The hardest part turns out to be defining what you mean by police. Assuming we mean “sworn officers” (with badges, guns, and general arrest powers), we need to exclude a lot of law enforcement employees who are civilian support staff. We also have to specify whether we mean full-time, part-time, or both.
But there’s a more important question about what these police do: are they federal, state, or local? Are they “general purpose” police, or are they “special” agencies, with specialized jurisdictions or functions (like university police, or arson investigators)?
Depending on the type of cops, you can get data that is better, younger, and more detailed — or the opposite.
The Bottom Line Up Front
According to the Bureau of Justice Statistics, in general purpose police agencies — local PDs, sheriff’s offices, and state police — there were 724,000 full-time sworn officers in 2016 — down from 724,000 officers in 2013.
That’s a rate of 217 full-time police per 100,000 residents in 2016, down from 229 per 100k in 2013. That is the lowest its been since the BJS began tracking them in 1990, but there are important caveats to consider.
First, this number does not include so-called “civilian” staff: nonsworn personnel who do not carry weapons or make arrests. Counting them, total police employment went up, from 1,045,000 in 2013 to 1,050,000 in 2016. Civilian staff now make up a third of general purpose police personnel, up from a quarter in 1990.
Second, this number of sworn officers does not include officers in “special police agencies” and federal law enforcement. Those two categories accounted for almost one fifth (19%) of law enforcement officers in 2008 (a share that was gradually increasing), but they haven’t been officially counted in the last decade.
In 2008, there were about 103,000 federal agents (not counting federal prison guards) and 57,000 full-time officers in special state and local agencies. Together with general purpose police, there were 865,000 full-time sworn law enforcement officers at all levels of government in 2008.
For way more than you wanted to know about US policing and law enforcement statistics, read on.
What Law Enforcement Looks Like in the USA
Cops in the United States fall into five basic categories: local police departments, sheriff’s offices, state police, special jurisdiction police, and federal law enforcement. (There’s a wildcard of 2-3k Texas constables, which are sometimes counted as officers or not, depending on the year; I chose to exclude them.)
In 2008, the last year we have data for all categories, the breakdown looked like this:
“General Purpose” Cops
Local PDs, sheriffs, and state police/highway patrol can be lumped together into the useful category of “general purpose” police. These are the cops you think of when someone says “call the police” or “a cop pulled me over” — they have a primary geographic jurisdiction where they are generally responsible for enforcing the state’s criminal laws.
Fortunately, sheriffs, local, and state police are included in every LEMAS report (unlike special police or Texas constables, which are inconsistently reported).
In 2016, according to the BJS, there were about 701,000 full-time sworn officers in general purpose categories.
“Special” police are a category until themselves. They either have a specialized purpose (like state bureaus of investigation, which increasingly handle investigations seen as too difficult or important for smaller local agencies) or a special jurisdiction (like university, school, or park police). They still have general arrest powers, like any other sworn officer, but they have a narrower focus or range.
Unfortunately, LEMAS only has estimates for special police from 1993-2004, so the CSLLEA figure in 2008 is the latest available. In 2008, there were about 57,000 full-time sworn officers in special police agencies.
“Special” is pretty vague, but from 1996-2008, CSLLEA has a breakdown of what these police do. Overall, it shows a significant increase in special police forces starting after 2000, rising 30% through 2008.
One notable trend inside the smaller categories: state bureaus of investigation increased five-fold, from fewer than 700 full-time officers in 2000 to over 3,500 in 2008. There was no similar growth in city or county special investigative agencies.
It would be interesting to see if the nationwide trend toward greater centralization of investigations in state bureaus has continued over the last decade, but without another census, there’s no easy way to know.
Federal law enforcement is often overlooked in discussions about cops because we generally don’t see or interact with federal agents. The feds seem to have a remote connection to the kinds of daily, local crimes that concern ordinary citizens.
But however opaque its operations, and however imperceptible its effects on ordinary crime, it’s an increasingly large chunk of US law enforcement resources, and we shouldn’t ignore it.
Unfortunately, the BJS hasn’t updated its data on the number of federal law enforcement officers (FLEOs) since 2008. But the data we have sketches a useful picture of federal policing.
In 2008, the BJS found about 103,000 full-time federal law enforcement officers, not counting the Bureau of Prisons. (Federal prison guards are considered law enforcement officers for some reason, but state corrections officers are not.)
The most significant change in federal law enforcement is the increasing share devoted to immigration and customs. In 1993, this category (then the INS and Customs Service) accounted for 33% of all FLEOs, but by 2008, the new ICE/CIS/CBP accounted for nearly half (48%).
It’s now possible to update some of my old projections about what’s happening to the size of police forces in the United States.
According to the LEMAS data, general purpose police have been in decline, relative to total population, since the late 1990s, but accelerating sharply in recent years.
Of course, special and federal law enforcement were (as of 2008) becoming an increasing share of US police, so this small decline in general purpose category could be offset by more cops elsewhere.
Combining the figures for full-time officers from LEMAS, CSLLEA, and CFLEO (for the available years), shows a picture like this:
As you can see, even with the additional data on general purpose police through 2016, we are still missing a big chunk of the picture.
Counting the number of cops is not a simple exercise, but it’s important to be precise and disaggregate the figures. It’s especially hard if you want a good answer that is younger than a decade. But this data is moderately better than what we had before.
It certainly looks like the number of police per capita has been sliding for decades, ever since the hiring surge sparked by the COPS program peaked in the 1990s. But lately, probably thanks to an aging police force and waves of retirements, that trend has accelerated. It’s possible that this downward trend could be offset somewhat by growth in special and federal agencies, or by increasing reliance on civilian staff.
*This post has been updated to include figures from the 2016 LEMAS report. But we will all die of old age before the BJS completes another CSLLEA. It blew its deadlines for 2012 and 2014, then stopped to compile a new list of agencies to poll in 2016, and is now staggering on to 2018, which might be completed before the heat death of universe (but no promises).