Tax revenues typically rise along with GDP. Governments may typically spend some of this additional revenue on discretionary items in the budget. Assuming this is true, how does this complicate the work of analysts trying to determine the multiplier effects of discretionary spending on GDP?