Harvard Kennedy School Shorenstein Center

Reported Voting Issues: St. Louis city, Missouri

About this Report

This report tracks self-reports of voting-related issues including misleading election messages, intimidation, and voter registration problems to help identify areas that may need a closer look. It is part of the broader 2026 Voting Information Trends Research Initiative, which examines how Americans access voting information and understand key election-related topics ahead of the 2026 U.S. elections.

We include counties where at least 50 people responded to our survey. Of 288 such counties, St. Louis city ranked 82nd highest in reports during this survey period. Findings reflect 94 respondents from St. Louis city, Missouri, surveyed between August 23 and October 31, 2026.

Important Note: This is not a nationally or locally representative sample. Respondents are recruited online, and counties rated higher risk by the Voting Rights Lab are oversampled. Figures are unweighted and reflect only those who responded, not the broader county population. Findings should be interpreted as signals that warrant further attention or verification, rather than as representative estimates or rankings of election administration, election integrity, or the prevalence of voting problems among county residents.

Monitored election problems

ProblemCounts when a respondent reports they…Time window
Registration challenge noticeGot a notice that their voter registration was challenged or canceled, or that they were removed from the rolls.Past month
Conflicting registration informationGot conflicting information about their voter registration.Past week or month
Discouraged from registeringSomeone tried to discourage them from registering or staying registered.Past day, week or month
Misleading calls or textsGot a text, email, message, call or robocall about the 2026 election that felt misleading, coercive or unusual.Past day, week or month
Intimidation or harassmentExperienced or witnessed harassment or intimidation connected with voting or elections.2026 election cycle
Trouble with a voting taskHad trouble registering, checking their registration, or requesting, returning or tracking a mail ballot.2026 election

Table of Contents

Select a name to go to that section.

SectionWhat it shows
Summary of FindingsShare of St. Louis city respondents reporting each problem, compared with the Missouri and national rates
Week by weekWeekly shares for the three problems asked about in the past day or week
Who reported a problemShares by gender, age, household income, race/ethnicity, education and party
Follow-up detailFollow-up answers for each problem reported by more than 10 people
Missouri overviewThe same measures for all respondents in Missouri, for comparison
National overviewThe same measures for all respondents nationally, for comparison
About the surveyWhat the survey monitors, how counties are ranked, and limits of the data

Summary of Findings

33.0% of St. Louis city respondents (31 of 94) reported at least one of six election problems, compared with 28.5% nationally. No single problem is higher than the national rate by more than the margin of error. Statewide, 26.0% of Missouri respondents (401 of 1,543) reported at least one problem; the chart shows St. Louis city, Missouri and national rates side by side. St. Louis city respondents make up 6.1% of all Missouri respondents (94 of 1,543) and 0.2% of all respondents nationally (94 of 61,908).

Share reporting each problem, St. Louis city, Missouri

Figure 1. Share of St. Louis city, Missouri respondents reporting each problem, with the Missouri and national shares, Aug 23–Oct 31, 2026. St. Louis city n = 94; Missouri n = 1,543 (including St. Louis city, 6% of the state's respondents); national n = 61,908. Margin of error for the county share reporting any problem: ±9.5 percentage points, so the county figure is within the margin of error of the national rate. Source: CHIP50 survey data.

Week by week

Weekly share reporting each problem, St. Louis city, Missouri

Figure 2. Weekly share of St. Louis city, Missouri respondents reporting each problem (teal line with circles), with Missouri statewide (red line with squares, including St. Louis city) and the national share (black dashed line). The shaded band is the 95% confidence interval for the national share, which is the range the weekly figure could vary based on the sample size alone. Hollow points are weeks with fewer than 10 St. Louis city respondents. Weeks start Monday; only problems asked about in the past day or week are shown (conflicting registration information: past week), each person counted once per week. Source: CHIP50 survey data.

Weekly groups in St. Louis city are small (1 to 41 people), and who answered these questions changes from week to week, so treat week-to-week movement as a signal to watch, not a finding. From Aug 23–Oct 17 to Oct 18–Oct 31, the share of St. Louis city respondents reporting any of these three problems went from 12.7% to 9.4% (71 and 32 people): no clear change, since the difference is within the margin of error.

Who reported a problem

Among groups of 30 or more people, reports were most common among people with household income $25,000 to $50,000 (39.4%) and least common among independents and others (25.6%), against 33.0% overall. Reports fell with education and age.

GroupReportedShare
All respondents31 of 9433.0%
Gender
Female15 of 5328.3%
Male16 of 4139.0%
Age
18 to 242 of 4–
25 to 4417 of 4637.0%
45 to 649 of 2931.0%
65 and over3 of 1520.0%
Household income
Under 25K6 of 2920.7%
25k to under 50k13 of 3339.4%
50K to under 75K5 of 1631.2%
75K to under 100K3 of 6–
100K and over4 of 1040.0%
GroupReportedShare
Race/ethnicity
Black or African American14 of 4928.6%
White or Caucasian16 of 4238.1%
Asian or Asian American3 of 7–
Hispanic or Latino1 of 2–
Other0 of 2–
Education
Some High School or Less1 of 4–
High School Graduate10 of 2737.0%
Some College7 of 2133.3%
College Degree10 of 3132.3%
Graduate Degree3 of 1127.3%
Party
Democrat13 of 4032.5%
Republican8 of 1553.3%
Independent/Other10 of 3925.6%

Share reporting at least one of the six problems, by group, St. Louis city, Missouri, Aug 23–Oct 31, 2026. Race/ethnicity is select-all-that-apply, so one person can appear in more than one group. Shares are shown for groups of 10 or more people. Source: CHIP50 survey data.

Follow-up detail

Follow-up detail is shown for each problem reported by more than 10 people. Not shown: registration challenge notices (2 people), conflicting registration information (no follow-up question), being discouraged from registering (7 people) and intimidation or harassment (7 people).

Trouble with a voting task, by task

Trouble registering9% (8)
Trouble checking registration7% (7)
Trouble requesting a mail ballot3% (3)
Trouble tracking a ballot2% (2)
Trouble returning a mail ballot1% (1)

Share of the 94 people asked.

Missouri Overview

August 23 to October 31, 2026. 1,543 respondents across Missouri, including people whose county is not on file.

26.0% of Missouri respondents (401 of 1543) reported at least one of six election problems, compared with 28.5% nationally. No single problem is higher than the national rate by more than the margin of error.

Share reporting each problem, Missouri

Figure 3. Share of Missouri respondents reporting each problem, with the national share, Aug 23–Oct 31, 2026. Missouri n = 1543; national n = 61,908. Margin of error for the state share reporting any problem: ±2.2 percentage points, so the state figure is lower than the national rate by more than the margin of error. Source: CHIP50 survey data.

Missouri: Week by week

Weekly share reporting each problem, Missouri

Figure 4. Weekly share of Missouri respondents reporting each problem (red line), with the national share (black dashed line). The shaded band is the 95% confidence interval for the national share, which is the range the weekly figure could vary based on the sample size alone. Hollow points are weeks with fewer than 10 Missouri respondents. Weeks start Monday; only problems asked about in the past day or week are shown (conflicting registration information: past week), each person counted once per week. Source: CHIP50 survey data.

Weekly groups in Missouri are small (31 to 530 people), and who answered these questions changes from week to week, so treat week-to-week movement as a signal to watch, not a finding.

Missouri: Who reported a problem

Among groups of 30 or more people, reports were most common among people aged 18 to 24 (39.4%) and least common among people without a high school diploma (18.9%), against 26.0% overall. Reports rose with education, and fell with age.

GroupReportedShare
All respondents401 of 154326.0%
Gender
Female204 of 88323.1%
Male197 of 66029.8%
Age
18 to 2439 of 9939.4%
25 to 44164 of 54530.1%
45 to 64132 of 58322.6%
65 and over66 of 31620.9%
Household income
Under 25K92 of 45420.3%
25k to under 50k117 of 48024.4%
50K to under 75K92 of 29731.0%
75K to under 100K50 of 14833.8%
100K and over50 of 16430.5%
GroupReportedShare
Race/ethnicity
White or Caucasian313 of 126724.7%
Black or African American88 of 25135.1%
Native American or Alaska Native11 of 4922.4%
Hispanic or Latino14 of 4928.6%
Asian or Asian American7 of 1936.8%
Other2 of 1020.0%
Pacific Islander or Native Hawaiian1 of 2–
Education
Some High School or Less20 of 10618.9%
High School Graduate110 of 53120.7%
Some College102 of 39825.6%
College Degree128 of 39032.8%
Graduate Degree41 of 11834.7%
Party
Democrat149 of 46831.8%
Republican125 of 51624.2%
Independent/Other127 of 55922.7%

Share reporting at least one of the six problems, by group, Missouri, Aug 23–Oct 31, 2026. Race/ethnicity is select-all-that-apply, so one person can appear in more than one group. Shares are shown for groups of 10 or more people. Source: CHIP50 survey data.

Missouri: Follow-up detail

Follow-up detail is shown for each problem reported by more than 10 people. Not shown: conflicting registration information (no follow-up question).

Registration notices: who sent the notice

Election office40% (12)
A private group or organization20% (6)
News or social media20% (6)
Not sure20% (6)
A political party or campaign17% (5)

Share of the 30 people asked.

Trouble with a voting task, by task

Trouble registering3% (47)
Trouble checking registration3% (39)
Trouble tracking a ballot1% (23)
Trouble requesting a mail ballot1% (19)
Trouble returning a mail ballot1% (12)

Share of the 1551 people asked.

Discouraged from registering: who discouraged them

A private group or organization42% (22)
A political party or campaign30% (16)
News or social media25% (13)
Election office23% (12)
Not sure6% (3)
Other6% (3)

Share of the 53 people asked.

National Overview

August 23 to October 31, 2026. 61,908 respondents who answered all six questions.

28.5% of respondents (17,645 of 61,908) reported at least one of six election problems. The most common were misleading calls or texts (17.6%) and intimidation or harassment (8.3%); the least common was registration challenge notices (2.7%).

Share of respondents reporting each problem, national

Figure 5. Share of respondents reporting each problem, national, Aug 23–Oct 31, 2026. Each person counted once; bases vary slightly by question (61,908 to 62,354). Margin of error for the national share reporting any problem: ±0.4 percentage points. Source: CHIP50 survey data.

National: Week by week

Weekly share reporting each problem, national

Figure 6. Weekly share of respondents reporting each problem, national. The shaded band is the 95% confidence interval for the national share, which is the range the weekly figure could vary based on the sample size alone. Weeks start Monday; only problems asked about in the past day or week are shown (conflicting registration information: past week), each person counted once per week, so weekly rates are lower than full-period rates. Source: CHIP50 survey data.

Misleading calls or texts went from 7.3% in the week of Aug 31 to 17.8% in the week of Oct 26; being discouraged from registering went from 2.3% in the week of Aug 31 to 2.4% in the week of Oct 26; conflicting registration information went from 2.2% in the week of Aug 31 to 2.8% in the week of Oct 26. Weeks have 2,968 to 16,146 respondents (the first, partial week 400), so the margin of error is narrow except in that first week. From Aug 23–Oct 17 to Oct 18–Oct 31, the share of respondents reporting any of these three problems went from 11.3% to 16.3% (47,225 and 20,908 people): a rise larger than the margin of error.

National: Who reported a problem

Among groups of 200 or more people, reports were most common among people aged 18 to 24 (44.9%) and least common among people 65 and over (20.6%), against 28.5% overall. Reports rose with income and education, and fell with age. By race and ethnicity, reports were most common among Hispanic or Latino respondents (35.6%, 1923 of 5404), 8 points higher than White or Caucasian respondents (27.2%, 12441 of 45788), a difference larger than the margin of error; Asian or Asian American respondents (32.8%), Black or African American respondents (32.3%), Native American or Alaska Native respondents (30.3%), Pacific Islander or Native Hawaiian respondents (30.2%) and Other respondents (30.1%) fell in between.

GroupPeopleShare
All respondents61,90828.5%
Gender
Female36,25226.5%
Male25,65631.4%
Age
18 to 243,52444.9%
25 to 4422,84234.0%
45 to 6423,40024.8%
65 and over12,14220.6%
Household income
Under 25K16,54024.0%
25k to under 50k17,44126.5%
50K to under 75K11,83230.1%
75K to under 100K6,76632.7%
100K and over9,32935.2%
GroupPeopleShare
Race/ethnicity
White or Caucasian45,78827.2%
Black or African American10,89332.3%
Hispanic or Latino5,40435.6%
Native American or Alaska Native2,07030.3%
Asian or Asian American1,80032.8%
Other53830.1%
Pacific Islander or Native Hawaiian35830.2%
Education
Some High School or Less2,90721.1%
High School Graduate19,46125.1%
Some College15,92227.8%
College Degree18,36332.1%
Graduate Degree5,25534.8%
Party
Democrat22,06131.3%
Republican18,90228.8%
Independent/Other20,87225.2%

Share reporting at least one of the six problems, by group, national. Race/ethnicity is select-all-that-apply, so one person can appear in more than one group. Source: CHIP50 survey data.

National: Follow-up detail

Follow-up detail for respondents nationwide who reported each problem. Not shown: conflicting registration information (no follow-up question).

Registration notices: who sent the notice

Election office47% (793)
A political party or campaign30% (504)
News or social media23% (383)
A private group or organization22% (376)
Not sure13% (215)
Other3% (50)

Share of the 1,699 people asked.

Trouble with a voting task, by task

Trouble registering4% (2,192)
Trouble checking registration3% (1,983)
Trouble requesting a mail ballot2% (1,256)
Trouble tracking a ballot2% (1,204)
Trouble returning a mail ballot1% (789)

Share of all 62,354 respondents asked.

Discouraged from registering: who discouraged them

News or social media34% (929)
A political party or campaign31% (838)
Election office26% (718)
A private group or organization26% (715)
Not sure10% (268)
Other9% (242)

Share of the 2,712 people asked.

About the survey

The 2026 Voting Information Trends research Initiative tracks how people get voting information and monitors possible election-related problems reported at the county level. We focus on issues respondents say they have personally experienced or observed, including misleading election communications, intimidation, voter registration problems, and difficulties completing voting-related tasks. The goal is to identify signals that may warrant closer attention.

We survey about 1,000 U.S. adults each day through an online, nonprobability sample covering all 50 states and Washington, D.C. State-level quotas are used for race, ethnicity, age, and gender. We also collect additional responses from selected counties identified as higher risk by the Voting Rights Lab (VRL), whose 2026 risk index considers factors including electoral competitiveness, election laws and procedures, litigation, and concerns involving state and local election actors.

How to read this report

Counties with at least 50 respondents are ranked by the share of unique respondents reporting at least one monitored problem during the data-collection period. We also report results for individual problems, weekly trends, demographic groups, and when sample sizes allow, additional follow-up detail.

Weekly trends are shown only for questions that ask about experiences in the past day or week, and each respondent is counted once per week. Figures include a 95% margin of error based on sample size. We describe a county as higher than the national rate only when the difference is larger than that margin.

When a respondent selects a problem but then states in an optional written follow-up that nothing actually happened, we recode the response as not reporting the problem. These contradictions are identified using a large language model (LLM) assisted coding and reviewed by a member of the research team.

About the data

These results are signals, not estimates of prevalence. The survey is not nationally or locally representative, and results should not be used to estimate how common a problem is in the country or in any county. Figures are unweighted and describe the people who responded. Unlike individual reports found on social media, these responses come from a standardized survey using the same questions and time periods across respondents, allowing us to compare patterns across places and against a national benchmark.

The purpose of the survey is to identify issues that may warrant independent investigation or verification. Election officials, journalists, and other organizations can then assess those reports and respond when appropriate.

Because the data are self-reported, a response does not by itself verify that the reported problem occurred. Respondents may also interpret questions differently from researchers. For example, some reports of “misleading calls or texts” describe campaign messages rather than false voting information, and some registration notices may come from campaigns or advocacy organizations rather than election offices. These findings should therefore be treated as signals of where to look more closely, not counts of verified incidents.

Use of AI in this report

The research team wrote and reviewed the “About the survey” section. The summaries of the results are filled in automatically from templates built with AI. All numbers come directly from the survey data. Automated text can contain errors, so readers should check the summaries against the tables and charts.