Yes: the questions they're forcing people to answer when they mention race in a post are not empty roadblocks that simply add friction to posting, but are also meant to sharpen the poster's thinking process by forcing them to focus on aspects other than race.

That seems like a much more reasonable and even-handed approach than just simply saying race should never be mentioned.

I think they should go one step further. Instead of requiring two or more of a given list of descriptors if race was mentioned, just require three or more of the given list in the example image (with the age, build, and race questions added) no matter what. Then everyone gets to think about what makes someone suspicious. There may be biases related to how people dress, or what hair style they have or what age they are or even what their build is that can also be reduced by this method.

Singling out the race question as one that needs to be handled delicately is the wrong way to go, in my opinion.

This is misguided. Not every poster can provide a detailed description of the person they're posting about. Maybe they didn't get a good look at the person, maybe they have a bad visual memory, or various other reasons. Requiring multiple descriptors of the person for every post would cut out a bunch of legitimate reports, and for no good reason.

Your suggestion seems to be predicated on the assumption that reducing the number of reports in general is a good thing, but there's no reason to be making that assumption. Reducing racial profiling is a good thing, but that does not imply that reducing other posts is good.

When race is mentioned maybe present a slightly grainy image of a random person of that race with a statement saying this person had been reported previously "Is this the same person? Yes/No". If confirmed announce that it was a random image "racism confirmed".

Hopefully this is a joke, but in case this needs to be pointed out, racism != an inability to distinguish members of race X.

Perhaps not, but it might be a significant component of it.

Think about it from the point of view of someone who doesn't interact with black people very often: they probably can't tell whether a person is a gangbanger or dangerous or not. If you can't distinguish a potentially dangerous black individual, then everyone black person becomes suspect, especially if you have a negative view of the group to begin with. People's natural risk aversion and fear of unfamiliarity just heightens this affect.

But presumably, that same person is more familiar with white people and can distinguish real threats, from, say, their neighbor or the kid down the block.

Nextdoor "suspicious activity" reports were notorious for racist reports that should never have been made. "Everyone be on the look out! I saw a black man walking around in our neighborhood, possibly looking for houses to rob."

If the UI helps people realize that a black man walking down the street is not suspicious activity and should not be reported as such, then it is doing its job effectively.

How exactly does that work, given Google Streetview

I really don't understand what you mean by this.

I don't know about your neighborhood, but it's almost a decade out of date here.

> Am I missing something obvious here?

The obvious thing you are missing is that "There is a <insert ethnicity> man walking around the neighborhood." without any further identifying characteristics is completely useless information as to "crime and safety" and can lead to both false positives and race-based alarmism.

If you actually see a suspicious person round the neighborhood, helpful information will be to describe how they look in more complete detail, and what is it about their appearance or behavior that causes you to think they are a crime and safety concern.

I understand and it makes sense that saying "I saw a black person walking around" is effectively useless and likely racist. The problem I see here is that it's only applied to race. If the same person made the same report and simply not added race (I'm only going by what I see in the screenshots, no clue whether it's more complex than that) then we'd have a similarly useless post because it offers no quality information either way. I don't understand why they wouldn't ask for the extra 2 out of 4 details for all "vague" reports, only for race. It's good, valuable information.

edit: as I think further on it, I probably find more questionable the whole existence of a feature to report "suspicious activity" than Nextdoor singling race out. I mean, were I to go out in a skirt (I'm a hairy dude) would I be reported for whatever suspicious activity means? If a hippie-ish friend of mine with dreadlocks came over, would that be "suspicious"? The more I think on it the more it seems like a means of fueling people's paranoia/racism. I read that they tried to address that with the crime report feature but I still get the feeling that the whole thing is misplaced.

Nextdoor isn't especially unique compared to other social media platforms. What does appear unique is its traction among highly localized user groups (i.e. neighborhoods). Both myself and friends found Nextdoor actually extremely useful in the course of recent home purchases, to get to know the target neighborhood and seek inexpensive repair/contractor services.

Given Nextdoor's very localized character, there can be unwritten context to users' posts not always present in other platforms like FB. Eg. very easy to imply "I know where you live" in a way the intended recipient may grasp, but group moderators do not. As another commenter said, having users click through some questions before committing a hot-headed posts seems wise. It would mirror a face-to-face interaction, i.e. "Did you really mean what you said?"

No idea, tho, on the justification for Nextdoor's metrics for reducing abuse.

> Both myself and friends found Nextdoor actually extremely useful in the course of recent home purchases, to get to know the target neighborhood and seek inexpensive repair/contractor services.

Did y'all use it before or after purchasing a home? I ran into Nextdoor while searching for ways to research potential homes, but ran into a verification wall that I didn't feel like bypassing. Had I jumped that hurdle it seemed like it may have been a good way to get a feel for potential neighborhoods.

Used Nextdoor after home purchase, since I only was only comfortable using their snail mail postcard verification. Nextdoor does indeed treat their location verification more seriously that other services, although that likely is warranted give the possibility (and consequences) of abuse.