groups in our sample. It is quite likely that the higher an individual's permanent income prospect the higher the cost of incarceration and consequently the less likely will be the willingness to risk the act of purchase (or consumption). To the extent that permanent income prospects at present and total expenditures are positively correlated, then we might expect the tendency to increase the consumption of marijuana with to;tal expenditures to be compensated by the higher cost of incarceration and the consequent lower willingness to pay this price in the face of such high opportunity costs.