VOA notes that

In the past, the downside of the time-consuming caucus process has been lower voter turnout. In the 2016 Iowa caucuses only 15.7% of the voting population participated. In contrast, more than 50% of New Hampshire voters cast ballots in the 2016 primary.

Is this (large) difference in participation consistently observed for other caucuses vs primaries in the US?

  • 3
    While the ultimate answer seems to be a foregone conclusion, I like the idea of testing our assumptions and validating or disproving them. +1 Feb 7, 2020 at 15:45
  • 1
    caucuses don't typically report raw participant counts, so it's harder to determine participation rates than comparing the vote total to the population of the state.
    – dandavis
    Feb 7, 2020 at 22:29
  • 1
    Are you only asking about overall participation rates, or if there are differences in age, class, race etc (ie, do caucus goers skew older and wealthier than primary voters)?
    – divibisan
    Mar 24, 2020 at 14:34

2 Answers 2


Four states switched their primaries from caucuses to primaries between 2016 and 2020: Maine, Minnesota, Colorado, and Utah, providing an additional test. Turnout increased dramatically in all four states (Vox.com):

| State |  2016   |  2020   | Increase |
| CO    | 122,000 | 755,000 |      6.2 |
| ME    |  47,000 | 194,000 |      4.1 |
| MN    | 205,000 | 745,000 |      3.6 |
| UT    |  77,000 | 175,000 |      2.3 |

Of course, the races are very different (two vs. many candidates, an incumbent president, no Republican primary), and Super Tuesday turnout grew over the same time period, but only about 1.3x overall (Vox.com). The US population grew by about 1.02x in the same time period.

And, as Don pointed out, participation in the 2016 Washington non-binding primary was 30x higher than the caucus, even though it was two months later, and non-binding. Although, as I commented, the primary ballots were mailed to every voter, so the barrier to participation is (rightfully) much lower than an in-person primary.


Yes, there is. There's actually an interesting test case with Washington state which, up until the 2020 elections, held both primaries and caucuses. In 2016, there were 26,345 participants in the caucuses who voted overwhelmingly for Sanders over Clinton. The primary had 802,754 who voted narrowly (non-bindingly) for Clinton over Sanders. The difference in numbers indicate that not only do caucuses have lower participation, but draw from a different sample of the population.

Source: https://en.wikipedia.org/wiki/2016_Washington_Democratic_caucuses

  • With the caveat that the Washington 2016 non-binding primary was conducted entirely by mail, like all elections in Washington. So it is not quite the same comparison as a caucus vs. in-person primary. Mar 24, 2020 at 17:04

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .