America’s police forces are in the spotlight. After the police shooting deaths of Michael Brown in Ferguson, Missouri, and 12-year-old Tamir Rice in Cleveland, as well as this past week’s decision by a grand jury not to indict the officer caught on video choking New York resident Eric Garner, who later died, Americans from the White House to the streets are debating or protesting police militarization, body cameras, lethal force — and whether enough is done to hold bad cops accountable.

It’s essential that we base these discussions on good data and sound presumptions. Police officers are human and fallible, just like the rest of us. How they behave and react in the aggregate is a product of the policies, procedures and guidelines set by police leadership, elected officials and ultimately the public. Here are five common misconceptions about policing today:

1 The job of a police officer is increasingly dangerous.

According to FBI statistics, 27 police officers were feloniously killed in 2013, the lowest raw number in more than 50 years. (The previous low was 41 in 2008.) If we go by officer homicides as a percentage of active-duty police, it was probably the safest year in a century. The number of cops killed on duty has been falling since the mid-1990s, consistent with the overall drop in violent crime in America. Assaults against police officers have been in decline as well.

We will probably see news stories in the coming weeks about a sharp increase in cops killed this year vs. 2013. Approximating from data from the National Law Enforcement Officers Memorial Fund, it is likely that about 50 police officers will be killed this year. That’s certainly a sharp increase over the 27 last year, but even if that toll is reached, it would still be one of the lowest since the early 1960s and in line with the general decline since the mid-1990s. The average number of cops feloniously killed per year over the past decade: 51.1.

2 YouTube videos and cellphone footage prove today’s cops are out of control.

Most criminologists believe that today’s police departments are more professional than ever before. Cops tend to get more training, and departments are guided by defined rules and procedures than they did in the past. Most decent-size police agencies have internal affairs departments, and a growing number of cities have installed citizen review boards.

That hardly means there are no problems in policing today, of course, or that these developments suffice to safeguard civil liberties. But it’s likely that the ubiquity of cellphone cameras and the diffusive power of social media are simply making us more aware of rule-breaking cops, rather than showing that there are more of them than before.

But even if there may be fewer rogue cops who abuse their authority and use force outside the bounds of department rules, it’s also true that, as a matter of policy, police use more force today than they have in the past. SWAT tactics, for example, are increasingly used for credit card fraud and other low-level offenses, administrative warrants, or even regulatory enforcement. Use-of-force training today puts less emphasis on conflict resolution and de-escalation, if they are addressed at all. The problem isn’t cops breaking the rules; the rules themselves are the problem.

3 With more criminals wielding heavy-duty weapons, police must militarize.

Multiple studies, including from the Justice Department, have shown that the guns used in homicides, including the killing of police officers, overwhelmingly tend to be small-caliber handguns. Moreover, gun ownership has increased over the past 20 years — the same period in which both the violent crime rate and the killing of police officers have been in decline.

One version of this argument advanced recently by Vox and the New Republic is that we can’t demilitarize the police without gun control. But even if it were true that criminals were arming themselves with bigger guns, it isn’t clear that gun control would demilitarize the police. First, gun-control legislation would probably not do much to keep guns out of the hands of violent criminals, particularly in the short term. Second, the argument assumes that the law enforcement community would accept such a bargain. That seems unlikely. Polls consistently show that large majorities of police officers oppose gun control, although big-city chiefs and the heads of some big police organizations support such policies. The National Rifle Association in particular includes a lot of cops in its membership and recently ran an article in favor of police militarization in its flagship magazine.

New gun-control laws may have other merits, but it’s unlikely that they would slow down the militarization of U.S. police.

4 Aggressive, confrontational policing is the best way to control crime.

Proponents of police militarization sometimes point out that the trend has occurred at the same time that crime has dropped dramatically — therefore, militarization must be working. But criminologists are still debating what has caused the decline in crime since the mid-1990s. In New York, crime fell without mass incarceration. In San Diego, it dropped without the “broken windows” policing employed in New York.

Moreover, the most notable manifestation of militaristic policing is the SWAT team. According to Eastern Kentucky University criminologist Peter Kraska, the number of annual SWAT deployments in the United States jumped more than 1,500 percent between the early 1980s and 2000. Yet according to Kraska’s data and a study this year from the American Civil Liberties Union, 60 percent to 80 percent of SWAT raids are to enforce warrants for drug crimes — and drug crimes are the one class of crime that hasn’t dropped since the 1990s.

The good news is that in places where it’s been tried, “community policing” — which stresses de-escalation, community involvement and solutions that don’t always involve more arrests, more raids and more street sweeps — has succeeded.

It happened in the early 1970s in Washington, D.C., where crime fell under the leadership of Police Chief Jerry Wilson, a community-policing advocate, while it increased just about everywhere else. In California, by the time Police Chief Joseph McNamara retired in 1991, he had used community policing to make San Jose the safest big city in America — with a police force that per capita was one of the smallest in the country. More recently, as the Washington Post pointed out this week, the number of stop-and-frisks in New York City has dropped by an incredible 94 percent since 2011 — with no noticeable effect on the crime rate.

5 Tasers and other “less lethal” weapons allow cops to use less force.

New technology and new weapons are only as good as the policies guiding their use. Tasers were initially touted as a substitute for lethal force, a way for cops to subdue violent suspects without killing them. Over time, however, they have become a compliance tool — used to quell dissent, move nonviolent protesters and punish people for talking back. A 2011 National Institute of Justice study found that cops use their Tasers too often and in inappropriate circumstances.

While there is no national data on Taser use, a 2012 Chicago Tribune report found that Taser use by suburban police doubled between 2008 and 2011. A 2011 New York Civil Liberties Union study found that nearly 60 percent of police Taser incidents in that state did not meet expert-recommended criteria for using the weapon. It’s also worth noting that Amnesty International has documented more than 500 cases in which a suspect died after being shocked with a Taser.

It would be one thing if all those stun gun barbs were being fired in place of bullets. But according to FBI statistics, the number of justifiable homicides by police has been increasing since about 2000. That data is also incomplete, but to the extent that it’s flawed, it probably undercounts such incidents.

Randy Balko, author of “Rise of the Warrior Cop: The Militarization of America’s Police Forces,” writes The Watch for The Washington Post, focusing on civil liberties and criminal justice system.