Couldn't be prouder to work with @AaronRegunberg and @David_Arkush on describing the crime that best captures Big Oil's conduct: causing or risking catastrophe. https://t.co/OmYlfD11d7
@markjrieke Great lesson in limitations of data-driven models. They are often better than nothing, but observed results are this sparse, strong causal inferences are hard. To be fair, few saw Harris or Walz doing this well or Trump campaign inflicting so much self-harm.
@catladypolitics You did the right thing, and false positives are not that unusual. To know if it really is a false positive, best to take another test 24 hrs after your first positive test.
@Groundwork@KittyRichardsDC@yasminv Thank you for bringing a refreshing dose of reality to this important discussion. The traditional austerity line is wrong on the facts and immoral in its application. I hope Biden kicks the fossilized policy hacks who've been so wrong for so long to the curb.
@emma_k_h@jenniferdoleac Congrats, Emma! VA is lucky to get you. And, I love the new piece on prosecutorial prediction mistakes. Please keep the puppy pix coming. :-)
@trippwhitbeck I love being correct, but crime rates respond to many, many things & those are hard to tease out. One reason I love the probable causation podcast is that Doleac and her guests try very hard to do that. I can still be wrong if crime rates go down!
@trippwhitbeck @brendanbooflynn @swagg__R@JoshKraushaar@mattyglesias High-volume, low-quality arrests are what clog courts. It's easy to look productive making arrests for petty offenses, so PDs do that. Redirect those resources to actually solving homicides, carjackings, etc. You can't sentence your way to higher-quality policing. Go Mammoths!
@mattyglesias It's the old adage: certainty and celerity over severity. To achieve that, move police (and thus prosecutors and judges) from the high-volume, low-quality work typically encouraged by modern PDs to investigating serious crimes. You can't fix the front end at the back end.
@maustermuhle Pay it forward. Fund intensive tutoring for kids below grade level. We know it works. Several top researchers just reviewed the evidence and advocated for just this at #SREE2022.
@maustermuhle Three quick points: (1) tickets should not be about revenue (see Ferguson), they should be about increasing compliance: (2) we don't have good data on what increases compliance; (3) let's test different approaches and collect that data.
@MusserRyan@maustermuhle Tickets are designed to provide notice of a fine and facilitate payment. They are not designed to change behavior, and we are not sure how effective they are in this context. I am glad DC is testing ways to change behavior instead of just making assumptions. This is good.
@maustermuhle This is great news. I love it when we actually test things to see what works. Everyone has an opinion. Almost no one has evidence. Let's go get some!
This is one of the many reasons @DCJusticeLab's work is so important: https://t.co/e3DgJFE6dK. Thank you @zeerobs & Stephen Rushin, and thank DC Justice Lab!
@LessCrime @Tom_Scott_Crim @JacobKaplanCrim Agree. Anyone can be factually wrong, but in the modern social science world refusing to share methods is ethically wrong. @CPPJournal should require that, upon request, authors share code & data sufficient for review & replication or issue a retraction.
@EpochTimes Except that Hogan won't share his code or data, and those who looked closer found either no effect or the opposite effect. https://t.co/pIIbS2ptY7. Hogan should release his data and code or retract his article.
Great open science practices by these folks, including the super-smart @Lab4Justice-affilliated researcher, @jj_naddeo. Grateful to these three for all the work they did helping us all assess the evidence openly and fairly!
In Charleston, SC, prosecutors wanted to ensure they were handling cases fairly & equitably. In 2020, JIL began partnering w/@SCSolicitor9 to explore data, ask hard questions, & look for better ways to support community safety. Here's a glimpse: https://t.co/1t7Tye1JjB
#justice