Finding 3 — Who didn't register, and why

What non-registrants say

The Current Population Survey (CPS) VOYNOTREG question asks respondents who are not registered to select the reason. We harmonize the codes into four policy-relevant categories (this registration question has no separate "logistical" bucket, unlike the registered-non-voter question in Finding 4):

access — missed deadlines, didn't know how, recent move (addressable by Automatic Voter Registration (AVR), Same-Day Registration (SDR), Online Voter Registration (OVR), and pre-registration for 16- and 17-year-olds)^[6]
engagement — not interested, vote wouldn't matter (addressable by civic infrastructure, outreach)
personal — illness, language barrier, ineligible (addressable by targeted accommodations)
other — miscellaneous / don't know

Pooled across 2000-2024, 18-29 non-registrants. The access category dominates across the pooled series, with engagement as the secondary cluster.

By race × gender (pooled 2000-2024)

For each race × gender group of non-registrants, the share citing access barriers (deadlines, didn't know how, recent move) vs engagement reasons (not interested / "won't matter"). Read each group across: access is the larger of the two in every group. Pooled 2000-2024; groups with unweighted n under 400 are indicative. The full four-category mix (all shares sum to 100%) is in the disclosure above.

What the group view shows

Access barriers × state policy availability

A non-registrant citing "missed the deadline" in a state with same-day registration is a different problem from the same response in a state without it. The Finding 6 — Policy lever page maps non-registrant shares against state National Conference of State Legislatures (NCSL) policy availability directly.^[6] The headline: access-category responses are consistently higher in states that lack same-day registration or online voter registration at the time of the cycle.

So what

For Secretaries of State. The top barrier cited by non-registrants in your state is, with high probability, operational — deadlines, knowledge gaps, and residence churn. This is your domain. Finding 6 pairs these responses with state NCSL policy flags; Recommendations lists the policy levers with evidence-backed effect ranges.

For foundation officers. Access-category non-registrants are the cheapest marginal voters in the intervention catalog — policy reform amortizes to $0-$5 per marginal voter; OVR portals to $5-$15. Compare to programmatic get-out-the-vote (GOTV) at $20-$500.^[13][14] The reason diagnostic validates that weighting.

For the public. Non-voting among young adults is not, in its plurality, a story about not caring. It is, in the first layer, a story about never reaching the rolls.

Methodology. VOYNOTREG harmonization maps the raw IPUMS codes^[1] to policy-relevant categories per scripts/process_cps.py. Pooled weighted shares use VOSUPPWT. The universe is citizens 18-29 who are not registered AND who provided a VOYNOTREG answer (i.e., excludes "NIU: registered" and "Refused / Don't know"). Unweighted group n reflects the denominator for the suppression rule. Full provenance: methodology page.