I’d argue just as Patio11 argues that zero payment fraud is a suboptimal amount of payment fraud, so too is zero tax fraud. And in the case of taxes, the losses

I’d argue just as Patio11 argues that zero payment fraud is a suboptimal amount of payment fraud, so too is zero tax fraud. And in the case of taxes, the losses (movement to low tax jurisdictions, winding down of enterprises, expending less hours on labour at the margin, etc) begin to heavily outweigh the gains (in marginal income) well before the “zero tax fraud” point.