In the study that has been referred to here, I think the authors looked at a cohort of people who graduated during one recession—probably the 1981-1982, or the 1990-1992 recession—and they tracked these people over time. They followed their earnings for the next 20 years and compared that cohort who graduated in a recession with another cohort who graduated right before a recession, and so in better economic circumstances.
They also tracked for 20 years the earnings of the lucky group who did not graduate in a recession. What they found is that about 20 years later, the cohort who had entered the labour market in a recession still had lower wages, after controlling for the same field of study—let's say comparing an engineer who had graduated in a recession with an engineer who had graduated in a good time. That's what the study did.