Inequality constraints in genetic algorithm toolbox

I am working on an optimization problem using genetic algorithm toolbox. I would like to know if there is a difference in the way GA toolbox handles the equality and inequality constraints. The reason I ask this is that, when I provide the nonlinear equality constraints as inequalities (like it is done in mixed integer programming), I can see performance improvements. I am curious to know why there is a performance improvement by including the nonlinear constraints as inequalities instead of equalities.

Thanks, Bala