Debate: should governments invest more in fusion?

Evidence cards

Each card gives verified facts with their source and date (all checked on 1 November 2026). Read them carefully: a card is evidence, but you decide what it shows.

Card A: Records in the laboratory

In its final experiments in autumn 2023, the JET tokamak in the UK released $69.26\ \text{MJ}$ of fusion heat in one pulse of about six seconds (EUROfusion, 8 February 2024). On 5 December 2022 the National Ignition Facility (NIF) in California released $3.15\ \text{MJ}$ from a fuel target hit by $2.05\ \text{MJ}$ of laser light, and in 2025 it reached $8.6\ \text{MJ}$.

Card B: The fuel

About one hydrogen atom in 6420 in sea water is deuterium, roughly $0.035\ \text{g}$ in every litre of water. The ITER Organization states that lithium reserves on land would allow fusion power plants to operate for more than 1000 years.

Card C: Waste and safety

According to the ITER Organization, fusion produces no high-activity, long-lived waste; the materials made radioactive by neutrons could be recycled within about 100 years. A fusion plasma holds only a few grams of fuel, and it cools within seconds if anything goes wrong.

Card D: A changing electricity system

In 2025, renewable sources gave $33.8\ \%$ of the world's electricity, more than coal ($33.0\ \%$). Wind and solar together gave $17.3\ \%$; their output depends on the weather (Ember, Global Electricity Review 2026). A grid with much variable power also needs storage, flexible demand or firm low-carbon power.

Card E: Private money

A survey of 56 fusion companies by the Fusion Industry Association (July 2026) found that they raised USD 4.48 billion in the year to July 2026 and USD 14.24 billion in total, of which about USD 0.98 billion came from public sources. The association said that the companies still need government support (World Nuclear News, 14 July 2026).

Card F: ITER's cost and schedule

When ITER was agreed in 2006, its construction was estimated at $5.9$ billion euros over ten years. Later estimates of construction and operation together are $18\text{–}22$ billion euros; other, higher estimates of up to $65$ billion US dollars are disputed by ITER. The 2024 baseline plans deuterium–deuterium plasmas from 2035 and deuterium–tritium operation from 2039, and added an estimated $5$ billion euros, partly for repairing faulty parts (Wikipedia summary of the 2024 baseline).

Card G: Gain is not electricity

NIF's lasers needed about $300\ \text{MJ}$ of electricity to produce the $2.05\ \text{MJ}$ of laser light. ITER aims for $500\ \text{MW}$ of fusion heat from $50\ \text{MW}$ of heating, but it will use over $300\ \text{MW}$ of electricity and will not produce any. As of November 2026, no fusion device has delivered electricity to a grid.

Card H: The climate clock

The IPCC concluded in 2018 that, to limit warming to $1.5\ ^\circ\text{C}$, net carbon dioxide emissions must fall by about $45\ \%$ from 2010 levels by 2030 and reach net zero around 2050. The UK's prototype fusion plant STEP targets first operations in 2040 (UK government, 12 June 2025).

Card I: From public laboratory to company

Commonwealth Fusion Systems was founded in 2018 as a spin-off from the Plasma Science and Fusion Center of the Massachusetts Institute of Technology (MIT). There, the Alcator C-Mod tokamak (1991–2016), proposed to the US Department of Energy in 1985, set a record plasma pressure of $2.05$ atmospheres in 2016; in 2014 a US federal spending law gave it 22 million dollars to keep running. In May 2023 the US Department of Energy gave the company, with seven others, milestone-based funding.

Key idea

A good debate argument has three parts: a claim (what you say), evidence (a fact with its source and date) and reasoning (why the evidence supports the claim). The same evidence can support both sides: private investment can show that fusion is attractive, or that governments can leave it to companies. So reasoning, not the fact alone, wins the argument.

Check the kind of evidence. An established result is stronger than an institutional plan, and a plan is usually stronger than a corporate claim, although a plan's own record matters too: ITER's schedule has slipped several times. Never use scientific gain $Q$ as if it were electricity. A strong speaker also rebuts: answers the best argument of the other side instead of ignoring it, and concedes a fair point when it is true.

How a structured debate works

The motion is "This house believes that governments should invest more public money in fusion research." The proposition (the team for) argues for it, the opposition (the team against) argues against it; the fact-checkers check numbers. Your role in the debate is not necessarily your own opinion: arguing the other side helps you understand it.

Role	Who	Task
chair	1	opens, gives the floor, keeps time
for	3	opening, rebuttal, closing
against	3	opening, rebuttal, closing
checkers	2	check numbers against the cards
judges	the rest	score, then vote

The order and the time limits are:

1. The chair opens and the audience votes once (for, against, undecided), 1 minute.
2. Opening speeches: proposition, then opposition, 2 minutes each.
3. Rebuttals: opposition, then proposition, 2 minutes each.
4. Open floor: questions from the judges to both teams, 5 minutes.
5. Fact-check report: correct facts and any mistakes, 2 minutes.
6. Closing speeches: opposition, then proposition, 90 seconds each.
7. Second vote and the judges' scores, 2 minutes.

The judges score each team from 0 to 3 on each criterion: evidence (at least two cards, with source and date), accuracy (results, plans and claims kept apart; gain is not electricity), reasoning (evidence linked to the motion; the other side answered) and language (clear, polite debate language). The scores give each team feedback and are used for assessment; the winner is decided by the votes: the team that changes more votes, not the team that started with more support, wins.

Common misconception

"The side with more facts wins." Facts without reasoning do not persuade. A single strong card, well explained and defended against the rebuttal, is worth more than five cards read aloud.

"Disagreeing means attacking the other speaker." In a debate you attack arguments, never people. Phrases such as "I see your point, but …" keep the discussion respectful.

Summary

• A debate argument has a claim, evidence (with source and date) and reasoning.
• Established results are stronger evidence than plans, and plans are usually stronger than corporate claims (check their record); scientific gain is not electricity.
• The same evidence can support both sides; the reasoning decides.
• Rebut the strongest argument of the other side and concede what is true.
• Whether to spend more public money on fusion is a question of values and priorities as well as of physics: well-informed people can disagree.