There have been a bunch of cases on this over the past couple of decades, and I don't have a big picture opinion about them. But it does seem to me that this particular opinion is quite unpersuasive.
The curfew barred under-17-year-olds who aren't accompanied by parents from being out in public from 11 pm to 5 am (or between midnight and 5 am Fridays and Saturdays); it had some exceptions, for employment, emergencies, various events, and exercise of "fundamental rights such as freedom of speech or religion or the right of assembly ... as opposed to general social association." Rochester's chief argument was that the curfew would reduce crime by and against under-17-year-olds. But the court didn't buy it:
Although the statistics show that minors are suspects and victims in roughly 10% of violent crimes committed between curfew hours (11:00 p.m. to 5:00 a.m.), what they really highlight is that minors are far more likely to commit or be victims of crime outside curfew hours [footnote: Looking at the hourly breakdown of minors as crime suspects and victims, more than three-quarters (75% to 86%) of all crimes that minors commit and are victims of take place during non-curfew hours.] and that it is the adults, rather than the minors, who commit and are victims of the vast majority of violent crime (83.6% and 87.8% respectively) during curfew hours. The crime statistics are also organized by days of the week and despite that minors are 64% to 160% more likely to be a victim and up to 375% more likely to be a suspect of violent crimes on Saturdays and Sundays as compared to a given weekday, surprisingly, the curfew is less prohibitive on weekends. We also note that the methodology and scope of the statistics are plainly over-inclusive for purposes of studying the effectiveness of the curfew.
To be sure, minors are affected by crime during curfew hours but from the obvious disconnect between the crime statistics and the nighttime curfew, it seems that "no effort ... [was] made by the [City] to ensure that the population targeted by the ordinance represented that part of the population causing trouble or that was being victimized." If, as the dissent argues, it is enough that from 2000 to 2005 a number of juveniles were victimized at night, then the same statistics would justify, perhaps even more strongly, imposing a juvenile curfew during all hours outside of school since far more victimization occur during those hours.
The last sentence I quoted, I think, helps show the weakness of the court's analysis. Of course a juvenile curfew during all hours outside of school would likely protect minors from crime more than a late-night curfew would. But the city isn't singleminded focused solely on protecting juveniles from crime. It also wants them to have fun -- to enjoy, to a considerable extent, the liberty that adults have.
The real reason for nighttime curfews isn't that 11 pm to 5 am are the times of maximum vulnerability for under-17-year-olds, and certainly not that under-17-year-olds are the main source, or even a disproportionate source, of crime during those hours. Rather, it's that Rochester government (and quite possibly Rochester voters) think that keeping minors off the street from 11 pm to 5 am is a fairly modest restriction on their liberty and happiness, in a way that keeping them off the street from 3 pm to 8 am would not be. This also explains why letting minors stay out until midnight on Fridays and Saturdays shouldn't be at all "surprising[]"; it's an accommodation of the desire to let under-17-year-olds enjoy themselves longer on non-school nights than on school nights.
Now maybe the government has no business protecting 16-year-olds this way, especially when their parents disagree with the government's protective plans. As I mentioned, I have no firm opinions on the scope of minors' rights to liberty of movement, or on parents' rights to allow their minor children freedom of movement without interference from the government.
But the court doesn't take this view, it seems to me. It says that children's rights to freedom of movement are substantially lesser than adults' rights, and are potentially restrictable if the law passes "intermediate scrutiny," which is to say is "'substantially related' to the achievement of 'important' government interests." It seems to suggest that if the law were sufficiently linked to the prevention of crime by or against under-17-year-olds, it would be constitutional. And it rests its rejection of the law on the grounds that "Quite simply, the proof offered by the City fails to support the aims of the curfew in this case," for the reasons I mentioned above.
This, it seems to me, is a common problem in cases that use intermediate scrutiny, or "means-ends scrutiny" more broadly (such as strict scrutiny): A court focuses on one interest, and argues that the law is a poor fit with that interest. But most laws aim at serving a bunch of different interests at once. Here, the government is trying to reduce crime by and against under-17-year-olds (hence the presence of some curfew). It's trying to reduce this crime with minimum impact on full-fledged adults, who the court seems to acknowledge have broader rights than minors do (hence the limitation of the curfew to minors, even though adults commit most of the crimes). And it's trying to reduce this crime with only a modest impact on 16-year-olds' ability to have fun.
If the law is to be struck down, it shouldn't be because the city lacks sufficient "crime statistics," or because adults commit many more crimes than minors, or because children are more often victimized on Saturdays and Sundays. Intermediate scrutiny, at least as the court has applied it, seems to have been more of a distraction than a helpful guide.
The dissent also seems right in pointing out (assuming the record evidence supports it -- I haven't looked at it) that "Of course minors are more likely to commit or be victims of crime outside curfew hours. For one thing, the curfew hours comprise only 40 out of the 168 hours in a week. As to the likelihood of becoming crime victims, most children are at home during the curfew hours, as the defendant Mayor noted. But it certainly does not follow that a child who goes out at night is less likely to become the victim of a crime than one who goes out during the day. Again, it is completely unsurprising that adults commit and are victims of most crimes during curfew hours. Adults commit more crimes than children at all hours. Indeed, this may simply be an instance of the general truth that adults, who make up some three-quarters of the population, are more likely to do anything." But that too shouldn't, I think, be the heart of the argument, for the reasons I gave above.
(The court also reasons that the law unduly interferes with parental rights, chiefly because it doesn't have an exception for children who have their parents' approval to be out. But under the court's view of parental rights, it seems that if the protecting-children rationale were accepted as a justification for restricting the children's rights, it would likely be accepted as a justification for restricting the parents' rights, which aren't that strongly protected in any event. That's why I keep the already longish post above focused on the children's rights question.)