WASHINGTON – The terrorist threat to the U.S. is not as dire as national security officials say, and Americans are being scared into believing they are in more danger than the facts suggest, some experts say.

“The truth is, you are 3,000 to 5,000 times more likely to be killed by an American with a gun on the street than to be killed by a jihad terrorist in the United States,” according to Peter Bergen, CNN’s national security analyst and vice president of progressive think tank New America.

But the November attacks in Paris in the Islamic State claimed responsibility for killing about 130 people and the San Bernardino shooting, in which Tashfeen Malik and husband Syed Rizwan Farooko killed 14 people, reignited fears.

Director of National Intelligence James Clapper told the Senate Armed Services Committee in February that the U.S. could expect another major terror attack next year. A few days later, CIA Director John Brennan said on “60 Minutes” when that the U.S. should be more fearful than ever because the Islamic State has access to chemical weapons.

Sen. Charles Schumer, D-N.Y., is blasting President Barack Obama for his proposed $300 million budget cut from a counterterrorism program, the Urban Areas Security Initiative, and lawmakers are fighting to eliminate the proposed cut.

Meanwhile, stories of terrorists like Nidal Hasan, who shot and killed 30 people at a Texas military base where he worked, have been translated into movies, most recently an HBO documentary series launched last month called “Homegrown,” which features cases of domestic terrorism.

Possibly as a result, 83 percent of U.S. voters, according to a 2016 Quinnipiac poll, believe that an impending major U.S. attack is “very likely.”

But Bergen and John Mueller, a national security expert at libertarian Cato Institute, believe that the government is spending too much on counterterrorism, pointing to the low number of terror attacks on U.S. soil.

The U.S. spends around $100 billion per year on deterring, disrupting and preventing domestic terrorism, with homeland security measures, up $75 billion annually since 9/11, according John Mueller. He conducted a cost-benefit analysis of the spending, noting there are about six to seven-terrorist deaths per year on average in the U.S. The analysis considered three factors: probability of a successful attack without a particular security measure, losses in a successful attack and the degree to which a security measure could prevent an attack.

Mueller found that for the $75 billion to be cost-effective, there should be 1,667 terrorist attacks per year.

“The government followed up over 10 million leads since 9/11, and maybe a thousand led to little, petty nut cases who couldn’t have done anything, but were just thinking of it,” Mueller said in a phone interview. “The intelligence community was telling reporters that they thought there was between 2,000 to 4,000 al-Qaida operatives, but that number was extremely close to zero.”

But the statistics of terrorism related deaths don’t comprise the full picture, because it’s important to consider what possible events could happen, said Henry Willis, a national security expert at global policy think tank RAND Corporation.

“Yes, six to seven people die per year on average, but this past year, 14 people at San Bernardino died. In 9/11, thousands of people died,” Willis said. “If a nuclear explosion went off, thousands of people would die again. It would be irresponsible to not try to prevent events even though they didn’t happen.”

There needs to be a more concrete examination of what works and doesn’t in homeland security, Willis said. There are very few studies of the efficacy of different counterterrorism programs to gauge which areas need to be improved, but using smart technology has been valuable in plot prevention, he said.

The government has invested in the tech sector in order to test preventive ways to detect terrorist attacks — a smart move for the money, Mueller said.

The Department of Homeland Security rounded up lab scientists and engineers in 2008 to test Future Attribute Screening Technology,  which gauges physiological responses like heart rate, breathing and temperature at travel checkpoints to target those who might be planning a crime and alerts Transportation Security Administration for secondary screening, said DHS spokesman John Verricho.

The next step for FAST, if there is a new wave of funding for the project, is to find a way to make the technology more efficient at calculating a person’s son’s intent. Experiments have successfully classified people at least 78 percent of the time, , said Verricho.

The Department of Defense and the DHS by have funded small and businesses in Silicon Valley and the CIA has bought the latest technology from-Q-Tel, a Virginia venture capital firm, Willis said.

“Behaving like a venture capital system and having proactive investment arms, instead of red tape … is productive,” Matt Gardner, the CEO of the California Technology Council, said.

But some experts suggest that the government needs to return to traditional investigative methods, such as low-cost methods of finding informants and wiretapping. The National Security Agency actually “minimally involved” in keeping the U.S. safe from terrorism, and the government’s past claims about the success of bulk surveillance are “overblown and even misleading”, according to an analysis of 225 domestic terrorism cases by Bergen.

“NSA surveillance of any kind…played an initiating role in only 7.5 percent of cases,” the report said. Traditional tools like community and family tips, informants and routine law enforcement initiated 60 percent of cases identified.

“It comes down to good old police work, informants and a community willing to report suspicious behavior,” Gary Lafree, a global terrorism researcher, said. “A lot of plots are foiled when people trust authority enough to report good information.”