Direct effects of money on aggregate demand: another look at the evidence