There are actually 7 studies in that article, and a link to more, however this particular article was written because of the one study that was done, and cited, 3 times.
All of these studies have the same problems, and have lots of criticism about their methodology, particularly in how to get this data. One of the biggest critiques being that they studied kids in bad homes the CS decided to not take in, vs ones they did. This is how they know a child is in a bad home, but not being ethically responsible for them staying. This automatically selects for less severe cases being the stay at homes, and the more severe being the ones taken. Then there are this issues in my last comment, like their estimations being wide. There are also many more when I started finding when putting the titles of these studies into google scholar and adding critique.
Basically these studies aren’t particularly useful because the data is hard to get (privacy laws, parents not wanting to participate, retraction of participation agreements before conclusion of data gathering, etc), the different groupings are already selected based on a varying scales of abuse severity, that it would not be ethical to select groups in a different fashion, and any experimental trials would be unethical. These foundational problems also make meta research faulty from the start. While they can pose some interesting questions, they are not able to make reliable qualitative calls on kids being removed from abusive homes because the ability to conduct this research is just not there in a way it would need to be.
Yeah I agree, to an extent. My extended family has had a LOT of interaction with CS over the years. Some bad, some good, some just, well they seem to be completely unconcerned about how my cousin’s daughter got pregnant when she was 12. The lady told my aunt to see it as a blessing from god. Police know the guy, but just sorta shrugged their shoulders about it, even knowing his is 21. We are still fighting about this, but this is rural florida… and florida has always seemed to be worse that other states. However knowing something bad is happening, and you being able to do something about it, should definitely be a considered factor in ethics. A lot of horrific things have flown under the radar because everyone takes the “not my business” stance. My experience over all though has been CS tends to be far more likely to miss a case where a child is being badly abused, than it is to take a child, at least long term, from a household where they may have done better.
As far as we can see with conclusive data, most of the time the predictor of a child’s long term success is heavily correlated to the very early part of their lives. So when CS generally get involved that ship has often sailed. Also, like I talked about before, the two groups are selected by what CS investigators have determined to be bad enough to justify long term loss of parental rights, and those who weren’t. The only way these researchers can know that a child is in a bad home is from third party report, IE CS. So that kind of undermines the whole idea right there, and most of these people recognize that. My data is about 10 years out of date now, but when I was doing data analysis for the corrections system, our department crossed paths with data, and reports, from CYS on the regular. At that time evidence of abuse/neglect among foster homes, and group homes, had a significantly lower per capita rating than the general public. CYS taking short term custody of children always seemed to be where most of the issues arose from. More often then not there wasn’t anything significant happening, at least there was no proof of such. However taking a kid, from their family, even if it is just like over night to be reviewed by medical staff, is stressful and potentially traumatizing event. Most places would get a report, and if it didn’t have like a laundry list of reporters, or some other stronger evidence, they mostly just wanted to see the condition of the house, and ask the child about it without their parents there. Rarely was this more than voluntary. However, we did have notable hot spots where this activity was way higher than normal. That was a political matter though, and my role was done, but they do investigate cases where things seem to be out of alignment with norms.
Nothing is perfect though, and when you are dealing with tens of millions of people, you will no doubt get thousands of errors, enough errors to make it seem like this is way more common than it is. There really isn’t a good solution to this.