Yes, one can measure the cost of encryption like this, just like you can measure the cost of lax gun controls or automobile traffic by their fatalities (= negative costs). But you also measure the cost of encryption and lax gun controls and automobile traffic by their benefits (positive costs). If the balance is sufficiently in the positive, the negative costs might get accepted. There is no logical point, to only reason from one side of the calculation.
I never heard Barr say:
"The cost of privacy is measured in 'victims' who might have been saved from crime if law enforcement had put an officer in every bedroom throughout the country."
"The cost of lax gun controls is measured in dead pupils who might have been saved from death if legislation had passed stricter gun control laws."