Read the summary of this post here

What if we checked what a fact actually proves before we let it change our minds?

Below are five claims about voting.  Each one was reported as fact by a real organization.  Before you read further, decide for each one:  True or not?

  1. A national think tank documented more than 1,400 proven cases of election fraud.
  2. A study found 21 million American adults do not have government-issued photo ID.
  3. Federal investigators estimated 100,000 fraudulent ballots were cast in one Chicago election.
  4. 84% of Americans support requiring photo ID to vote.
  5. Ohio's Secretary of State flagged 521 cases of potential noncitizen voting.

Write down your answers.  We will come back to them.

Facts Still Matter

Opinions are everywhere, and most of them cost nothing to produce.  Solutions that last have to be built on facts. 

This is why every factual claim we publish is tied to its original source - the statute, the court record, the database, or the study itself - and we list those sources so you can check our work.

WARNING:  A claim can include a real fact and still mislead you. 

Sometimes someone lied.  Other times it’s simply a mistake.

More often the problem is where the fact came from or how it was presented:  An exaggeration, a partial fact, one side of the data with the other side left out, or a fact followed by an opinion that is made to sound like part of it.  Any of these can stir up fear, and fear leads to misunderstandings, poor decisions, and stress.

If we care about facts - and Simple Liberty does - we have to look at the details and measure them against a clear standard of truth.

Data Integrity

The more reliable the data, the closer it gets us to the truth.

Surveys are some of the easiest data to manipulate when a person or organization wants to justify or prove something - even when they know their presupposition is not entirely true. 

Sometimes the presupposition is simply a foolish mistake. 

Here are five ways survey data gets twisted from simple truth:

  1. Wording.  How a question is asked changes the answer.  Ask about fraud first, and support for voter ID goes up.  Ask about voters being turned away first, and it goes down.
  2. Who gets asked.  A survey can oversample groups likely to give the answer its sponsor wants.  Even honest surveys now rely heavily on opt-in online panels, and the people who sign up for those panels are different from the people who don't.
  3. Who answers.  Response rates below one in ten are common, and the nine who hang up are not random.  When is the last time you answered a survey?  Busy people, who might be successful and sharp, often don't bother.  Plus, many people are worried about scams and fraud - as a result, they do not participate in surveys from unknown groups.
  4. Misreporting.  Someone may say they have no driver's license when one is sitting in a drawer, or say they have one that expired years ago.
  5. Small groups.  A poll's advertised margin of error covers the whole sample.  A finding about one small group inside that sample can be off by three times as much, or more.

Data from other sources can be intentionally manipulated - or accidentally misrepresented too - each in its own way:

  1. Fraud databases count the cases that were caught and prosecuted.  They undercount whatever went undetected.  This can make fraud appear rarer than it is, and make the kinds of fraud that do get caught look more typical than they are.
  2. Audits find what their matching method is designed to find, which is not everything that exists.
  3. Court findings of fact describe the evidence in one case, in one state, at one point in time.  They don't automatically describe the whole country today or include details that could change the conclusion someone is trying to confirm.
  4. Academic studies depend on the researchers' modeling choices, and journals are more likely to publish a dramatic finding than a study that finds nothing.  At least two widely cited voting studies fell apart when other researchers reran the numbers.  Also, like any other survey or study, academic studies can be initiated to prove a belief or assumption of the researcher(s).
  5. News stories report the charge.  They rarely go back to report whether it held up.

ENDNOTES:  1-5

How Much Can Data Be Misrepresented?

Far enough that two sides can quote the same database and reach opposite conclusions.

In the voter ID debate, one side reports, accurately, that a national fraud database contains more than 1,400 proven cases. 

The other side reports, also accurately, that a separate review of thousands of fraud cases found roughly ten of the one kind of fraud a photo ID at the polling place would stop - a person showing up and voting as someone else.

Both statements are true, and both leave something out.  Often that is deliberate, because the goal is to win you over rather than inform you.

WHAT YOU CAN DO:  The foundation of clear thought is to pause, think, and start asking questions. 

Then, use our 4Q Scrolls process to consider the data.  Read further below for a summary of 4Q Scrolls or click here for a more complete explanation.

ENDNOTE:  6

The Answers

All five statements are accurate.  Not all of them prove what they seem to claim.

That is why "True or not?" was the wrong question.  The better question is:  What does this fact actually prove?

  1. Fact presented:  A national think tank documented more than 1,400 proven cases of election fraud.

Answer:  The database is real, and each case ended in a conviction, a legal finding, or an overturned election.  It shows that election fraud happens. 

However, the soundbite does not present how many people voted in those elections, which would help you understand the amount of fraud in that time period.  The 1,400 is a partial truth.

  1. Fact presented:  A study found 21 million American adults do not have government-issued photo ID.

Answer:  The 987 survey answers are real.  About 11% of those surveyed said they lacked a current government-issued photo ID, and that percentage was applied to the full voting-age citizen population to reach 21 million.  

That is a common way to estimate, but this one has real limits:  A small sample, answers people reported about themselves (see Misreporting above), and data nearly twenty years old.

The 21 million is an estimate, not a fact - and an outdated one.

  1. Fact presented:  Federal investigators estimated 100,000 fraudulent ballots were cast in one Chicago election.

Answer:  A federal prosecutor estimated 100,000.  The number proven in court was 63 convictions - still the largest voter fraud prosecution in Justice Department history. The two numbers measure different things: Convictions count people, and one corrupt election worker can cast many fraudulent ballots.

Even so, the 100,000 is an estimate, not a proven fact.  Check the source to see how it was reached. 

  1. Fact presented:  84% of Americans support requiring photo ID to vote.

Answer:  This one deserves more trust than most survey numbers, because multiple polls have confirmed it.  Three polling firms, with different methods and different leanings, found support between 77% and 84% within a few months of each other. 

When independent sources land in the same place, the finding is more reliable - although watch the sources.  Multiple organizations can be sincere, or not.

  1. Fact presented:  Ohio's Secretary of State flagged 521 cases of potential noncitizen voting.

Answer:  521 cases were flagged.  A later review found most were clerical errors or paperwork that had not caught up yet, and one case led to a charge. 

The 521 flagged cases were not a fact of fraud.  Only one (0.19%) led to a charge.    

The raw facts and estimates may still have value.  However, to make good decisions we must discern facts from estimates, partial truths, and lies.  None of the five claims above was a lie - but many claims you will hear are.

ENDNOTES:  7-11

How to Investigate a Fact

People make claims based on assumptions, bias, partial facts, and misunderstandings every day.

When someone makes a claim, pause and think about it.

Unless it is necessary for your survival to make a quick decision, follow this four-step process to discern the truth based on facts - not just feelings.  We call this process the 4Q Scrolls.

Q1 - MEANING  |  What do you mean?

Ask questions to confirm what the person or group means by their claim. 

The outcome of Meaning is to sincerely consider the person or group’s opinion and understand their claim.

Q2 - EVIDENCE  |  Why do you believe that’s true?

Ask questions to gather facts that support or contradict a claim.  Be alert - partial truths, exaggerations, assumptions, and lies are not evidence that supports a claim.

Depending on the person’s claim, do your research.  AI can help you find sources quickly.  However, it makes mistakes. 

The outcome of Evidence is to determine whether the claim is fully true, partly true, or unsupported.  Not to judge anyone, but to discern the truth of the claim.

Q3 - IMPACT  |  Who and what are affected?

Ask questions to consider what happens if the claim becomes reality - who and what are affected.  Measure the change by magnitude, the probability it happens as expected, and the time it takes to implement.

Every choice for something is a choice against something else.  Every decision carries positive and negative outcomes - large or small.

The outcome of Impact is to know who gains, who loses, and by how much.

Q4 - SOLUTION  |  What is best for the long-term common good?

Ask questions to determine the best way to serve the long-term common good based on the facts. 

This step brings together what Meaning, Evidence, and Impact produced - including your own research - so the solution is based on facts rather than opinions or assumptions.

The outcome of Solution is to name the solution that the facts support as best for the common good - today and into the future.

Test 4Q Scrolls yourself.

Every example in this post is real and documented, with sources listed in the endnotes below.  Apply the 4Q Scrolls process to any of them - or any other claims you want to test.

What's Next?

Simple Liberty is regular citizens researching major issues, one at a time, and proposing solutions to America's biggest problems.

In the posts that follow, we apply this same approach to the three branches of government our Founders designed to serve us, not rule us - beginning with the President, then Congress, then the courts.

Read our Commoners Corner posts, and join the discussion in The Commons, where we present and debate each issue.

We can do this.

Lasting restoration occurs when we - American citizens - strengthen our character, comprehend the facts, and gratefully steward the people's government for the common good.

Simple Liberty unites everyday Americans to research the facts behind our nation's biggest problems and develop solutions for the common good

…because politicians and billionaires won't do it for us.

Join Us

ENDNOTES

These references are meant to introduce you to the data, not provide every source.  We encourage you to explore more on your own, and welcome your submissions as comments on this post.

Data Integrity

(1) Question Wording Changes the Answer

Writing Survey Questions, Pew Research Center - Explains how the wording and order of questions can shift results, with examples from Pew's own testing.  |  https://www.pewresearch.org/our-methods/u-s-surveys/writing-survey-questions/

(2) Opt-In Online Panels

Online Opt-In Polls Can Produce Misleading Results, Pew Research Center, 2024 - Found opt-in online samples produced notably less accurate estimates than probability-based samples, especially for young adults and Hispanic adults.  |  https://www.pewresearch.org/short-reads/2024/03/05/online-opt-in-polls-can-produce-misleading-results-especially-for-young-people-and-hispanic-adults/

(3) Low Response Rates

What Low Response Rates Mean for Telephone Surveys, Pew Research Center, 2017 - Documents telephone response rates falling to about 9 percent, and examines how that affects accuracy.  |  https://www.pewresearch.org/methods/2017/05/15/what-low-response-rates-mean-for-telephone-surveys/

(4) Margins of Error for Small Groups

5 Key Things to Know About the Margin of Error in Election Polls, Pew Research Center, 2016 - Explains that the margin of error for a subgroup is larger than for the full sample, because the subgroup is smaller.  |  https://www.pewresearch.org/short-reads/2016/09/08/understanding-the-margin-of-error-in-election-polls/

(5) Studies That Fell Apart

The Perils of Cherry Picking Low Frequency Events in Large Sample Surveys, Ansolabehere, Luks, and Schaffner, Electoral Studies, 2015 - Showed that a widely reported 2014 estimate of noncitizen voting was driven by survey response errors.  |  https://web.stanford.edu/group/bps/cgi-bin/wordpress/wp-content/uploads/2015/04/The-Perils-of-cherry-picking-low-frequency-events-in-surveys.pdf

Also see Obstacles to Estimating Voter ID Laws' Effect on Turnout, Grimmer and others, Journal of Politics, 2018 - Found that a widely cited 2017 voter ID study's results did not hold up when the data was reanalyzed.  |  https://doi.org/10.1086/696618

How Much Can Data Be Misrepresented?

(6) The Same Facts, Opposite Conclusions

Source to confirm:  the "roughly ten cases" of in-person impersonation figure.  See note in the cover message.  |  https://votingrights.news21.com/article/election-fraud/

The Answers

(7) The Fraud Database

Election Fraud Database Tops 1,400 Cases, The Heritage Foundation - Each entry is a proven instance of election fraud resulting in a criminal conviction, civil penalty, official finding, or overturned election.  The authors note the database is a sampling, not a complete count.  |  https://www.heritage.org/election-integrity/commentary/election-fraud-database-tops-1400-cases

(8) The 21 Million Estimate

Citizens Without Proof, Brennan Center for Justice, November 2006 - A telephone survey of 987 voting-age citizens, conducted by Opinion Research Corporation November 16-19, 2006, found about 11 percent lacked current government-issued photo ID, which the Center projected to roughly 21 million Americans.  |  https://www.brennancenter.org/media/6697/download/download_file_39242.pdf?inline=1

(9) The Chicago Estimate and the Count

Where There's Smoke, There's Fire:  100,000 Stolen Votes in Chicago, The Heritage Foundation - Describes the federal investigation and the prosecutor's estimate, alongside the 63 convictions that resulted.  |  https://www.heritage.org/election-integrity/report/where-theres-smoke-theres-fire-100000-stolen-votes-chicago

(10) Support for Voter ID

Voter ID Fact Brief, Gigafact - Reviews multiple national polls showing between 77 and 84 percent of Americans support photo ID requirements to vote.  |  https://gigafact.org/fact-briefs/voter-id-americans-support-wisconsin-poll-photo-identification/

(11) The Ohio Referrals

Ohio Sec. of State LaRose Flagged More Than 520 Cases of Noncitizen Voter Fraud; Only One Was Legit, Ohio Capital Journal, September 27, 2023 - Reports the outcome of the referrals, most of which traced to clerical errors and paperwork lag.  |  https://ohiocapitaljournal.com/2023/09/27/ohio-sec-of-state-larose-flagged-more-than-520-cases-of-noncitizen-voter-fraud-only-one-was-legit/

 

Image source: Seacalm/Shutterstock.com