A question from Corporate Finance, by Berk&Demarzo

Galt Motors currently produces 500,000 electric motors a year and expects output levels to remain steady in the future. It buys a part from an outside supplier at a price of $2.50 each. The plant manager believes that it would be cheaper to make the part rather than buy it. Direct in-house production costs are estimated to be only $1.80 per the part. The necessary machinery would cost $700,000 today (year 0) and can produce output from the following year (year 1). It would be obsolete in 10 years (in year 10). This investment would be depreciated to zero for tax purposes using a 10-year straight line depreciation. The plant manager estimates that the operation would require additional working capital of $40,000 from year 0 (even though there is no production at year 0) but argues that you do not need the net working capital for the last year of operation (year 10). The expected proceeds from scrapping the machinery after 10 years are estimated to be $10,000. Galt Motors pays tax at a rate of 35% and has an opportunity cost of capital of 14%.

b) How much is the incremental cash flow that Galt Motors will incur in year 4 if they elect to manufacture the part in house is closest to:

c) How much is the incremental cash flow that Galt Motors will incur in year 10 if they elect to manufacture the part in house?

The answers for (b) and (c) are 500,000*(2.50-1.80)+0.35*70,000=374,500 and 500,000*(2.50-1.80)+0.35*70,000+40,000=414,500 each. My question is why the tax rate is not applied to the amount of CoGS(which is 500,000*(2.50-1.80)). As gross profit increases when CoGS decreases, I thought that a decrease in CoGS would lead to an increase in income tax, and the answer should have been 500,000*(2.50-1.80)*(1-0.35)+70,000*0.35. Why are the answers like that? I'd appreciate if anyone gives reply to this question.