Sorry TM but I completely disagree with you about the length of time that the child lived with the parents. This issue is something that is hammered home in continuing education on a regular basis...and hammered home in due diligence continuing ed. I have also been through this issue with dozens of our firm's client's in paper audits over the past few years.
Yes, I reviewed the publication and it's wording, and I do find that troublesome, since it seems to contradict well established due diligence guidelines regarding the claiming of children. However, I am not going to accept it because again, it contradicts the serious due diligence guidelines in continuing ed.
Sorry, LdiJ but I disagree with you and your continuing ed once again. I can't comment on whether your understanding of the continuing ed was correct since I haven't seen the materials you got or heard what the presenters said. But I know what the statute and regulations say. The language in Publication 501 should be troubling to you because the IRS did not pick that wording by accident. It comes from the statute and regs itself. IRC § 152(c)(4)(B) provides the tie breaker for the admittedly less common situation of both parents claiming the exemption but the rule for divorced and separated parents does not apply. But that is indeed the situation we have here if the mother and child moved out in early December (and the OP's facts are not entirely clear on that). Subparagraph (B) states:
(B) More than 1 parent claiming qualifying child.--If the parents claiming any qualifying child do not file a joint return together, such child shall be treated as the qualifying child of--
(i) the parent with whom the child resided for the longest period of time during the taxable year, or
(ii) if the child resides with both parents for the same amount of time during such taxable year, the parent with the highest adjusted gross income.
The applicable regulation, Treas. Reg. § 1.152-4(a) says the same thing:
(a) In general. A taxpayer may claim a dependency deduction for a child (as defined in section 152(f)(1)) only if the child is the qualifying child of the taxpayer under section 152(c) or the qualifying relative of the taxpayer under section 152(d). Section 152(c)(4)(B) provides that a child who is claimed as a qualifying child by parents who do not file a joint return together is treated as the qualifying child of the parent with whom the child resides for a longer period of time during the taxable year or, if the child resides with both parents for an equal period of time, of the parent with the higher adjusted gross income. However, a child is treated as the qualifying child or qualifying relative of the noncustodial parent if the custodial parent releases a claim to the exemption under section 152(e) and this section.
(Italics added).
The language of the regulation and statute seems very clear to me. The choice of tie breaker here is not accidental. It puts this situation essentially on par with the outcome in the rule for divorced and separated parents where the custodial parent (the parent with whom the child lived most of the year) gets the exemption by default. So tell me, how does your continuing ed effectively write out the residence requirement out of the statute and use the AGI as the first tie breaker? Upon what guidance does it rely upon to go straight to the AGI here? Because given the statute and regulations cited above, I'm not seeing any way to get around that.
As for the audits you've seen, I cannot comment as I don't what the facts of those were to know which particular dependency rules would apply and whether the IRS examiner and taxpayer reps really understood the rule of § 152(c)(4)(B).