Age and Age^2/100

[Pepito Grillo]

Hello,
I am currently doing a panel data regression using fixed effects and have recently tumbled upon some past literature on the same topic in which they use age and age^2/100 as control variables, I was wondering why is this so (I understand age can be used as a control) but especially I am confused as why they use age^2/100, also considering that the value of age^2/100 will be less than the age variable.
Many thanks

[Nick Cox]

I guess that dividing by 100 was nothing other than a way to ensure that coefficients and SEs were not too small to be easy to report.

[Pepito Grillo]

Thank you Nick
Still confused on why they would use age and age^2 though?

[daniel klein]

Including squared terms is one way of modeling a non-linear relationship; more precisely an (inverted) u-.shape.

Best
Daniel

[Nick Cox]

Without disagreeing with daniel klein I note that a turning point (maximum or minimum) does not have to be evident within the observed data range for a quadratic to be useful as catching curvature.

[Pepito Grillo]

Thank you for the help.