Donald Trump tried very hard to position his 2020 reelection bid as a referendum on police.

He was explicit about it. He claimed that the choice between him and his opponent, Joe Biden, came down to Americans’ immediate safety, with a Biden election meaning the country was “going to have crime like you never saw.” Over and over he talked about how much he, by contrast, loved the police. He joined other Republicans in responding to calls for reducing police funding by saying he would expand department budgets. When he received the endorsement of police groups, he touted them frequently.

Police officers are often a sought-after endorsement, but Trump went further, seeking to almost literally roll law enforcement into his campaign. At rallies he’d call out law enforcement the way he might call out a local Republican elected official, with the crowd cheering far more robustly for the police.

It may not be a coincidence, then, that police officers gave far more in reported contributions than they had in years past — and mostly to Trump and his party.

Before we look at the numbers, there are some critical caveats to mention. The first is that, for all of the rules governing political contributions, a lot of what’s raised isn’t specifically reported. Campaigns only have to reveal individual contributions that pass a certain dollar amount, meaning that if you send a check for $100 to a campaign, it may never end up in a Federal Election Commission donor report.

The second caveat is that identifying police officers or employees of police departments relies on how people describe themselves when they do have to report a contribution. Donors who give to a campaign or a political committee who pass the threshold for reporting must identify their employer and occupation, allowing us to pick out everyone who says their job is “police officer” or who works for, say, the San Jose Police Department. But a typo — “ploice” — can muck things up.

From 2004 to 2018, there was a pattern in how police officers or employees of police departments (which we’ll just refer to inaccurately as “police” from here) made political contributions. Democratic and Republican candidates or committees (meaning things like the National Republican Senatorial Committee) got about the same level of contributions from police, with Republicans getting slightly more in presidential years.

Then came 2020.

In 2016, there were about 7,000 contributions from police. In 2020, there were more than 46,000, totaling more than $2.75 million.

Across the country, the balance of contributions between the parties shifted to the right in 2020. In the previous four federal cycles, the average amount given to Republicans out of contributions that went to either identified Republican or Democratic committees was fairly even. In 2012, Republican committees got an average of 50.2 percent of contributed money across the 50 states. In 2016, the average was 57.7 percent.

Last year it was 77.7 percent.

Now it’s time for another big caveat. The bulk of those contributions in 2020 didn’t go directly to Trump, but rather to a newly formed fundraising platform called WinRed. WinRed, created to emulate a successful Democratic tool called ActBlue, didn’t exist before 2019. Before that point, small donations were collected directly by campaigns that often didn’t report contributions under the threshold. So the advent of WinRed may simply be capturing small donations that normally weren’t included in federal reports.

Of the 25,000-plus contributions from police reported by WinRed, 88 percent were under $100, well below the reporting limit. Only 66 percent of contributions from police to Trump’s campaign were under $100; for the Trump campaign’s joint fundraising committee with the Republican Party, the percentage was 63 percent.

The result is that a massive portion of the donations made by police in 2020 came through WinRed and to Republican candidates (including Trump). But we don’t know how much of the increase in 2020 was a function of WinRed’s creation.

There’s another interesting aspect to these contributions. As we explained above, we included contributions from people who identified their jobs as police or who identified their employers as police. So someone who works for a police department in a capacity that isn’t directly related to law enforcement is included here. And over the past 16 years, those employees have been less likely to give to Republicans than have those whose occupations are explicitly described as “police.”

Even with the group whose employer is “police,” though, there was an uptick in contributions to Republicans last year.

The available data are frustratingly incomplete. We can’t say with certainty what percentage of those contributions might have simply gone unreported in past years. It’s possible, in other words, that the giving was no more lopsided in favor of Republicans last year than normal, it was just documented for the first time.

It’s also very possible that a president and party that seized on criticism of police as a primary rhetorical point ended up motivating a lot more police officers to make political contributions than normal.