No, because the relationship between incarceration and crime is non-linear, in much the same way as tax revenues and income tax rates.
In most forms of economics or other models of social systems, the key observables are equilibrium values. That means there's a turning point in their graph. The output function of firms is like that: the most profitable production level is not 0 and it's not necessarily "as much as the factory can possible produce." There's an optimal level.
This will perhaps be difficult for you to understand so I will simplify: if there is an optimal level of something, it means the relationship is non-linear and the slippery slope doesn't apply. There's an optimal level of taxation. We can debate what it is exactly, but clearly tax revenue is 0 at 0% taxation and 100% taxation (one supposes in certain special circumstances, 100% taxation could still produce revenue but as a general rule, of course not). Thus, there must be some tax level at which raising or lowering the rate would decrease revenue.