Definition
The majority trap is Rosling's name for the mistake of treating "most people in category X" as a precise claim when it is actually compatible with very different lives at the top and bottom of that category.
Saying "most people in a Level 2 country have access to electricity" can be true while still hiding the fact that the bottom quartile experiences daily outages and the top quartile runs air conditioners. "Most" tells you about the center of a distribution; it tells you nothing about its width.
Why it matters
How it works
Once a stereotype has been refined into a more careful claim about a group, the next trap is treating that claim as if it covered everyone uniformly. "Most people in this country" smooths over a distribution that may include a small wealthy elite, a sizable middle, and a substantial poor minority. The single word "most" obscures all of that texture.
The trap matters most in policy and aid. If you design a program for "the majority of households," you can produce real improvements while the bottom 20% experience nothing — or worse, lose access to services that previously included them. Rosling's income-level framing is partly a response to this trap: by replacing two buckets with four, and by always asking who is in the bottom of each level, the analyst keeps the tails visible.
The corrective question is simple and works everywhere: when someone says "most" or "the majority," ask what fraction, and what the bottom-tail and top-tail look like inside the group. If the answer is "55% and the tails differ enormously," the majority claim was hiding the story.