Hi all, I’m building an agricultural sector model. Calibration is via positive mathematical programming. For the construction of the slope and intercept of the cost function I get the marginal values of the calibration constraint. But these are negative! Any hints why this may be the case?

Thanks!

Hello,

Marginal values are frequently negative. If we exclude the zero marginal values, I suppose marginals are negative about half the time.

Why do you find it surprising that your marginals are negative?

-Steve