Scientific Survey Shows Voters Widely Accept Misinformation Spread By the Media
By James D. Agresti
January 2, 2020
NBC News reporter and political director Chuck Todd recently railed against “misinformation” and singled out President Trump and “the right” for having an “incentive structure” to spread it. Todd, who according to NBC, “is responsible for all aspects of the network’s political coverage,” also stated that Republicans criticize the media for “sport” and “the loudest chanters of fake news” are “the ones who, under a lie detector, would probably take our word over any word they’ve heard from the other side on whether something was poisonous or not.”
Speaking directly to those unsupported claims, a scientific survey commissioned by Just Facts shows that many people are indeed misinformed—but contrary to Todd—this is a bipartisan affair. In fact, the survey found that the most commonly believed misinformation accords with left-leaning narratives spread by the press, and Democrat voters are more likely to accept these falsehoods than Trump voters. Furthermore, sizable portions of Trump voters have swallowed some of these media-promoted liberal fictions, as well as some conservative ones.
The findings are from a nationally representative annual survey commissioned by Just Facts, a non-profit research and educational institute. The survey was conducted by Triton Polling & Research, an academic research firm that used sound methodologies to assess U.S. residents who regularly vote.
While most polls measure public opinion, this unique one measures voters’ knowledge of major issues facing the nation—such as education, taxes, healthcare, national debt, pollution, government spending, Social Security, global warming, energy, and hunger. Every year, the poll includes a new question about a prevalent, controversial issue. This year, it is about poverty.
For each question, voters were offered a selection of two or more answers, one of which was true. Voters also had the opportunity to say they were unsure.
On average, voters gave the correct answer 39% of the time, gave an incorrect answer 54% of the time, and said they were unsure 6% of the time. A majority of voters gave the correct answer to only 5 of the 24 questions.
The highest levels of misinformation were found on questions relating to child hunger, tax burdens, poverty, landfills, health insurance copayments, and two elements of Social Security finances. For these 7 questions, 75% or more of voters provided an incorrect answer.
Among 8 of the 10 questions in which the electorate was most deluded, the wrong answers they gave accorded with progressive storylines propagated by the media. Moreover, these answers were often far removed from reality, not just slightly mistaken.
For example, 79% of voters think that the middle class pays a greater portion of their income in federal taxes than the top 1%. Yet, the Congressional Budget Office, the U.S. Treasury, and the Tax Policy Center have all documented that households in the top 1% pay an average federal tax rate that is about 2.5 times higher than that of the middle class. More specifically, the latest Congressional Budget Office data on federal taxes shows that on average in 2016:
Nevertheless, media outlets commonly report the opposite based on deceptive studies that exclude large portions of people’s taxes and/or incomes. A remarkable 93% of Democrat voters have accepted this canard, as well as 65% of Trump voters.
The survey also recorded voters’ ages, genders, and who they planned to vote for in the upcoming presidential election (Donald Trump, the eventual Democrat nominee, or a third-party candidate). This allows the survey to pinpoint segments of society that are most and least informed about specific issues.
The results show deep partisan and demographic divides, with different groups being more or less knowledgeable depending upon the questions. In total, the rates at which voters gave the correct answers varied from a high of 46% for Trump voters to a low of 32% for Democrat voters, with others falling in between as follows
The sample sizes of unsure and third-party voters were too small to produce meaningful data.
Question 1: Relative to other nations, how do you believe U.S. fourth graders rank in terms of their reading and math ability? Are they in the bottom 50% or in the top 50%?
Correct Answer: Top 50%. In international tests administered to students in dozens of nations, U.S. fourth graders rank in the top 30% of nations for reading and for math. Confusion about this issue may stem from the fact that the relative performance of U.S. students declines over time, and by the age of 15, they drop to the bottom 50% in reading and to the bottom 20% in math. This suggests that the problems of the U.S. education system may occur in the later years, not the early years, as many have claimed.
Correct answer given by 47% of all voters, 46% of Democrat voters, 48% of Trump voters, 46% of males, 48% of females, 53% of 18 to 34 year olds, 49% of 35 to 64 year olds, and 42% of 65+ year olds.
Question 2: On average across the United States, how much do you think public schools spend per year to educate each classroom of students? Less or more than $150,000 per classroom per year?
Correct Answer: More than $150,000. The average cost to educate a classroom of public school students is about $332,000 per year. In contrast to a drumbeat of media stories decrying education funding cuts, Department of Education data shows that the average inflation-adjusted spending per public school student has risen by more than three times since 1960.
Correct answer given by 36% of all voters, 26% of Democrat voters, 45% of Trump voters, 46% of males, 28% of females, 25% of 18 to 34 year olds, 40% of 35 to 64 year olds, and 33% of 65+ year olds.
Question 3: In your mind, what portion of 17- to 24-year-olds in the U.S. are unqualified for military service because of poor physical fitness, weak educational skills, illegal drug usage, medical conditions, or criminal records?
Correct Answer: More than half. According to various agencies within the Department of Defense, two-thirds to three-quarters of all 17- to 24-year-olds are unqualified for military service because of poor physical fitness, weak educational skills, illegal drug usage, medical conditions, or criminal records.
Correct answer given by 43% of all voters, 37% of Democrat voters, 51% of Trump voters, 44% of males, 42% of females, 40% of 18 to 34 year olds, 46% of 35 to 64 year olds, and 41% of 65+ year olds.
Question 4: When conventional public schools are subject to school choice programs that allow students to leave for private or charter schools, do you think the children who remain in the public schools academically decline?
Correct Answer: No. At least 21 high-quality studies have been performed on the academic outcomes of students who remain in public schools that are subject to school choice programs. All but one of the studies found neutral-to-positive results, and none found negative results. This is consistent with the theory that school choice stimulates competition that helps public schools to improve.
Correct answer given by 46% of all voters, 40% of Democrat voters, 51% of Trump voters, 49% of males, 43% of females, 53% of 18 to 34 year olds, 47% of 35 to 64 year olds, and 43% of 65+ year olds.
Question 5: The average U.S. household spends about $30,000 per year on food, housing, and clothing combined. If we broke down all combined federal, state, and local taxes to a per household cost, do you think this would amount to more or less than an average of $30,000 per household per year?
Correct Answer: More than $30,000. In 2018, federal, state and local governments collected a combined total of $5.1 trillion in taxes or an average of $40,000 for every household in the U.S.
Correct answer given by 43% of all voters, 36% of Democrat voters, 47% of Trump voters, 42% of males, 44% of females, 51% of 18 to 34 year olds, 46% of 35 to 64 year olds, and 38% of 65+ year olds.
Question 6: On average, who would you say pays a greater portion of their income in federal taxes: The middle class or the upper 1% of income earners?
Correct Answer: The upper 1%. The Congressional Budget Office, the U.S. Treasury, and the Tax Policy Center have all documented that households in the top 1% of income pay an average effective federal tax rate of about 33%, while middle-income households pay about 13%. These tax rates account for nearly all income and federal taxes. Claims to the contrary—often voiced by politicians and the media—are based on misleading calculations that exclude large portions of people’s taxes and/or incomes.
Correct answer given by 18% of all voters, 6% of Democrat voters, 30% of Trump voters, 21% of males, 15% of females, 11% of 18 to 34 year olds, 19% of 35 to 64 year olds, and 19% of 65+ year olds.
Question 7: Now, changing the subject from taxes to spending, suppose we broke down all government spending to a per household cost—do you think the combined spending of federal, state and local governments amounts to more or less than $40,000 per household per year?
Correct Answer: More than $40,000. In 2018, federal, state and local governments spent a combined total of $6.9 trillion, or an average of about $54,000 for every household in the United States. For reference, the average U.S. household spends about $45,000 per year on food, housing, clothing, transportation, and healthcare combined.
Correct answer given by 48% of all voters, 44% of Democrat voters, 53% of Trump voters, 53% of males, 43% of females, 53% of 18 to 34 year olds, 52% of 35 to 64 year olds, and 43% of 65+ year olds.
Question 8: Do you think the federal government spends more money on social programs, such as Medicare, education, and food stamps—or does the federal government spend more money on national defense, such as the Army, Navy, and missile defense?
Correct Answer: Social programs. In 2018, 62% of federal spending was for social programs, and 18% was for national defense. In 1960, the opposite was true, and 53% of federal spending was for national defense, while 21% was for social programs. Reporters sometimes mislead the public about the composition of federal spending by using a subset of spending that omits the vast majority of social programs.
Correct answer given by 36% of all voters, 14% of Democrat voters, 59% of Trump voters, 40% of males, 33% of females, 23% of 18 to 34 year olds, 36% of 35 to 64 year olds, and 41% of 65+ year olds.
Question 9: What about federal government debt? The average U.S. household owes about $122,000 in consumer debt, such as mortgages and credit cards. Thinking about all federal government debt broken down to a per household basis, do you think the average federal debt per U.S. household amounts to more or less than the average consumer debt per U.S. household?
Correct Answer: More than $122,000. Federal debt is now $23.1 trillion or about $180,000 for every household in the United States. Such levels of debt can have far-reaching negative effects on wages, living standards, and government and personal financial security.
Correct answer given by 77% of all voters, 76% of Democrat voters, 81% of Trump voters, 75% of males, 80% of females, 84% of 18 to 34 year olds, 79% of 35 to 64 year olds, and 75% of 65+ year olds.
Question 10: From the time that the Great Recession ended in 2009, which do you think has grown at a faster rate, the U.S. economy or the national debt?
Correct answer given by 77% of all voters, 89% of Democrat voters, 65% of Trump voters, 71% of males, 82% of females, 76% of 18 to 34 year olds, 77% of 35 to 64 year olds, and 77% of 65+ year olds.
During Obama’s presidency, the party split on this question was reversed, and 88% of Republican voters and 73% of Democrat voters answered correctly.
Question 11: Would you say the earth has become measurably warmer since the 1980s?
Correct Answer: Yes. According to both satellite-measured data and ground-level thermometers, the earth’s average temperature has increased by about 0.7 to 0.9 degrees Fahrenheit since the 1980s. This increase is greater than the range of measurement uncertainty. Providing a sense of scale for this change, a temperature analysis of a glacier in Greenland found that it was about 22ºF colder during the last ice age than it is now.
Correct answer given by 62% of all voters, 94% of Democrat voters, 29% of Trump voters, 57% of males, 67% of females, 83% of 18 to 34 year olds, 57% of 35 to 64 year olds, and 64% of 65+ year olds.
Question 12: Again, thinking about the whole planet, do you think the number and intensity of hurricanes and tropical storms have generally increased since the 1980s?
Correct Answer: No. Comprehensive global data shows that the number and intensity of cyclones and hurricanes has been roughly level since the 1980s. This data was originally published in the journal Geophysical Research Letters in 2011 and updated this year. Likewise, the Intergovernmental Panel on Climate Change has reported: “There is low confidence in any observed long-term (i.e., 40 years or more) increases in tropical cyclone activity (i.e., intensity, frequency, duration), after accounting for past changes in observing capabilities.” Regional data that extends back for more than century shows the same. Nonetheless, media outlets have spread false claims to the contrary by cherry-picking anecdotes and reporting computer model predictions instead of hard facts.
Correct answer given by 32% of all voters, 4% of Democrat voters, 59% of Trump voters, 40% of males, 25% of females, 19% of 18 to 34 year olds, 36% of 35 to 64 year olds, and 30% of 65+ year olds.
Question 13: Now, just thinking about the United States, in your opinion, is the air generally more polluted than it was in the 1980s?
Correct Answer: No. EPA data shows that ambient levels of all criteria air pollutants have declined significantly since the 1980s. Criteria air pollutions are those that are deemed by the administrator of the EPA to be widespread and to “cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare….” Likewise, combined emissions of hazardous air pollutants have declined by about 50% since the 1990s. Lower pollution levels can improve human health and reduce problems like learning deficits and behavioral disorders.
Correct answer given by 56% of all voters, 44% of Democrat voters, 67% of Trump voters, 67% of males, 46% of females, 47% of 18 to 34 year olds, 63% of 35 to 64 year olds, and 49% of 65+ year olds.
Question 14: If the U.S. stopped recycling and buried all of its municipal trash for the next 100 years in a single landfill that was 30 feet high, how much of the nation’s land area would you think this landfill would cover? Less than 1%, 1% to less than 5%, or more than 5%?
Correct Answer: Less than 1%. At the current U.S. population growth rate and the current per-person trash production rate, the landfill would cover 0.06% of the nation’s land area. More realistically, the actual area in use will be an order of magnitude smaller, because:
Despite these facts, media outlets have been reporting for more than 30 years that the U.S. is “nearly out of space” to dispose of trash. NBC News recently published a series of videos called Wasteland that claims U.S. landfills are “set to reach max capacity by 2030” and “scientists are racing against time to find new ways to hack them for the future.” The video presents no source or evidence to support these statements.
Correct answer given by 7% of all voters, 8% of Democrat voters, 9% of Trump voters, 12% of males, 3% of females, 6% of 18 to 34 year olds, 8% of 35 to 64 year olds, and 6% of 65+ year olds.
Question 15: Without government subsidies, which of these technologies do you think is the least expensive method for generating electricity? Wind turbines, solar panels, or natural gas power plants?
Correct Answer: Natural gas power plants. Determining the costs of electricity-generating technologies is complex, but data from the U.S. Energy Information Administration shows that natural gas is considerably less expensive than wind, and wind is considerably less expensive than solar. Affordable energy has many important benefits, and for poorer people, it can mean the difference between life and death.
Correct answer given by 40% of all voters, 23% of Democrat voters, 57% of Trump voters, 53% of males, 29% of females, 25% of 18 to 34 year olds, 43% of 35 to 64 year olds, and 41% of 65+ year olds.
Question 16: Without government subsidies, which of these fuels do you believe is least expensive for powering automobiles? Gasoline, ethanol, or biodiesel?
Correct Answer: Gasoline. Data from the U.S. Department of Energy and U.S. Energy Information Administration show that in 2017, the unsubsidized cost of ethanol was 32% more than gasoline, and the unsubsidized cost of biodiesel was 119% more than gasoline.
Correct answer given by 46% of all voters, 35% of Democrat voters, 55% of Trump voters, 56% of males, 37% of females, 30% of 18 to 34 year olds, 48% of 35 to 64 year olds, and 49% of 65+ year olds.
Question 17: Worldwide, which of these technologies generates the most electricity? Solar panels, natural gas power plants, coal power plants, or nuclear power plants?
Correct Answer: Coal power plants. Due to the low cost and widespread availability of coal, coal power plants generate about 35% of the world’s electricity, as compared to 23% for natural gas, 11% for nuclear, and 4% for solar.
Correct answer given by 40% of all voters, 34% of Democrat voters, 43% of Trump voters, 55% of males, 26% of females, 36% of 18 to 34 year olds, 42% of 35 to 64 year olds, and 36% of 65+ year olds.
Question 18: On an average day, what portion of U.S. households with children do you believe will have at least one child who experiences hunger? Less than 1%, 1% to 10%, or more than 10%?
Correct Answer: Less than 1%. Per the latest data from the USDA, 0.14% or less than one out of every 700 U.S. households with children have any child who experiences hunger on an average day. This includes children who are hungry due to poverty, not those who skip meals because they are late for school, don’t feel like eating, or are trying to lose weight.
Media stories claiming that child hunger is much more prevalent often falsely equate USDA data on “food insecurity” with “hunger.” Yet, the USDA explicitly states that “low food security” means “reports of reduced quality, variety, or desirability of diet” but “little or no indication of reduced food intake.” Prior to 2006, the USDA’s label for such households reflected this reality and was called “food insecurity without hunger.”
Correct answer given by 12% of all voters, 2% of Democrat voters, 22% of Trump voters, 15% of males, 9% of females, 3% of 18 to 34 year olds, 12% of 35 to 64 year olds, and 13% of 65+ year olds.
Question 19: Do you think Social Security’s financial problems stem from politicians looting the program and spending the money on other programs?
Correct Answer: No. By law, all Social Security taxes and revenues can be used only for the Social Security program, and the federal government has never failed to abide by this law. What some call “looting” is actually a legal requirement (established in the original Social Security of 1935) that all of the program’s surpluses be loaned to the federal government. The government is required to pay back this money with interest, and it has been doing this since 2010.
Contrary to the notion that politicians have looted Social Security, they have actually added money to the program by repeatedly raising its payroll tax rate, increasing its inflation-adjusted taxable maximum, and injecting other taxes to its income stream. This has placed progressively higher Social Security tax burdens on successive generations of American workers. Yet, the program is still facing insolvency, mainly because the ratio of workers paying taxes to people receiving benefits has fallen by three times since 1955 and is projected to fall further.
Correct answer given by 20% of all voters, 28% of Democrat voters, 13% of Trump voters, 25% of males, 15% of females, 24% of 18 to 34 year olds, 23% of 35 to 64 year olds, and 15% of 65+ year olds.
Question 20: Some policymakers are proposing that individuals be allowed to save and invest some of their Social Security taxes in personal accounts instead of paying these taxes to the Social Security program. In your view, do you think such proposals generally improve or harm the finances of the Social Security program?
Correct Answer: Improve. As shown by analyses conducted by the chief actuary of the Social Security Administration and a bipartisan presidential commission, proposals to give Social Security an element of personal ownership generally strengthen the program’s finances. Although some tax revenues that would have gone to the program instead go to people’s personal retirement accounts, these tax revenues are more than offset by the savings of not paying these individuals full benefits. Media outlets like NBC and the New York Times have misleadingly portrayed personal ownership proposals in a negative light.
Correct answer given by 22% of all voters, 11% of Democrat voters, 33% of Trump voters, 28% of males, 17% of females, 31% of 18 to 34 year olds, 20% of 35 to 64 year olds, and 23% of 65+ year olds.
Question 21: In 1960, governments paid for 24% of all healthcare costs in the U.S. Do you think governments now pay a greater portion or a lesser portion of all healthcare costs in the U.S.?
Correct Answer: A greater portion. In 2017, governments paid for 49% of all healthcare expenses in the United States.
Correct answer given by 57% of all voters, 42% of Democrat voters, 70% of Trump voters, 62% of males, 52% of females, 46% of 18 to 34 year olds, 57% of 35 to 64 year olds, and 59% of 65+ year olds.
Question 22: When health insurance copayments are high, people tend to spend less on healthcare. Do you think that this reduced spending typically has a negative impact on their health?
Correct Answer: No. Multiple studies have shown that when copayments are high, people generally spend less money on their healthcare without negatively impacting their health. This is because when people directly pay for more of their healthcare bills, they are more likely to be responsible consumers and use only those services that actually benefit their health. An exception to this rule is the poorest 6% of the population, who do experience negative effects when copayments are increased.
Correct answer given by 13% of all voters, 4% of Democrat voters, 22% of Trump voters, 18% of males, 9% of females, 14% of 18 to 34 year olds, 14% of 35 to 64 year olds, and 12% of 65+ year olds.
Question 23: In 2010, Congress passed and President Obama signed the Affordable Care Act, also known as “Obamacare.” This law uses price controls to save money in the Medicare program. Do you think these price controls will worsen Medicare patients’ access to care?
Correct Answer: Yes. As explained by Medicare’s actuaries, the price controls in the Affordable Care Act will cut Medicare prices for many medical services over the next three generations to “less than half of their level under the prior law.” The actuaries have been clear that this will likely cause “withdrawal of providers from the Medicare market” and “severe problems with beneficiary access to care.”
Correct answer given by 50% of all voters, 17% of Democrat voters, 80% of Trump voters, 53% of males, 46% of females, 38% of 18 to 34 year olds, 52% of 35 to 64 year olds, and 49% of 65+ year olds.
Question 24: Including government benefits and private charity, how much worth of goods and services do the poorest 20% of U.S. households consume on average each year? Less than $20,000, $20,000 to $40,000, or more than $40,000?
According to the latest government data, the poorest 20% of U.S. households consumed an average of $57,049 of goods and services per household in 2010. The much-lower poverty statistics often cited by the media exclude about 80% of the material resources of the poor because they:
Correct answer given by 13% of all voters, 6% of Democrat voters, 20% of Trump voters, 13% of males, 14% of females, 15% of 18 to 34 year olds, 16% of 35 to 64 year olds, and 9% of 65+ year olds.
The survey was conducted by Triton Polling & Research, an academic research firm that serves scholars, corporations, and political campaigns. The responses were obtained through live telephone surveys of 700 likely voters across the U.S. during December 2–11, 2019. This sample size is large enough to accurately represent the U.S. population. Likely voters are people who say they vote “every time there is an opportunity” or in “most” elections.
The margin of sampling error for the total pool of respondents is ±4% with at least 95% confidence. The margins of error for the subsets are 6% for Democrat voters, 6% for Trump voters, 5% for males, 5% for females, 12% for 18 to 34 year olds, 5% for 35 to 64 year olds, and 6% for 65+ year olds.
The survey results presented in this article are slightly weighted to match the ages and genders of likely voters. The political parties and geographic locations of the survey respondents almost precisely match the population of likely voters. Thus, there is no need for weighting based upon these variables. The complete weighted and unweighted results are available here: