Program participants sometimes OVERestimate disincentives; good new book

I commonly hear it claimed that health insurance assistance does not discourage work because the rules are too complicated for people to understand that work does not pay. This claim is full of errors of logic and fact (see here and scroll to Q8). But let's look at one: why should we assume that people underestimate work disincentives rather than overestimate them?

Here's one story in which one man, Dave Campbell, decided to reduce his work hours in order to reduce his income in order to qualify for Medicaid, when it now appears that he probably would have qualified without reducing income.

... when Dave returned to work a few months after the accident, he pared his hours down to meet the $2,100 [monthly] level: why work more when it would all go to Medi-Cal [California's Medicaid program]...

Dave had been getting advice from a social worker and Medi-Cal expert, as well as his very smart Harvard/MIT professor/social-policy expert sister. Nevertheless, it appears that he/they overestimated the program's disincentives (as the disincentives applied to him).

I have not yet read Professor Campbell's new book that contains the full story. I expect that it is very good, because I did read her earlier book and found it to be excellent!