My old room mate did cocaine studies with pregnant mice. He'd inject pregnant mice with cocaine and then raise the mouse pups and place electrodes in on of the "pleasure centers" of their brains. He would then measure the curve of how hard the mice were willing to spin a wheel for a given amount of current into their brains.
The model was that mice willing to beat their bodies more for the same amount of stimulation were more susceptible to a wide range of addictions. He found that pups exposed to cocaine in utero did in fact as adults spin the wheel harder for a given amount of stimulation, indicating higher susceptibility to a wide range of addictions.
My old room mate would have really liked to perform the same study with nicotine, since many many more human mothers dose their fetuses with nicotine as compared to cocaine. For what the wild speculations of an experienced researcher are worth, he suspected that mouse pups exposed to nicotine would also be more susceptible to addiction (supposing he was actually measuring susceptibility to addiction).
However, politicians and lobbyists have made it much easier to get federal grant money for cocaine studies vs. nicotine studies, despite nicotine having a much larger impact on society.
On a side note: in 1999, 4.7% of US 8th graders were willing to admit to having used cocaine, so the test group appears to be below US averages for cocaine use, despite their mothers using cocaine. I imagine that being predicated upon having mothers caring enough to place their children in these studies, and mothers responsible enough to stay in contact with researchers, and the subjects knowing they were being studied, skewed the drug usage portion of the study. Nationally, (for those outside of medical trials) I can't imagine the cocaine use rate for those whose mothers used cocaine to be below the rate for those whose mothers did not use cocaine.