The IRS recently issued Private Letter Ruling 201811009, which provides helpful insight into how the IRS construes the term “control” for purposes of determining whether two affiliated 501(c)(3) organizations are “related” for purposes of the definition of “refunding issue.”
The ruling involved a 501(c)(3) university (“Seller”) that sold its medical center to another 501(c)(3) organization (“Buyer”). The Buyer was operationally independent of the Seller, but the Seller could appoint 30% of the Buyer’s board and the Seller also had approval rights over certain major Buyer decisions, such as major transactions and changes to the mission of the Buyer. If the Buyer and the Seller were treated as related, the proposed bonds (“Proposed Bonds”) to be issued for the Buyer to finance the purchase of the Seller’s medical center would be treated as refunding bonds, and they therefore could not qualify as tax-exempt bonds. This was because the Seller had previously used part of the proceeds of prior bonds to finance the medical center, and those bonds had previously been advance refunded. As readers of this blog know, post-1985 qualified 501(c)(3) bonds could be advance refunded once and only once until tax-exempt advance refundings were repealed last year. Read below to see how the IRS tackled the analysis of control, which is still relevant even though tax-exempt advance refundings aren’t permitted anymore.
If new tax-exempt bonds that are issued for the benefit of one 501(c)(3) organization are used to pay debt service on old tax-exempt bonds issued for the benefit of a different 501(c)(3) organization, then the new tax-exempt bonds will not be considered a “refunding issue” unless the two 501(c)(3) organizations are “related parties.” Under the Treasury Regulations, two 501(c)(3) organizations are related if they are members of the same “controlled group.” The ruling explores this concept.
Before looking at the “control” test for 501(c)(3)’s that was the heart of the ruling, a quick summary of some related tax-exempt bond rules may help set the context. The term “refunding issue” generally refers to “an issue of obligations the proceeds of which are used to pay principal, interest, or redemption price on another issue.” Treas. Reg. § 1.150-1(d)(1). Because this definition looks to the use of the proceeds, an issue can be a refunding issue even if the obligors are different parties. However, an issue generally cannot be a refunding issue if the obligors are not related. Treas. Reg. § 1.150-1(d)(2)(ii)(A). If one 501(c)(3) organization controls another, the two will be considered to be related parties. See definition of “related party” at Treas. Reg. § 1.150-1(b). If the Seller controlled the Buyer (or vice versa), the two would be related parties and the Proposed Bonds would be a refunding issue.
This puts the focus on what “control” means. The test for control is found at Treas. Reg. § 1.150-1(e), and generally provides that control is determined based on all facts and circumstances. The control test also provides a presumption that an entity or group of entities (the controlling entity) will generally control another entity or group of entities (the controlled entity) if the controlling entity either (1) has the right to approve and remove without cause a controlling portion of the governing body of the controlled entity or (2) has the right or power to require the use of funds or assets of the controlled entity for any purpose of the controlling entity.
With respect to the Buyer’s board composition, the IRS noted that the Buyer had an 11-member board of directors, and that the bylaws provided for a minimum of 11 directors and a maximum of 17 directors, of which 30% could be appointed by the Seller. The President of the Buyer and the Chancellor of the Seller each served on the Buyer’s board. Any director appointed by the Seller could be removed by the Seller without cause. Directors other than the President, the Chancellor, and those appointed by the Seller were appointed based on a majority vote of the Buyer’s board, could serve a term of no more than three years, and were limited to three successive terms. After the initial term of the Buyer’s board, no employee of the Seller other than the Chancellor could serve as a director of the Buyer.
With respect to the Buyer’s operations, the IRS noted that the Buyer independently sets its own operating and capital budgets, independently issues debt, and independently expends funds. The Seller could not require the Buyer to spend money for any Seller purpose, and could not hire, fire or set the compensation of the Buyer’s employees. The Buyer was solely responsible for collection of its accounts receivable and payment of its liabilities.
However, the Seller did have approval rights over certain major decisions of the Buyer. These approval rights were conceptually similar to investor protection rights customarily given to passive investors such as limited partners and minority stockholders. The specific approval rights that the Seller had with respect to the Buyer were over: (1) major corporate transactions by the Buyer not within the ordinary course of the Buyer’s business, (2) actions that would result in a change of the Buyer’s tax-exempt status, (3) academic affiliations by the Buyer with any educational institution other than the Seller, (4) material changes to the Buyer’s purposes, (5) any change to the fundamental, nonprofit, charitable, tax-exempt mission of the Buyer, (6) any action that would grant another entity or organization the right to appoint directors of the Buyer, (7) a joint operating agreement or similar arrangement under which the Buyer’s governance is substantially subject to a board or similar body that the Buyer does not control, and (8) the sale or transfer of all or substantially all the assets of the Buyer to a third party.
The IRS concluded that the Seller did not control the Buyer. In its analysis, the IRS noted that the Seller did not control the board of the Buyer (it could only appoint and remove without cause four of the Buyer’s 11 directors). Thus, the Seller was not considered to control the Buyer based on control of the board, which would have created a presumption under the control test that the Seller was in control of the Buyer. The remaining analysis was more nuanced. The ruling noted that the Buyer had operational independence from the Seller, which supported the conclusion that the Seller did not directly control the Buyer. However, the ruling acknowledged that the Seller’s approval rights were “a form of control” over the Buyer. The IRS characterized the Seller’s rights as a “power to bar the Buyer from taking certain actions” rather than a “power to cause the Buyer to act.” This let the IRS reach a favorable conclusion and avoid the conclusion that the Seller had “the right or power to require the use of funds or assets” of the Buyer “for any purpose” of the Seller.
 Four out of 11 board seats is approximately 36%, and three out of 11 is approximately 27%. Although the ruling does not address the point, there was presumably some sort of rule in the bylaws to address cases where 30% of the board was not a whole number. The right to appoint a fractional part of a board member would be unfortunate for everyone involved, especially the board member who got split up.