How do you feel about the current number of trees in your city? Too many? About right? Not enough? You probably didn’t answer “Too many.”
In a survey of San Franciscans (courtesy of San Francisco’s Open Data initiative), 3.1% of of respondents said San Francisco has too many trees.
So. Who thinks there’s too many trees? Who are those 3.1%?
We looked around the dataset to try to paint a portrait of this segment of the population, who we’ll call Lumberjacks.
Are they from a certain neighborhood?
For the most part this is a city-wide phenomenon. Lumberjacks do tend to be concentrated in Districts 9, 10, and 11 (which were grouped together in the survey we’re analyzing).
But it’s not really a very large effect: 4.9% of Southeasterners are not fans of trees. That’s statistically significantly but not too meaningfully different from the 3.3% average across the city.
What else is demographically interesting about lumberjacks?
A couple things stand out. Older folks are less likely to care for trees than younger folks (though it’s unclear if this is best interpreted as being about age, the era in which one grew up, or something else).
The other notable demographic difference is by income. Lower income respondents are more likely than higher income respondents to be lumberjacks. Here’s the proportion of each income bracket that dislikes trees:
To be very clear: income is not the end of the story. Every other effect discussed in this blog post holds firm even while controlling for income level. In fact, every effect discussed in this post is still relevant even while controlling for every other effect in this post (though most of them do tend to be correlated).
Are they just cranky people generally?
Yes, sort of. Even controlling for income and age, lumberjacks are very consistently less satisfied than typical with just about everything.
The survey asked respondents to grade San Francisco’s city services on an A-B-C-D-F scale (i.e., the grading scale in U.S. schools). Across every one of 24 grades they gave the city, lumberjacks graded statistically significantly more harshly than non-lumberjacks did, typically by a third of a letter grade. For example, on average they grade the condition of the city’s pavement as a D+, while others graded it a C-.
To give you some idea of the breadth of city services lumberjacks are less happy about than non-lumberjacks, here’s a representative selection:
Labeling lumberjacks “cranky people generally ” is perhaps overly harsh. Perhaps they were just in a bad mood when they took the survey. Perhaps they’re just more demanding citizens. But “cranky” seems as reasonable an interpretation as any.
Lumberjacks tend to be older, lower income, more southeasterly, and crankier. But those are only tendencies. There were plenty of lumberjacks who didn’t fit those patterns.
So. Look to your left. Look to your right. If you’re in a room in San Francisco with hundreds of people, at least one of those people probably wishes San Francisco had fewer trees.
Thanks for coming on this journey with us. This survey dataset is a goldmine of fun analyses. Consider this Part 1 in a series.
“If trees could scream, would we be so cavalier about cutting them down? We might, if they screamed all the time for no good reason.”
Late edit #1: As one Hacker News commenter points out, some folks oppose trees because of the association between trees and gentrification. He links to a really interesting article about the phenomenon.
Late edit #2: We should have called them Treeshuggers (or maybe Treetotalers), not Lumberjacks. Props again to Hacker News commenters for the concern about “Lumberjack” and the suggestions. Late edit #3: As a gesture of respect to the lumberjack community, we’ve donated ~$25 to the Kickstarter campaign for the lumberjack-starring game Fist of Awesome.
1. Some folks may want fewer trees because they know it costs the city to maintain them, but lumberjacks really do tend to dislike greenery. For example, only 40% reported visiting a city park at least once a month in the past year, as opposed to 64% of non-lumberjacks people. They also tend to rate the quality of the city’s parks particularly poorly. Late edit: Some folks quite reasonably expressed concern that these 3.1% of respondents were actually just folks who miskeyed or mis-circled a survey response. In fact about a quarter of the survey responses were answered by phone, and those responses showed the same patterns of lumberjackiness as other results.
2. Note that the survey specifically asked about the number of trees in the respondent’s neighborhood, then the number of trees citywide. And respondents did tend to answer those questions differently. So we wouldn’t naively expect different neighborhoods to have different perceptions about the appropriate number of trees citywide, which is the question we’re examining.
3. Late edit: Though we rarely mention it, every finding here is backed by statistical significance testing, and the findings are quite robust. For example, the p-values associated with this table were below 0.00001 for every year of the survey.