Why Is Consensus Harder Than Voting?

The G20 began in 1999 as a meeting of finance ministers and central bank governors, assembled after the Asian financial crisis made it obvious that the existing club of rich economies could no longer discuss the world economy without the rest of it in the room. It only became a leaders' summit in 2008, when another crisis made the point again.

Note what it never acquired along the way. No charter. No secretariat of any size. No treaty obligations. And no vote.

What happens when nobody can be outvoted

In a body with a vote, the arithmetic does the final work. You count, you win or you lose, and the losing side lives with it. That produces a particular style of argument: build a coalition big enough, then stop persuading.

Remove the vote and the arithmetic disappears. A G20 communique is agreed by everyone or it does not exist. The smallest delegation in the room has the same power as the largest, because the largest cannot outvote anyone. Persuasion never stops being necessary.

This is why real communiques read the way they do. The language is careful because every clause survived somebody's objection. Students often find this frustrating for a day and then find it fascinating, because the skill it demands is not one school rewards.

The skill it actually trains

Finding language that two people who disagree can both sign is a specific and unusual ability. It is not compromise in the weak sense - not splitting the difference until nobody minds and nobody cares. It requires understanding the other position well enough to know which part of it is load-bearing and which part is decoration.

A student who can do that can also negotiate a salary, run a team, chair a meeting and survive a family argument. Very little else in a school timetable trains it.

What the room feels like

Quieter than people expect, and more economic. Sessions turn on growth, trade, debt, energy and climate finance rather than on rhetoric. The delegate who dominates the floor is often not the one who ends up in the final text, because the final text belongs to whoever was trusted enough to draft it.

This suits a student who enjoyed the corridor conversations at a Model UN more than the formal speeches. It is also, in our experience, a gentler first experience for a capable student who is still building confidence in front of a room - the pressure is on the drafting rather than the performance.

Our Model G20 runs in person at FLAME University, Pune. If the written form appeals more than the room, Model NITI Aayog ends in a published paper instead - the two are compared in what comes next after a first MUN.

The G20's critics call it a talking shop with no enforcement. They are right about the mechanism and wrong about what follows from it. An agreement nobody can be forced into is one everybody has to actually be convinced of, which is a harder thing to build and a more durable thing to have built.

Dates and registration are on Upcoming Programs.

Key takeaways

• Voting tallies preferences; consensus requires everyone to accept the outcome, making them fundamentally different decision mechanisms
• The Byzantine Generals Problem shows that distributed agreement is mathematically harder than simple counting when participants may be unreliable
• Consensus scales poorly because it requires pairwise communication, growing effort roughly with the square of group size
• Real systems like blockchains use economic incentives and fault-tolerant protocols to manage the difficulty of achieving consensus
• The choice between voting and consensus depends on whether speed or buy-in is more important for the decision at hand

Test yourself