Legalized judicial racism…

I'll address you earlier response later. A lot to go through.

Re: above. Statistics are the measurement tools of science, though statistics on their own are simply statistics. Can easily be abused, no doubt. Same time, you can't have science without statistics because science by definition is knowledge gained through methods of measurement and testing. and those methods and measurements are based on statistics. Likewise w/physical or scientific laws such as Newton's. They're statistics as well, becoming laws through repeated observations.

In important ways, I agree w/you about polling, at least on the surface, e.g. poop in poop out etc, potential for bias/thumb on the scale etc. And as such, I don't discount the existence of many potential problems.

I'd flip your argument around and suggest that you are basically stating hypotheses to be tested, not inferences, because it's too broad-brushed and argument without specific examples. No doubt that there are, e.g. fatally flawed/biased polls, but that doesn't make all polls fatally flawed or biased.

 

I have two published studies on judicial sentencing and bias (many other published studies on bias in other parts of the cj system). In one study, we had access to all the judicial files including pre-sentence investigation reports. And we had this access because it was a federal judge on the east coast who commissioned the study and hired me to complete it. Took four of us (my grad asst & I, and two clerks) nearly a year just to create the data set.

The other study where I was second author of three was led by a senior colleague who was hired by a midwest state to collect on all sentences across the state. By the time I came aboard he had already compiled an enormous data set which actually took considerable funding and labor to put it together.

Thing is, even if, for example, we don't have 'all the possible data' (such as what you mention above about "attitudes"), it's a ridiculously tall order to try to argue that bias didn't exist when the numbers were strong enough to confidently reject the null hypothesis.

Absolutely conclusive? No, yet too robust against arguments that no bias existed.

Also have a massive 10-year data set collected by my wife who was a senior researcher at our county probation department. More than 15 thousand cases and basically every bit of data one can have from soup to nuts and in between...judicial decisions, psi, and probation decisions etc. We're not allowed to publish it because probation doesn't want it out there--I can understand way. Paints a not nice picture, but one that cannot be dismissed out of hand as somehow being biased or non-rigorous.

That said, I know from first hand experience how hard it is to pry information from courts. But it's not impossible and I'm not even close to the only researcher who has done it.

 

It would probably fall under "contextual factors" in your list, but one thing that came to mind was urban vs. rural settings.

Edit: You also have it as "court location" in your last item.

 

Urban/Rural is also among those factors (yes, court location).

Interesting thing about it too is that rural crime is likely undercounted relative to urban crime. Nature of the beast with less of a chance of getting caught, less crimes on the record etc...

 

Yes, the judge volunteered, but not out of courtesy. It was his idea and he invited me to do the research. It was out of a concern about sentencing bias (his own and the court) so he wanted to better understand what was happening. He was extremely gracious despite probably not necessarily enjoying the findings. Still, I received a nice formal thank you letter from him that I framed. BTW, it was quite progressive for this judge who wasn't necessarily known for it.

True, they are often exempt from FOIA. Not suggesting that it's easy to get a lot of data, but it happens, and quite often. Still, all data sets have limitations and flaws. Nature of the beast.

But you know what else is flawed: perceptions and anecdotes. People are often unaware of their biases and/or misperceive things across time because, e.g. things like singular or unique cases can bias those perceptions.

Think about it for a moment in the context of eyewitness accounts in criminal cases and how/why they've come under attack. Humans often misperceive/misremember things and because they are often easily manipulated by those with authority.

I don't doubt at all that many things aren't captured (again, nature of the beast). Same time, there is no substantive argument to be had that uncaptured information would alter the results in one study let alone thousands.

Honestly, yours/your wife's experience doesn't surprise me at all. Though it def reaffirms how the system operates.

You raise an excellent point re: ignorance & bias. They often go hand in hand, one following the other.

Thing is, bias as measured at the aggregate and/or system level is based on sentencing outcomes, controlling for variables mentioned above. In a sense, it is an objectively measured bias...which may or may not reflect ignorance.

Again, not trying to claim absolutes--but the literature consistently points in one direction, which I don't believe can be easily dismissed.

Thanks! Likewise.

Completely agree about bias being in every profession even in research. Humans are bias, and to the individual this often means some/many biases in which they are unaware.

Repeated over time and across individuals in the aggregate (in the cj system) and it's not hard to imagine or demonstrate how this produces problems that undermine ideals of justice/due process/fair treatment.

Gotta disagree in important ways about bias against every definable group. There are significant differences (i.e. variation) whether we're talking rates or in the categorical nature of the type of bias itself.

For example, the war on drugs has produced a mind-numbing explosion of drug arrests that began in the 1980s. This applied to people of all racial/ethnic groups/walks of life. But not by any stretch has this bias been equal across groups. The stark differences against blacks, e.g. is impossible to ignore or explain away despite the dearth of evidence that blacks are more likely to use drugs, and certainly not at use/sale rate that would comport with the bias in drug arrests.

Long and short, not all bias is the same/equal.

 

Flawed, maybe, but not de facto fatally. Doesn't just apply to the cj system/courts. Literally no study in any scientific pursuit ever can capture all the data about everything. It's an impossibility. Setting the standard at such an extreme otherwise: worthless science is unreasonable and wrong on many levels because it misunderstands knowledge and science. After all, all of your knowledge, all of mine, anyone/everyone is based on incomplete information and as such, if we're held to this impossible standard, there is faiap nothing that anyone can ever conclude about anything.

Yes, laws vary from state to state and within the state itself. Generally if not specifically agree with your characterization except up to the point where your suggestion seems to be that this makes it impossible to study.

But science is piecemeal and varies in what is being compared, how it is being compared etc. I need not be concerned, e.g. with CA's laws or sentencing outcomes based on said laws (other than generally informing understanding) if I am studying sentencing outcomes in NY. Further research or comparative research between different locales/systems is helpful toward understanding similarities and differences, but research is often narrowly tailored (by necessary design), as such, there is no expectation to answer all of the questions everywhere for all time.

Disagree about being "too fact specific." Though differences in all sorts contribute to making research a challenging endeavor and by extension, limits what can be concluded in any one study.

Nature of science is in probability and demonstrating via controlled study that findings are "real."

I get where you're coming from and respect your experience. It matters. But yeah, we don't agree on several things.

 

I'd have to disagree. Mens rea is a legal construction and element defining what is criminal. Bias does not have to have intent to be bias in the same way a crime requires intent to be a crime.

The system as it were is a human system (though often not a humane one) that reflects the biases, flaws, and limitations of such a human endeavor. The system also reflects biases that are built into the law that governs the system. Systemic bias is thus determined from patterns of bias that emerge from the aggregation of individual decisions which encompasses and thus reflects these forces/influences.

Clinton might not have personally intended to incarcerate his base, but that is what happened. This reflects biases as well as imo a profound misunderstanding about drugs, addiction, and human behavior.

 

/Bias

 

There are unique qualities to the cj sytem for sure. Same time, humans are not necessarily unique in the cj system compared to other systems (e.g. health care, politics, education) because humans embody human traits , tendencies and foibles. Like I said, it's a challenge to study the cj system but not impossible. It's not so mysterious or unknowable, imo. Indeed, from my own experience, it is a knowable & understandable microcosm of larger society.

 

I am trying to read and digest this conversation and I find it interesting and ..dense.
However, more than anything I want to applaud both of you for the civility, maturity and even kindness extended to one another through out this discussion. Sadly those qualities are seldom present in a disagreement/argument/discussion/ debate anymore.. especially on a message board.
Sincerely appreciate it.

 

Bias comes in many forms and you need not have intent to have or be biased. Often is the case people are unaware of their biases or are so habituated into thinking in biased ways that they are blind to it.

Agree on both counts. The latter being a strength of scientific research, not a weakness.