What's going on in Sunnyvale?

The big news on local TV today is the "59% rise in violent crime" in Sunnyvale. That caught my attention, because a) it sounds like another bogus use of statistics and b) as we've seen in Los Altos next door, crime has been generally low and not shooting up.

The news report alluded to "California DOJ" data, but failed to actually point to it. Have no fear, gentle reader, for we have trod where KTVU fears to go: straight to the data.

The DOJ does, in fact, provide this data, although it takes a little work to get to. You can download it here, if you are interested: https://openjustice.doj.ca.gov/exploration/crime-statistics 

Next you have to filter that down to only show Sunnyvale. Excel, Google Sheets, and LibreOffice all have a way of doing that, using "filters." Once you set them up, you get a little pull-down menu for each column, and you can select only the values you want. That's what I did, for each of the 5 cities shown below. Then I plotted up the data.

 

 

Before we dive in, let's review the caveats on this data. 

• These are raw counts, not adjusted for population. Los Altos is way down low in part because Los Altos is much smaller than Sunnyvale. So we can't use this to decide which town is safer. But we can use it to look at how things change over time. So focus on the trend in each line, rather than the absolute values.
• That being said, we have to be careful about trends as well, because the population of these cities has changed over time. Ideally I would go find population data by year for each city and make the corrections, but I haven't done that yet.
  
• COVID happened in there. That really messed up a lot of things with crime data, because the whole way society worked had a major hiccup. So, anything in the 2020-2022 time frame you have to take with a grain of salt.
• The way that the FBI collects this data changed, slowly, between 2017 and 2021. I think most of this data has been sufficiently corrected, but it's another thing to keep in mind. 
  

So crime is up in Sunnyvale?

First of all, and no surprise to old hands at crime data, crime was a lot worse in the 90s. However, it looks like there might be something to this claim about recent rises in crime in Sunnyvale.

I was able to reproduce the 59% number in the news report. You get that by comparing the total number of violent crimes in 2023 to the average number of violent crimes in the 10 years between 2014 and 2023. That's a little weird, because you are comparing 2023 to an average that includes 2023, but it doesn't make that much difference. Another concern is the time frame -- how did they pick 10 years? The 59% is pretty sensitive to the choice of the time window.

Still, looking at the graph, it's clear that in Sunnyvale and Santa Clara there was a big increase. Santa Clara looks like it went back down, but Sunnyvale didn't. There was a smaller effect in Mountain View. Definitely something that is worth exploring, so let's ask some other questions that might shed light on what's going on here.

What about other kinds of crime?

In that same data download there is information about property crimes as well. We can plot that up the same way:

Well, perhaps there is a slow rise from the glory days of 2010, but nothing as dramatic as we saw in the violent crime plot.  Remember, these are not corrected for population. We can also see the huge drop in 2020, when property crimes were thwarted because everyone was home all the time.

Notably, property crime is down in Sunnyvale from 2023 versus 2022.

Why didn't anyone mention this?

The "59% rise" statistic appears to come from the Sunnyvale Public Safety Officer's Association--that is, the police union. The page where they put up the statistics is advocating for one of the mayoral candidates in the recent election. So this is a political choice, and that 59% number was highlighted in support of the candidate who wants to increase the budget for the Sunnyvale Department of Public Safety (Sunnyvale's combined police/fire/EMT organization) and hire a lot more officers.

So, should I be worried or not?

Let's dig a little deeper into the violent crime stats. It turns out that the California DOJ gives you one more layer of detail, telling us about the numbers of different categories of violent crime.

 

Wow! This rise in crime is almost entirely due to an increase in aggravated assault. That's when someone attacks someone else, not in the commission of a robbery, like in a fight or domestic violence. This is a sharp increase, and it's likely that the Chief has a fair bit of insight into what is going on there.

It also says that you aren't a lot more likely to get mugged in Sunnyvale, and as we saw in the chart above, your house is not more likely to get broken into.

However, you don't get that impression by reading the KTVU article:

"From 2022 to 2023, the data shows that violent crime in Sunnyvale rose by 15.6%, which includes rapes, robbery and aggravated assaults. Back in June, police say about 20 people ran into PNG Jewelers, smashing display cases and escaping with hands full of jewelry in a matter of minutes."

As we can plainly see, this increase has little to do with rapes or robbery, and nothing to do with a smash-and-grab at a Jewelers. And while they got a statement from the City of Sunnyvale, they do not appear to have interviewed anyone representing the Department of Public Safety, or the Chief.

Bottom line: there has been a real increase in aggravated assault, but other forms of crime have been reasonably unchanged. Overall, crime is still much lower than it was in the 90s (the good old days weren't that good) and a little higher than they were in 2010--although some of that rise may be due to changes in the total population size.

Do you have any really nerdy things to add, that I should ignore unless I'm a nerd?

Well, duh. There was a change in methodology that the FBI uses to collect data. It's really hard to collect data from police departments across the entire country, in a way that is uniform and consistent. Starting in around 2017, they moved from a methodology called "summary reporting" to a methodology called "incident-based reporting." This was done to make the data more accurate, but any time you change methodologies, even for the better, it complicates your data analysis. Jeff Asher has done some good work on reconciling these two. The California DOJ site seems to be saying that it has translated older data from the summary style to the incident-based style. They probably did that pretty well. If I had a grad student, I'd ask them to dig into what those translations are, and convince themselves that nothing in that process would distort our results.

As long as I'm inventing grad students, I'd have another one dig up estimates of population for these cities for each year. That data is surprisingly hard to find, because (and this is a good thing) no one is really keeping track on yearly basis about who lives where. Census data only comes once every 10 years; yearly data, when it exists, was often estimated or extrapolated from other trends, which can be pretty error-prone. Most stuff seems to think that our population in the South Bay has been fairly stable over the last few years, perhaps decreasing a few percent. Of course, the population is actually much larger now than it was in 1990, meaning that the good old days were even worse.