Minor overtime deduction test changes caused by taxcalc PR 3016

[martinholmer]

These changes should have been included in PR #511.
No change in national TMD data files.

[donboyd5]

@martinholmer, I updated tmd to upstream master, ensured it is using taxcalc 6.5.3, pulled pr 514, ran make clean and make data (all tests passed), then ran test_imputed_variables.py verbose. Results are below. Everything passes. The 3 values updated in this PR are hit exactly; 3 other values are within tolerance but not exact.

ded	stat	actual	expected	diff
OTM	totben	23.68	23.68	0
OTM	affpct	8.56	8.56	0
OTM	affben	1413	1413	0
TIP	totben	7.08	7.08	0
TIP	affpct	2.59	2.59	0
TIP	affben	1397	1397	0
ALI	totben	7.88	7.87	+0.01
ALI	affpct	12.28	12.28	0
ALI	affben	328	327	+1
ALL	totben	58.92	58.92	0
ALL	affpct	28.25	28.24	+0.01
ALL	affben	1065	1065	0

[martinholmer]

@donboyd5, Thanks for checking the tests on your computer using PR #514.
Do you want to do the merge now?

[donboyd5]

@donboyd5, Thanks for checking the tests on your computer using PR #514. Do you want to do the merge now?

Gladly, @martinholmer. Just merged, and deleted branch.