Guns on display at a gun store in Miami on June 29, 2016. (Alan Diaz/AP)

On the first anniversary of the massacre at Marjory Stoneman Douglas High School in Parkland, Fla., NPR, “PBS NewsHour” and Marist released a poll exploring views of various approaches to gun ownership in the United States. Many of the findings are what we’ve come to expect: broad support for expanded background checks in the purchasing process; split views by party on the necessity for new restrictions on gun ownership. The extended advocacy of the surviving teenagers in Parkland notwithstanding, opinions on guns remain fairly fixed.

At the end of that poll, though, was an interesting question. The pollsters asked respondents a question about a matter of fact: Did they think that the rate of gun murders in the United States was higher or lower than it had been 25 years ago?

Twenty-five years ago was 1994, the year that President Bill Clinton signed into law a sweeping crime bill targeting violent crime and a ban on assault weapons. It was a response to a surge in violent crime that had occurred over several decades.

This history was probably not immediately at hand for respondents. So how did the poll respondents view gun violence now as opposed to then? Most believed that the rate of gun murders now is higher than it was at that point.

(Philip Bump/The Washington Post)

Notably, Democrats were much more likely than Republicans to say that the rate of gun murders was higher now than it was then. This despite President Trump’s focus during the 2016 campaign on the extent of violent crime in U.S. cities.

The responses by type of community — city, suburban, rural — were overall fairly similar, as were responses by education level. There was a fascinating split, though, between suburban men and women. Suburban and small-town women were 29 points more likely to say that the gun murder rate now is higher than it was in 1994.

(Philip Bump/The Washington Post)

To be explicit: This is wrong. Data from the FBI breaks out reported murders by type of weapon, showing a drop in the raw count of firearm murders since a peak in 1993. (Both the Marist poll and the FBI data are predicated on “murder” as a metric, a term The Washington Post generally reserves for homicides involving a criminal conviction.)

(Philip Bump/The Washington Post)

(Also note that while high-profile mass murder events such as in Parkland usually involve rifles such as the AR-15, most firearm murders — three-quarters of them since 1991 — are committed with handguns.)

There was, as Trump has noted, an uptick in firearm murders beginning in 2015, with an increased percentage of murders being committed with firearms. But in 2017, the most recent year for which data are available, the total number of firearm murders was 10,982, compared with 16,136 in 1994.

But remember, too, that the question asked about the murder rate — that is, the number of murders as a function of the population. In that regard, the drop is even steeper. In 1994, the FBI data suggest a rate of about 6.2 firearm murders per 100,000 people. In 2017, the rate was 3.38 murders, up from the 2014 low in which the rate was less than half of that in 1994.

(Philip Bump/The Washington Post)

In other words, gun homicides have dropped substantially over the past 25 years — but most Americans believe the opposite to be true. Why? Perhaps in part because of the media focus on multiple-victim shooting incidents in recent years. Perhaps, too, because of the number and deadliness of those incidents. We’ve noted before that the number of fatalities in major mass-shooting incidents has increased dramatically in recent years; it’s possible that people are conflating increases in frequency and deadliness of mass shootings with the United States getting more dangerous generally.

It’s hard not to assume that this influences the politics of the gun debate. Democrats are both much more likely to support new gun restrictions and to believe that the rate of gun murders is higher than it was in 1994, a year that was near the peak of violent crime in recorded U.S. history. Does the latter inform the former?

The good news is that gun homicides are down substantially both as a raw total and as a function of population since 1994. The bad news is that many Americans don’t seem to realize that.