4 ms·

I agree with your point. But the quoted statistic is confusing. Under what specifics would 300 people = 0.5% of the total prison population? As I understand it

I agree with your point. But the quoted statistic is confusing. Under what specifics would 300 people = 0.5% of the total prison population? As I understand it, the total prison population of the USA is over 1 million, and the subset designated 'federal prisoners' is, by itself, over 200,000.

hedora 3y ago
That 0.5% figure is wrong. It’s closer to 20%. (350,000 people) Here is a write up of many prisoner stats from 2023. See slideshow 3: https://www.prisonpolicy.org/reports/pie2023.html https://www.prisonpolicy.org/reports/pie2023.html Note that terminology in this space is very important and subtle. 7 million people are arrested each year, and > 600K are sent to prison, but the total number of people in prison (and number that are incarcerated one way or another) is much lower than those numbers suggest. The total incarcerated population is about 1.75 million.
- d0mine 3y ago
  2% of the total population arrested each year sounds like a lot! If we believe chatgpt, then the median time in prison is a couple of months. I'm unsure about the math: "The arrival rate $\lambda$ is the rate at which people arrive in the queue, and the service rate $\mu$ is the rate at which people are served and leave the queue. In a stable system, the service rate must be greater than the arrival rate, otherwise the queue will grow indefinitely. Given that the total number of people in the queue $n$ is 1750000 and the arrival rate $\lambda$ is 600000 people per year, we can find the service rate $\mu$ by dividing the total number of people in the queue by the arrival rate: $\mu = \frac{n}{\lambda} = \frac{1750000}{600000} \approx 2.92$ people per year. This means that on average, about 2.92 people are served and leave the queue every year. The median waiting time in the queue can be calculated using the formula for the median of an exponential distribution, which is $\frac{\ln(2)}{\mu}$. So, the median waiting time in the queue in years is $\frac{\ln(2)}{2.92} \approx 0.237$ years."
  - chmod775 3y ago
    > 2% of the total population arrested each year sounds like a lot! It's probably mostly people getting arrested multiple times. But yeah, there's a lot of arrests where police in other countries would have just ID'd people and sent them on their way. Arrests are simply the default operating procedure in the US, which is likely related to the fact that the lack of reliable national ID makes it harder to identify people on the spot.
  - nmfisher 3y ago
    > If we believe chatgpt We don’t. It’s usually not hard to source actual statistics (and when it is, that’s usually an indicator that ChatGPT is completely wrong).

hedora 3y ago

That 0.5% figure is wrong. It’s closer to 20%. (350,000 people) Here is a write up of many prisoner stats from 2023. See slideshow 3: https://www.prisonpolicy.org/reports/pie2023.html https://www.prisonpolicy.org/reports/pie2023.html Note that terminology in this space is very important and subtle. 7 million people are arrested each year, and > 600K are sent to prison, but the total number of people in prison (and number that are incarcerated one way or another) is much lower than those numbers suggest. The total incarcerated population is about 1.75 million.

d0mine 3y ago
2% of the total population arrested each year sounds like a lot! If we believe chatgpt, then the median time in prison is a couple of months. I'm unsure about the math: "The arrival rate $\lambda$ is the rate at which people arrive in the queue, and the service rate $\mu$ is the rate at which people are served and leave the queue. In a stable system, the service rate must be greater than the arrival rate, otherwise the queue will grow indefinitely. Given that the total number of people in the queue $n$ is 1750000 and the arrival rate $\lambda$ is 600000 people per year, we can find the service rate $\mu$ by dividing the total number of people in the queue by the arrival rate: $\mu = \frac{n}{\lambda} = \frac{1750000}{600000} \approx 2.92$ people per year. This means that on average, about 2.92 people are served and leave the queue every year. The median waiting time in the queue can be calculated using the formula for the median of an exponential distribution, which is $\frac{\ln(2)}{\mu}$. So, the median waiting time in the queue in years is $\frac{\ln(2)}{2.92} \approx 0.237$ years."
- chmod775 3y ago
  > 2% of the total population arrested each year sounds like a lot! It's probably mostly people getting arrested multiple times. But yeah, there's a lot of arrests where police in other countries would have just ID'd people and sent them on their way. Arrests are simply the default operating procedure in the US, which is likely related to the fact that the lack of reliable national ID makes it harder to identify people on the spot.
- nmfisher 3y ago
  > If we believe chatgpt We don’t. It’s usually not hard to source actual statistics (and when it is, that’s usually an indicator that ChatGPT is completely wrong).