PhDs in gloomy times

If you are nowadays a PhD student in Economics you will be glad to know that according to this paper from Michael Boehm and Martin Watzinger, you are probably going to be a more productive researcher (in terms of academic publications) than if you were graduating in a glorious economic period:

“Economists starting or graduating from their PhD in a recession are significantly more productive over the long term than economists starting or graduating in a boom.”

Why?

“It is well documented that graduates enter different occupations in recessions than in booms.” In recessions they move to more stable industries like academics. Therefore, “the quality of talent in a relatively more stable industry [like academics] increases in recession”.

So, this increase on PhD’s productivity is coming from a shift on talented people career decision, from business sectors to academic sectors. Thus, this is a positive consequence of economic recessions, when “talented individuals choose to work in entrepreneurship or research instead of rent-seeking sectors, this may have some social benefits that subtract from the immense adverse effects of downturns”

The data analysis

In the first column, the outcome variable qi,t is the average publication output of a cohort of graduates from university i in year t. In the second column, it is the average propensity to decide in favor of an academic career after the PhD, and in the third column, qi,t is the average productivity of those who stayed in academia after the PhD. Application and Graduation rows takes into account the date of the measurement. Application means that the measure was made when the students applied for the PhD and Graduation when graduated.

Unemployment change, both at time of application and at graduation, has a significantly positive effect on research productivity at the five and one percent level, respectively.

There's something unclear about it. If people graduated in gloomy times are more produtive than those from normal times why are their salaries lower in the long-term? As Oyer said