Beware of “record” numbers claims in a headline
The headline du jour: The Number of Americans Identifying as Liberal Jumps to Record High.
Why am I calling for a skeptical view? Because, ladies and gentlemen, there are more adults in the U.S. this year than last year. So even if the percentages remained constant, there would be “more” of whatever is being measured.
In today’s story, Gallup claims a 1% increase in the number of people who self-identify as liberals:
Americans continue to be more likely to identify as conservatives (38%) than as liberals (23%). But the conservative advantage is down to 15 percentage points as liberal identification edged up to its highest level since Gallup began regularly measuring ideology in the current format in 1992.
The key to any year-to-year comparison is to use a deflated measurement, which is often a simple percentage. With a 1% increase in the number of people who self-identify as liberal, Gallup’s statement seems true. But is it?
For results based on the total sample of national adults, the margin of sampling error is ±1 percentage point at the 95% confidence level.
The sampling error means that today’s liberal state could be 22% — the same as last year’s — or 24%.
Gallup’s data show that America is pretty evenly (over time) divided in terms of self-identification of party. The range shows that the biggest change over time has been in people who self-identify as moderate — a drop in 9 points:
- Conservative: 36% (1992) to 40% (2011) — 4 point change
- Moderate: 43% (1992) to 34% (2013) — 9 point change
- Liberal: 17% (1992) to 23% (2013) — 6 point change
The more interesting question coming out of Gallup’s poll is this: “why do fewer American identify as moderate today than 20 years ago?” The answer might be found in the positions of the parties or it might be in the types of political rhetoric that media choose to broadcast. Or that the cultural meaning of each term has evolved (my bet).