulator has multiple households, each making decisions independently.
  2. Households have different levels of productivity, which can change over time.
  3. They earn wages from companies (production), buy goods (consumption), and save money in financial institutions (investment).
  4. Households pay taxes to the government.
  5. Their well-being depends on balancing work and consumption:
    • - Working too much reduces quality of life.
    • - Higher consumption improves happiness.

Government

  1. The government controls tax policies and public spending to manage the economy.
  2. It collects taxes from households and borrows money from financial institutions.
  3. The government invests in companies to promote production and economic growth.
  4. Its goal is to increase GDP (economic output) and reduce wealth inequality.

Financial Institutions

  1. Financial institutions (such as banks) accept deposits from households and pay interest.
  2. They lend money to governments and companies.

Companies

  1. Companies hire workers and pay them wages.
  2. They sell goods to households.
  3. They receive investments from the government and loans from financial institutions.

For a detailed description of the relationship, see GitHub Link: github link

Track

The challenge consists of two tracks, focusing on decision making issues in the economy and society from different roles.

Track 1: Optimal Tax Policy-Government Track


Track 1 focuses on optimizing fiscal policy in a simulated macroeconomic environment. Participants will control a government agent responsible for adjusting income tax, asset tax, and government spending to drive economic growth while ensuring social fairness and stability. The government’s primary objective is to maximize GDP while maintaining a balanced wealth distribution. To simulate a complex economy, the environment includes heterogeneous households following three predefined strategies in varying proportions. The government agent must dynamically adjust tax and spending policies in response to economic conditions to achieve optimal outcomes. The performance of each government agent will be evaluated based on economic effectiveness within the simulation, with rankings determined by overall economic performance.

Track webpage: http://www.jidiai.cn/compete_detail?compete=55

Track 2: Optimal working and saving policy-Household Track


Track 2 focuses on optimizing individual decision-making within an economic system. Participants will design household agents that adjust their working hours and wealth investment based on observed economic and social information—such as average wealth, income, and education levels across different social classes (e.g., wealthy, middle class, and low-income groups)—to maximize personal benefits. The environment consists of 100 independent household agents, each making decisions autonomously. The objective is to maximize personal well-being, which depends on both consumption levels and working hours—higher consumption and shorter working hours lead to greater individual benefits. An optimal household agent must strategically balance work and consumption by leveraging observed information, personal assets, and productivity.

During the competition, participant-designed household agents will interact with three predefined household policies in a heterogeneous economic environment. Performance will be evaluated based on overall economic benefits within the simulation, with rankings determined accordingly.

Track webpage: http://www.jidiai.cn/compete_detail?compete=56

How To Participate

The competition will be held online on the Jidi multi-agent open-source platform (Jidi Platform). The Jidi Platform provides participating competitors with fair agent evaluation and real-time ranking functions online. Users only need to register an account on the Jidi Platform and sign up to participate in the challenge.

Each participating team consists of 1-3 members, including a team leader and two team members. The steps to participate are:

  1. All participating members must register a Jidi account using an email/mobile phone number and enter their real names and organizational institutions;
  2. The team leader logs into their Jidi account and fills in the team name, organizational institution, and team members' email/mobile phone number in the "Sign Up Now" section of the arena, and submission of this information completes the registration;
  3. Write and train the agent locally, and submit them for ranking in the subjects of Optimal tax-government and optimal tax-households (this does not affect arena ranking, only for daily testing of agent levels);
  4. Before the deadline for each warm-up match and official match, log into the team leader's Jidi account to submit the agent code, related parameter files, etc., in the "Agent Submission" section, and check in "My Page-Submission List" whether the agent has passed verification (multiple submissions are allowed, but a new submission must wait until the previous verification is complete to avoid incorrect verification results);
  5. After each phase, check the competition results such as scores, rankings, and match replays in the "Results".

Come and Sign up for the competition!

Timeline

Open for registration: 2025/03/14, 10:30 (UTC+8)
Warm-up Round 1 Submission Deadline: 2025/03/24, 10:30 (UTC+8)
Warm-up Round 2 Submission Deadline: 2025/04/07, 10:30 (UTC+8)
Submission Deadline for Round 1 (Tentative): 2025/04/22, 10:30 (UTC+8)
Submission Deadline for Round 2 (Tentative): 2025/05/06, 10:30 (UTC+8)

Award

According to the competition rankings, we will award certificates to the top 6 teams and invite them to present their work on-site at the AAMAS 2025 conference.

Organizers

Haifeng Zhang

Institute of Automation, Chinese Academy of Sciences

Nanjing Artificial Intelligence Research of IA

Email: haifeng.zhang@ia.ac.cn

Qirui Mi

Institute of Automation, Chinese Academy of Sciences

Email: miqirui2021@ia.ac.cn

Wentian Fan

Nanjing Artificial Intelligence Research of IA

Email:fanwentian23@mails.ucas.ac.cn

Shuai Wu

Nanjing Artificial Intelligence Research of IA

Email:wushuai@airia.cn

Zhiyu Zhao

Institute of Automation, Chinese Academy of Sciences

Email: zhaozhiyu2022@ia.ac.cn