Average tax rate: ATR=YT=Yp+tY=Yp+t.
For the tax system to be progressive, ATR must increase as Y increases, i.e. dYd(ATR)>0.
dYd(ATR)=−Y2p
For this derivative to be positive (ATR rising with income), we need −p>0, i.e. p<0.
Intuitively: if p is negative, it acts like a fixed tax credit/rebate (everyone owes tY but gets ∣p∣ back), which matters proportionally more for low-income earners (a bigger share of their income) than high-income earners — so their effective average rate is lower, and the average rate rises as income rises. This is exactly what defines progressivity here.
Therefore the answer is (B) p must be negative.
The trap: (A) must be positive
Students sometimes reason intuitively that 'a positive constant p, like a fixed tax amount everyone pays, sounds like it would make the system more burdensome on the poor and thus progressive' — but the actual math shows the opposite: a positive p makes the tax regressive (ATR falls with income), while it's the negative p (a rebate/credit) that produces progressivity, which is the less intuitive direction without working through the calculus.