Should have been more precise. By "social spending" I mean sum of (tax revenue) + (donations to charities), since 1% decrease in revenue translates to >1% increase in donations to charity. Agreed that this says nothing about "net social benefits," since it depends on the value of government vs. social spending.

My point is only that the elasticity is above 1, since if it were below 1 then there would be literally no justification for deductions. In that case, 1% reduction in tax revenue from deductions would lead to a less than 1% increase in giving, so the government could increase aggregate funding for charities by killing deductions and issuing grants. Elasticity >1 opens the door for deductions being sensible depending on objectives and use of funds by gvt vs. charities.

Note this is not a spending multiplier, so the comparison with government spending multipliers is irrelevant. The relevant comparison there would be, for example, GDP (or ideally the "social") impact of each dollar in charitable spending. I don't know what that is and it probably varies by charity.