Dubious crime statistics: a disturbing trend of police and media misleading the public

December 3, 2022

People can come up with statistics to prove anything. 14 percent of people know that.

Homer Simpson was right. Misleading statistics are the hobgoblins of small minds – a fact police, politicians, and an often-uncritical media are happy to exploit.

I first wrote about the media and misleading crime statistics in 2014. The Ottawa Citizen then reported a 28 percent decrease in December break-and-enters. The Ottawa Police’s public relations department quickly pointed out that the decline was due to increased incarceration of offenders – more bad guys behind bars, fewer break-and-enters.

At first, a 28 percent reduction in break-ins sounds impressive until you look at the sample size because it turned out that there were 23 break-and-enters in December 2013 compared to 32 in 2012 — in other words, only nine fewer break-and-enters. The limited nature of this tiny sample makes it very difficult to separate cause from effect and signal from noise.

Small sample size-driven crime statistics may make for convenient narratives and easy newspaper copy, but they do almost nothing to help make sense of the complexities of the real world. It is good that the Giants didn’t cut Willie Mays after he started his major league career 1-for-26 at the plate.

Who knows if the nine fewer break-and-enters in December 2013 were due to bad weather or skilled police work – the Ottawa Citizen never even thought to ask this basic question.

Here is the grift: Police are in a win-win situation. When crime goes down, it is because of the excellent police work, and they are spending our tax dollars well. And when crime goes up, the police need more money.

Last month the Ottawa Citizen reported on what the police described as the “disturbing trend” of increasing impaired driving-related arrests.

But a different story emerges when we take a minute to look back at the data.

The police reported 463 arrests for impaired driving as of November 2022. In 2021 there were 447 impaired arrests; in 2020, there were 503 impaired arrests; and the 5-year average for impaired driving arrests is 500.

More impaired driving arrests will have occurred in 2022 than in 2021, but a one-year increase is not a “disturbing trend.” In fact, in 2021, there was an 11 percent decrease in impaired driving arrests from 2020.

The police say Ottawa’s impaired driving rate “continues to climb” — they have not.

Any impaired driving offence is too many, but not everyone charged is found guilty. And notably, while charges from impaired driving increased in 2022, charges for refusing to provide breath samples decreased, which may explain the slight one-year increase in impaired charges. More people are giving breath samples at the roadside, so there are more impairment charges.

And, of course, we are again dealing with small sample sizes; has there been more enforcement and more post-pandemic driving? Are there other explanations for the slight one-year increase? Can we separate signal from noise?

There are all complex questions that the Ottawa Citizen, in its uncritical parroting of the police narrative, didn’t even bother to ask.

The police are not the only ones who twist to fit their agenda. It’s also a classic political strategy.

Conservative Leader Pierre Poilievre promised a solution to the epidemic of opioid overdose deaths in his “Everything feels broken” video. He then twisted and distorted health data to push for a tough-on-crime approach that would shut down safe supply programs and shutter safe consumption sites.

Poilievre claimed that there had been a “massive increase in crime” and “a 300 percent increase in drug overdose deaths” since Justin Trudeau took power. He compared this to Alberta, where a Conservative government “managed to cut overdoses in half.”

Alberta has only cut overdoses almost in half and only if you compare July 2022, where there were 92 deaths, with November 2021, where there were 174 deaths. As Paul Wells notes in his excellent debunking of Poilievre’s claims, 92 is more deaths in any month before May 2020. Overall, opioid deaths in Alberta are still up from pre-pandemic levels.

Even taking Poilievre’s data manipulation at face value, just how has Alberta achieved its modest recent success – part of the answer is through safe supply programs and government-run consumption sites.

Historically speaking, we live in one of the safest periods in history. Canada’s Crime Severity Index, a measure of the seriousness of police-reported crime, has decreased 6 per cent in the last decade and a staggering 31 per cent since 2000. Violent crime rates have also declined over the previous decade and an 11 per cent drop over the last 20 years. This fact can be lost when we focus too narrowly on monthly or yearly numbers.

But nuisance and honesty don’t make for easy political talking points.

To steal a quote, facts are stubborn things, but statistics are pliable. If everything really feels broken, police and politicians are all too happy to exploit statistics to their advantage.