A regression is of linear form
y = bo + b1x1 + e
It denotes house prices (dependent) on square-feet (independent) which are graphed on Y and X axis' respectively.
Marginal effects measure a slope's gradient, or how much Y changes to X, but not the rate of change, which is where elasticity is useful and (in this example) defined as
e = (%changeiny/%changeinx) = (changeiny/changeinx)•x/y = b•(x/y), b = marginal effect.
I am having difficulty "seeing" the theoretical and algebraic reasoning for infinite elasticity when applied to independant/dependant variables. How can it be understood generally using this linear example? Is it only possible where y=0?