Answer: Let's first look at the regulation in question:
An individual shall not be considered to own an interest in an organization owned, directly or indirectly, by or for his or her spouse on any day of a taxable year of such organization, provided that each of the following conditions are satisfied with respect to such taxable year:
(A) Such individual does not, at any time during such taxable year, own directly any interest in such organization
Notice that the opening of this regulation uses the phrase "directly or indirectly." Indirect ownership is generally regarded as ownership by operation of the attribution rules. Thus, if I have an option to acquire stock which I do not in fact own, I am considered to be its owner under the attribution rules and I indirectly own the stock. From there, it can be attributed to another person, such as my wife.
Having said that, we then consider direct ownership. The direct owner is someone who owns the stock without regard to the attribution rules. Direct ownership is determined by reference to state law. The Code uses it the same way. For example, Section 1563(d)(2) states:
For purposes of determining whether a corporation is a member of a brother-sister controlled group of corporations (within the meaning of subsection (a)(2)), stock owned by a person who is an individual, estate, or trust means--
(A) stock owned directly by such person, and
(B) stock owned with the application of subsection (e) [the attribution rules].
So, when the regulation says that unless certain conditions are met, my wife is deemed to own any stock I own, whether I own it directly or indirectly, it means she is treated as owning the stock if I am either the legal owner or the owner under the attribution rules.
One of the conditions of what I call the noninvolvement exception, which is outlined in the regulation you cite and in section 1563 of the Code, is that my wife cannot own any of the stock "directly." In other words, my wife cannot legally own any of the stock, even though she may own it indirectly, such as by attribution from one of our children or from a trust.
In my opinion, community property ownership is direct ownership. In saying this, it is important to see this issue in the context of the entire Code. Before Congress allowed married couples to file joint returns, couples in community property states were able to derive substantial tax benefits by splitting their income on two separate returns. The courts held that under state law each spouse owned half the income, and so only needed to claim half the income. Spouses in other states could not do this splitting, which is one reason Congress created the joint return.
In the case of Aero Industrial Co., Inc., (1980) TC Memo 1980-116, the court determined that for purposes of section 1563 community property ownership was direct ownership. The Aero Industrial case involved a mother who owned 2/3 of one business and 100% of another. The remaining 1/3 was owned by her son-in-law. There is no attribution between the son-in-law and the mother. Because the son-in-law did not own any stock of one corporation, the rule in the U.S. Supreme Court case of Vogel Fertilizer v. Commissioner excludes him from consideration. The mother owned less than 80% of one of the corporations and so, at first glance, it appeared there was not a controlled group.
However, the son-in-law received his stock as compensation for services rendered to the company. As such, under state law, his stock was community property, even though it was in his name alone. Hence, under state law, the mother owned 67%, her daughter owned 16%, and the son-in-law owned 16%. The daughter's shares, which she owned directly (and not by attribution from her husband) were attributed to her mother. The mother was deemed to own 83% of the company, thus creating a controlled group. (Had the daughter owned her shares, indirectly, by operation of the attribution rules, attributing her ownership to her mother would have been prohibited under the rules against double family attribution.)
Here's an excerpt of the discussion of this issue in Chapter 7 of my book, Who's the Employer?:
A question arises with regard to community property. Community property ownership is direct ownership. Under the laws of each community property state, if stock is held as community property, each spouse has an equal ownership of the stock. Under certain circumstances, one spouse or the other may have the sole right to manage the stock, but that does not change the fact that both spouses own it. Generally, the title under which the stock is registered or held is irrelevant.
Medical Corporation Community Property. John married Mary as John was finishing medical school. He now has a successful incorporated medical practice. All of the stock of the practice is in John’s name. Mary is a prosperous accountant but cannot practice medicine. The couple lives in California, a community property state. Under California law, in the absence of an agreement to the contrary, the stock in John's professional corporation is community property. If John and Mary divorce, Mary will be entitled to half of the value of the practice. She is now entitled to half of the income from the practice. It does not matter that she cannot be a shareholder of a medical corporation. Under California law, she has an ownership interest equal to John’s in the practice.
This means that both spouses have a direct ownership in community property assets. Their ownership does not come through attribution. It comes by operation of state law, just as any other ownership does. This was the holding of the court in Aero Industrial Co.
Hence, if stock is held as community property, it is impossible for the stock to meet the first condition. Therefore, until there is a legal separation or a divorce, each spouse is deemed to own 100% of any stock held by either or both of them as community property. Each owns 50% directly, and 50% by spousal attribution.
Some practitioners take a different approach, believing that community property ownership should not be treated as direct ownership. These practitioners feel that Congress “clearly intended” to create a spousal exception and would not have wanted to make a difference between community and separate property states. The author disagrees. Congress knows how to put community property and separate property jurisdiction on an equal footing and did not do so. The courts have held for years that community ownership is real, direct ownership, sufficient to allow for perhaps the most important distinction between separate and community property states, income splitting even without a joint return. Some practitioners argue that we should not be concerned because the IRS has not, to our knowledge, taken this position on audit to date. Frankly, the actuarial audit cases should be sufficient warning to make us suspicious of that reasoning, especially when the Code is clear.
Medical corporation continued. Continuing the facts of the last example, suppose that Mary owns 100% of the stock of her accountancy corporation. If either corporation is held as community property, then the two are a controlled group for ordinary income tax purposes and employee benefit purposes.
There is a way around this rule for those in community property states. They can have their stock treated as separate property. Usually, this is accomplished through an agreement between the parties, signed before or after the marriage.
However, such an agreement has serious side effects. If the parties divorce, and one corporation is worth more than the other, what the parties receive in the divorce will be affected. If one spouse dies, the basis of the stock in the corporations will be different than if the stock had been held as community property (usually to the survivor’s detriment).
Truly, entering into such an agreement to clear up a pension problem is a case of the “tail wagging the dog.” Such an agreement should not be considered, even in a friendly situation, unless husband and wife are separately represented by experienced counsel. In fact, in California, such an agreement is potentially invalid after marriage unless each party has separate counsel.
Postmarital agreement. Continuing the same example, suppose John and Mary enter into a legally binding agreement that each spouse owns the stock in his or her corporation as separate property. As of the date that agreement is signed, assuming the other conditions of the noninvolvement exception are met, the two corporations are no longer in a controlled group.