A Crime Wave in Sunnyvale? Think again.

This is an addendum to Part 2 of my endorsement series, where I endorsed Councilmember Gustav Larsson over his opponent, Parks Commissioner Henry Alexander III. Normally I would just leave it at that, but for some reason this past week Henry decided to tag me on Facebook about crime in Sunnyvale, and, to quote Oscar Wilde, I can resist everything except temptation.

Some background: This past September, Sunnyvale’s Public Safety Officer Association issued a letter saying they were concerned with the pace of corporate development in Sunnyvale, especially Google’s proposed campus in North Bayshore. Henry Alexander III and Josh Grossman both pounced on this letter as evidence that Sunnyvale is growing too fast and that we need to hit the brakes. Of course, this “sky is falling” narrative is somewhat undermined by the fact that the PSOA has endorsed Gustav Larsson and Glenn Hendricks.

Henry in particular made quite a bit of hay about this on Facebook, using it to preach his anti-density message. Read the whole discussion — it’s really something.

Anyway, I waded in with some other pro-housing advocates, because, well.

me IRL

Among other things, Alexander claimed that “Sunnyvale has reached critical mass,” asked me “What are you and your friends willing to do to stay in Sunnyvale?” when I challenged him over rising rents, and appeared to suggest that he’d be better suited for office because his opponent is unmarried:

Being married over 21 years and raising our 3 kids in Sunnyvale perhaps gives me a different optic than my opponent.

A great deal of the conversation also turned on whether Sunnyvale was the “safest city in America” (as it’s been ranked for three years running by… SmartAsset.com? Sure, whatever). Most of us, myself included, thought it was fairly absurd to make crime an issue here.

Anyway, as these things do, the conversation fizzled. I sparred with Henry and some of his supporters a bit and moved on with my life.

Until last Thursday night, that is, when I got a notification that I had been tagged on a post by Henry Alexander III, in that old thread.

Thread necromancy is generally considered to be unholy.

He linked some stats released by the Department of Public Safety and asked me if I had any thoughts as to why crime in Sunnyvale has increased over the past year in particular, with an 11% overall increase since 2013. Well, that sounds bad, doesn’t it? So I decided to do what I do, and dig a bit into the numbers.

After a rather tedious amount of data entry and formatting work, I was finally able to get a spreadsheet together (linked). 2017 saw a violent crime rate of 1.22‰ and a property crime rate of 18.06‰, up from 1.07‰ and 14.47‰, respectively, in 2016. Those are increases of 14% and 25% in a year, respectively.

Whenever one’s examining behavior in large populations over time, there is a certain degree of fluctuation. This is true of everything from baby names to standardized test scores to, you guessed it, crime rates. So I did the first thing that anyone who’s survived a lab course (or five) does instinctively whenever they’re presented with a series of measurements: compute the average and standard deviation. From 2008–2016, the average violent crime rate in Sunnyvale was 1.14 ± 0.11‰, and the average property crime rate was 16.86 ± 2.11‰. (That little ‰, by the way, means per mille, or per thousand. It’s a lot shorter than writing out “per thousand residents” all the time).

Then I did the second thing we always do: I made a graph. The right axis measures violent crime while the left measures property crime. The colored bands represent the average range for each metric — that is, its 2008–2016 average ± one standard deviation.

Sunnyvale Crime rate, 2008–2017

As a rule of thumb, when we’re looking at datasets, fluctuations within a single standard deviation of the mean aren’t significant. Furthermore, 2012 and 2014 saw crime spikes of similar magnitude to 2017’s. So I, sweet innocent that I am, posted the graph and said that, while I couldn’t speak to causality, without the 2018 data, 2017’s crime increase does not appear to be statistically significant.

Henry obviously wasn’t satisfied with my response. He replied:

Do you still believe that there is no correlation of increased population growth with increased crime? If your answer is a firm no, then I’ll leave it there. I just want for our city to be careful when considering the effect of increases in density.

Ah, of course, that’s what he was getting at. Growth causes crime. More people, more problems. I’m not sure where that “still” comes from — as far as I can remember I don’t think I’d ever previously discussed crime rates with Henry. Rather than just responding “LOL no” (dear reader, I was sorely tempted), I went back to my spreadsheet and started tweaking.

In what can only be described as an act of infinite kindness, DPS included population numbers in their report for the years in question. So I added them into my spreadsheet and made another graph: a set of simple sparklines showing violent crime rate, property crime rate, and population.

2008–2017. Which of these things is not like the others?

What matters here is the shape of the graphs. Population growth in Sunnyvale has been mostly constant at around 1% per year over the past ten years, with some signs of leveling off recently. Crime, on the other hand? Near as I could tell crime rates were just fluctuating randomly, mostly within the confines of their standard deviations.

So, not only do I believe there’s no correlation between increased population and increased crime rates, the numbers Henry himself linked me show that there’s no correlation. Sunnyvale’s nearly continuous growth, year on year on year, cannot be responsible for the ups and downs of our crime rates.

Fear and Loathing in Sunnyvale

At this point, it’s very clear what Henry’s game is: he’s fearmongering. He’s cherry-picking crime statistics and presenting them out of context to suggest that Sunnyvale is somehow in the grip of a terrible crime wave and that it’s the current council’s fault for allowing population growth.

Unsurprisingly, Henry hasn’t proven to be particularly receptive to contradictory evidence. His last reply in that thread?

I suppose I’ll stop presenting those facts when the term “Sunnyvale is the safest City” becomes accurate. Until then I still believe that Sunnyvale should be aware. But thanks for letting me know where you stand.

No Henry, thank you for letting me know where you stand. It doesn’t matter to Henry that the numbers — the facts — don’t bear his story out. It doesn’t matter that they completely disprove his narrative. Rather than reevaluating, he’s doubled down again and again — by tagging me in the first place, by asking me explicitly whether I thought population growth was correlated with increased crime, and then by ignoring the actual evidence I provided while reiterating his “facts”.

as always, dril was there first

The thing is, it’s not that Henry doesn’t understand the numbers. It’s that he just doesn’t care. Contrary to what he claims, he’s not presenting facts; he’s selling feelings. Henry’s entire electoral strategy, more so even than Josh Grossman’s, is based around appealing to longtime residents who are frightened and upset by the change they’re seeing in Sunnyvale, and whose memories of the Sunnyvale of the past run distinctly toward the nostalgic. Henry is playing on those fears to conjure an image of a city being overrun with dastardly newcomers, bringing traffic and crime to a formerly Elysian paradise.

If all of this sounds familiar, that’s because it is. Fearmongering was the core of Donald Trump’s campaign, as laid out in his 2016 convention speech and repeated ad nauseam since he rode down that gold escalator in 2015. Fear of crime played at least as important a role in his campaign as fear of terrorism. Combined with an appeal to a mythologized golden age, that rhetorical strategy was — barely — enough to edge Donald Trump into the White House. His administration has continued to make fear of immigrants and newcomers the centerpiece of its political strategy.

Henry clearly hopes to repeat that trick in Sunnyvale.

This Election Day, let’s prove him wrong.


Crime Rates vs. Total Crimes

It’s worth discussing briefly why everything I’ve looked at so far involves crime rates, per population, not the total number of crimes. The short answer is that, if some more or less constant fraction of the population is criminally inclined, obviously the total amount of crime will grow with the population. Since the population has been increasing, if the crime rate is remaining more or less constant that naturally implies that the total number of crimes must be increasing as well. However, as long as the crime rate does not increase, no one is at higher risk of being the victim of a crime. So while it’s obviously not great to see total crimes increase, it is tolerable so long as it remains proportional to population. That’s why the crime rate is the usual safety metric, rather than absolute number of crimes.

Regional Trends?

One thing I did not consider in my discussion with Henry was regional crime trends. The County of Santa Clara’s report on crime in 2017, however, goes into some detail on this. That report shows that the spike in crime that Sunnyvale saw in 2017 was part of a regional increase that affected our neighbors.

In each case, Sunnyvale’s numbers are more or less in line with those from Mountain View, Palo Alto, and the city of Santa Clara. Given that all four cities are wealthy and that it is easy to move between them, this makes a certain degree of intuitive sense. For the most part, criminals are not going to carefully choose which city to commit their crimes in. They’ll do so based on opportunity and where they happen to be at a given moment.

One major source of the spike in property crime in 2017, both regionally and in Sunnyvale in particular, was from larceny and theft. The total number of larcenies in Sunnyvale jumped 39% in 2017, from 1,347 to 1,873. Why? Well, it turns out larceny is the category that vehicle break-ins fall under. It appears that this spike was the work of well-organized gangs. One that was broken up in January 2018 was shipping stolen laptops and tablets to Vietnam by the thousand.

(CW: discussion of rape and sexual assault)

More disturbing is the countywide increase in reports of rape. A February article from San Jose Inside examined potential causes for this spike. To summarize, it seems that the increase is at least partially the result of increased reporting. Better resources for victims and the #MeToo movement may have encouraged more victims to come forward than did so previously.

The trial of Brock Turner for sexual assault and subsequent recall of Judge Aaron Persky may also have played a role in increasing reporting. Furthermore, in 2013 the FBI considerably broadened the definition of “forcible rape”, which previously had been limited to “the carnal knowledge of a female forcibly and against her will”; they estimated that the old definition excluded as many as 40% of rapes. Which makes sense, considering that male victims were entirely excluded.

Given that, it seems likely that the past few years’ increase in rapes in Sunnyvale (from 14 in 2014 to 30 in 2017) can substantially be explained by increased reporting rather than increased frequency.