A large majority of Americans say they will participate in the census: 87% say they definitely or probably will fill out and return their forms, or have already done so. But in a recent Pew Research Center survey, 12% still said they may not participate — 5% might or might not, 4% probably will not and 3% definitely will not fill out and return their census forms.1

Age is the demographic characteristic most strongly associated with participation. The median age of those who may not participate is 31, compared with a median age of 47 among likely participants. Put differently, among the 12% who say they might or will not participate in the census, 44% are younger than age 30. By comparison, among the 87% who say they will participate, only 17% are ages 18-29.

Home ownership — which is correlated with age — is also related to census participation. Only 39% of those who are uncertain or say they will not participate own their home, while 40% rent and 20% have some other arrangement (including living in a dorm or with their parents). By comparison, a majority (64%) of those who will definitely or probably participate own their homes.

Those who might not participate also have lower levels of education and income than those who are likely to participate. About seven-in-ten (71%) who may not participate have a high school education or less, while 28% have attended college. By comparison, 47% of those who intend to participate have a high school education or less and 52% have attended college. A similar pattern is evident when it comes to family income. But there are no significant differences between these two groups by gender, race or ethnicity.

In a multivariate analysis to see which factors matter most in predicting those who may or will not participate in the census, age is the most important predictor. Even when controlling for other demographic characteristics, analysis finds that younger Americans are more likely to say they might not participate. For those younger than age 30, the probability of likely non-participation is .18, compared with a probability of only .04 for those age 50 and older.

Education also is a factor in predicted census participation, even controlling for age and other demographic characteristics. For those with a high school education or less, the probability of not participating is .11 whereas those with a college education have a probability of .05. Gender, party and race and ethnicity are not significantly related to whether someone might not participate. And after controlling for other demographic characteristics, income and home ownership, are not significantly related to intention to participate.

Attitudinal Differences about the Census

There are also significant attitudinal differences between those who may not participate and those who say they will participate. Only 28% of those who might not participate think the census is very important for the United States, 36% say it is somewhat important and 21% say it is not too or not at all important. Nearly three-fourths (72%) of those who intend to participate think the census is very important. Similarly, far fewer of those unlikely to participate think participating in the census is a civic responsibility (42% vs. 84% of those who intend to participate).

Awareness is also a significant factor. About a third (32%) of those who may or will not participate have never heard of the census, compared with only 8% of those who intend to participate.

Only 16% of those who may or will not participate think their participation will personally benefit them while 10% think it will harm them. Among those who intend to participate, 35% think completing their census forms will personally benefit them and only 4% think their participation will harm them. A similar pattern is evident on views about whether participating in the census will benefit their community. Less than a third (30%) of those who might not participate think it will benefit their community, compared with 67% of those who intend to participate.

For more analysis of public views on the 2010 census, see “With Growing Awareness of Census, Most Ready to Fill Out Forms” and “Most View Census Positively, But Some Have Doubts“. Also see “All Things Census,” a gathering place for frequent postings about census methodology, findings and resources at census.pewresearch.org/social-trends.


1. This group of likely non-participants includes those who say in a follow-up question that someone else in their household either may fill out the form (this is 25% of this group, or 3% of the total sample). The results reported here would not change substantially if this small group were excluded from likely non-participants.