When there’s a mass shooting of the kind that killed 17 in Parkland, Florida, last month, public health officials, scientists, and now even teenage survivors call for more research into gun violence.

But often these pleas fall on deaf ears.

The Centers for Disease Control and Prevention (CDC) spends a large amount of time and effort to track data on a variety public health threats, from gastrointestinal virus outbreaks on cruise ships to sleep disorders.

But comprehensive data on gun injuries and deaths remain elusive.

Now as survivors and other students plan to protest in part for more research during the “March for Our Lives” protest, we look at the reasons why the United States has so little information on gun injuries and deaths.

According to the CDC, thousands of people die in gun-related incidents every year in the United States. The best estimate is that in 2015 — the most recent year that data is available — approximately 12,979 people died due to firearm homicides. Another 22,018 people died by suicide using a firearm.

Tens of thousands more people are injured by firearms, whether intentionally or by accident.

Yet there’s no comprehensive or recent federal data on gun-related injuries or deaths — and that includes the CDC data above.

What data on gun-related injuries and deaths does the CDC have?

The CDC currently records the number of gun-related injuries and deaths in 40 states, the District of Columbia, and Puerto Rico, as part of their National Violent Death Reporting System database (NVDRS.)

The figures from 10 states aren’t included in the database, which could mean the total is a gross underestimate of the true tally.

“We can’t say very much about the states we don’t have data on,” said David Hemenway, PhD, a gun violence researcher and professor of health policy at Harvard T.H. Chan School of Public Health’s Department of Health Policy and Management.

That lack of data makes it hard to draw conclusions about gun violence in the United States as a whole.

“My old stats teacher used to say to me, ‘It’s easy to lie with statistics, but it is a lot easier to lie without statistics,’” Hemenway said.

Why doesn’t the CDC have better data on gun-related injuries and deaths?

Three words: the Dickey Amendment. The amendment was a rider on a 1996 congressional spending bill, dubbed in honor of its sponsor, Arkansas Republican Congressman Jay Dickey.

The CDC has approximately a $6 billion budget, of which 0.25 percent goes to the National Violent Death Reporting System database (NVDRS.)

What is the Dickey Amendment?

The amendment was the result of intense lobbying by the National Rifle Association (NRA) in response to a 1993 study, which the CDC funded. The study’s authors found that owning a gun was linked to a higher risk of homicide in the home.

The NRA argued that the study’s findings were biased and constituted anti-gun advocacy. They lobbied the government to shut down the CDC’s National Center for Injury Prevention and Control entirely.

To compromise, Congress settled on the Dickey Amendment, which states: “None of the funds made available for injury prevention and control at the Centers for Disease Control and Prevention may be used to advocate or promote gun control.”

This explicit language was more a warning shot than a new rule.

“The Dickey Amendment was a signal,” said Hemenway. “It says something that was already known: that you can’t use any federal research dollars for lobbying on anything. It’s a reminder that they’re watching you.”

At the same time, Congress lowered the CDC’s budget for the coming fiscal year by $2.6 million — the exact amount the agency had spent on gun-related research the year before the amendment was passed. The CDC did get their $2.6 million back, but by then, the damage had been done.

There was and continues to be no ban on gun-related research. There didn’t have to be.

The CDC got the message.

“It made it really clear that if the CDC does any research on guns, it is going to be called in front of Congress and the Appropriations Committee and it is going to be threatened that large amounts of its funding is going to be eliminated,” said Hemenway.

Dickey later regretted the effect of the Dickey Amendment on research. In 2012, he co-wrote an editorial published in The Washington Post calling for more research into gun violence.

What about other public health agencies?

In the wake of the Dec. 2012 Sandy Hook Elementary School massacre in Newtown, Connecticut, President Barack Obama issued a memorandum calling for the Department of Health and Human Services (HHS) to make gun violence a research priority.

In response, the National Institutes of Health (NIH), an agency under the HHS, said it would solicit funding proposals for “research on violence with particular focus on firearm violence.” The plan was to fund projects starting between 2014 and 2017.

“This federal initiative showed an openness and willingness to support this important research,” said Amanda Nickerson, PhD, director of the Alberti Center for Bullying Abuse Prevention at the University at Buffalo.

Nickerson is on a team that was awarded almost $4 million for a study on violence exposure and substance use.

“I doubt that we would have submitted this specific proposal without the request for application,” said Nickerson. Her grant will run until 2020. She’s unsure what will happen then.

“Federal funding for research is increasingly difficult to obtain, and for those of us who study issues related to violence, victimization, sexual orientation, and other topics that have the potential to generate political controversy, it is even more risky,” she said.

As of now, there’s no indication from the NIH as to whether there will be another round of funding made available for new projects in the future, said Hemenway.

The NIH spends billions to study not just common causes of death like cancer, but also diseases that are less fatal like West Nile virus or conditions like seasonal allergies that aren’t deadly. But they’ve dedicated just a fraction of their funds to gun violence research.

How can we get better data on guns?

To get better data, there has to be the willingness to collect it. But there also needs to be money to pay for it.

In 2015, the Congressional Budget Office estimated it would cost $12 million to expand the CDC’s NVDRS database to include all 50 states and American territories. Doing so would dramatically improve the quality of the data by making it representative of the United States as a whole.

The CDC has never allocated the money to fund this effort.

“The CDC doesn’t give money for gun research,” said Hemenway.

For researchers who are forging ahead despite the dearth of federal funds, private foundations provide a welcome alternative source of money. However, these options are still rare.

“We are living hand to mouth,” said Hemenway.

Hemenway said he’s on a team of four people and, along with another team member, has to personally raise two-thirds of his salary.

Some researchers go further: Dr. Garen Wintemute, another prominent gun-violence researcher at University of California at Davis, famously donated more than $1 million of his own money to the school’s Violence Prevention Research Program.

The other problem for the field is that the researchers who are collecting the data are heading into the end of their careers. Wintemute is in his 60s, while Hemenway is in his 70s.

If there’s no money to fund them, junior investigators like assistant professor Rinad Beidas, PhD, might not be able to study the issue at all.

Beidas, an assistant professor at the University of Pennsylvania’s Perelman School of Medicine, was awarded one of the NIH grants to study suicide prevention.

“As scientists, you have to follow where the resources are,” she said. “You’re not going to get the people like me, who are newer to the area and interested in firearm violence.”

What consequences does all this have for the U.S. public?

One way of looking at it, Hemenway said, is in comparison to injuries and deaths on the road. Deaths from motor vehicle collisions have steadily decreased over the last 30 years, despite the fact that more people drive now.

The reason, he said, is good data and research. Both have informed public health interventions, like road safety classes, graduated driving licenses, and better safety technology on the road and in cars.

Meanwhile, the rate of firearm-related deaths over the same period has remained constant. If the same logic were applied to the issue of gun violence, things might be very different.

“Every time this happens, everyone says, ‘How could this have happened? Why can’t we identify folks who are going to go on to engage in [gun violence]?’ If we don’t have the science to elucidate the answer to the questions, this is just going to keep happening over and over again,” said Beidas. “And it feels very hopeless.”