Reply #1: Advocacy research.

[Igel]

It happens. Sometimes the researcher doesn't even know it's happening, but he winds up making claims and assertions about causality--or setting things up so that the reader makes the inferences, and finds making such inferences almost obligatory--based entirely on correlations and gaps in the data.

It's not just a social-science problem. Not sure if it's even primarily a social science problem.

The illegal immigrants I've worked with have all been very poorly educated. This leads to bad jobs, unstable jobs: If you have a 6th grade education, "career" isn't the word that generally comes to mind. Parental education is a prime factor in the kid's cognitive development. They've all had fairly traditional ways of interacting with kids, ways that don't respect a wide variety of differences and grant a lot of dignity to the kids. Often they don't have stable social networks--they don't necessarily trust compatriots, and they have no others to easily turn to. They aren't comfortable dealing with authorities: In their home countries the assumption was that the authorities had no respect for them, that social services were mostly a crock, and avoiding them was best. Add in the fear of being deported, and you wind up with a lot of authority avoidance.

But only the very last bit has *anything* to do with their status as illegal immigrants.

Legal immigrants, on the other hand, are mostly in two groups: The relatives of current residents, in which case they have a social network ready made upon arrival. Or they're allowed in because they're reasonably well educated and fill a need in the US other than doing scut work--so they have a job, and their jobs are usually more stable and bette paid. The parents are better educated, on average, and are more likely to have moderately or drastically non-traditional ways of dealing with kids, and have had enough authority not to fear authority or distrust it. In other words, it's not their immigration status that's causing their kids to be better off.

Suddenly making all 10-12 million illegal immigrants legalw wouldn't solve most of the problem. Even among the native born, English-speaking populations if you find groups with the same educational profile as most of the illegal immigrants you'll find kids that are really quite at risk.