Data from the last ten years of county and municipal elections in North Carolina reveals that voters rarely reject proposals to issue general obligation (“GO”) bonds at the ballot box. From November 2012 through and including November 2022, North Carolina voters approved 202 of the 213 (94.8%) county and municipal GO bond referenda put before them. Out of a proposed issuance volume of approximately $14.592 billion, voters authorized counties and municipalities to issue up to $14.376 billion in GO debt (or 98.5% of all GO bond authority sought).
This blog post explains (1) what a GO bond is, (2) why and how North Carolina local governments generally must obtain voter approval prior to issuing a GO bond, and (3) some of the information a local government currently must disclose when placing GO bond referenda before voters. Using elections results from county and municipal elections held over the last ten years, it also analyzes the high rate with which North Carolina voters have approved county and municipal proposals to issue GO bonds in the last decade.
What is a General Obligation (“GO”) Bond?
The North Carolina General Assembly has constitutional authority to authorize local governments to borrow money. See N.C. Const. art. V, § 4. Employing that authority, the General Assembly has permitted counties and municipalities to issue seven types of debt: (1) general obligation bonds (G.S. Chapter 159, Article 4); (2) revenue bonds (G.S. Chapter 159, Article 5); (3) special obligation bonds (G.S. Chapter 159, Article 7A); (4) project development financing instruments (G.S. Chapter 159, Article 6); (5) bond, tax, revenue, and grant anticipation notes (G.S. Chapter 159, Article 9); (6) installment financing contracts or limited obligation bonds (G.S. 160A-20), and (7) bonds or notes issued to the federal or state government to repay a loan from an agency of either entity (G.S. 160A-17.1(a)).
When a county or municipality issues the first type of debt—a general obligation bond—it pledges its “faith and credit” as security for the repayment of all borrowed principal and interest, meaning it promises to take all actions within its power—including levying any amount of property (ad valorem) tax necessary—to repay the debt that it incurred. See G.S. 159-46. This pledge creates a “general obligation” of the borrowing local government—and for that reason, bonds carrying such a promise are called “general obligation” or “GO” bonds.
North Carolina counties, municipalities, and certain other types of local governments may only issue GO bonds for specific purposes set out in state law. See generally G.S. 159-48(b) (listing purposes for which counties and municipalities both may issue GO bonds); (c) (county-specific purposes); (d) (municipality-specific purposes). The three most common types of issuances of GO bonds (by aggregate dollar amount) are county issuances to finance public elementary and secondary school facilities, county issuances to finance community colleges, and county and municipal issuances to finance parks and recreation facilities.
Voter Approval Requirement for GO Bonds
Except in limited circumstances described below, the General Assembly lacks constitutional authority to permit counties and municipalities to “contract debts secured by a pledge of [their] faith and credit unless approved by a majority of the qualified voters of the unit who vote thereon.” N.C. Const. art. V, § 4(2) (emphasis added). Therefore, in authorizing North Carolina counties and municipalities to issue GO bonds—which are secured by a pledge of a borrowing unit’s “faith and credit”—the General Assembly has generally required counties and municipalities to secure the “affirmative vote or a majority of those who vote thereon” in a referendum. G.S. 159-61(a); see also G.S. 159-49 (listing purposes for which voter approval of general obligation bonds is required).
Exceptions to the Voter Approval Requirement
Under the North Carolina Constitution, the General Assembly may permit counties and municipalities to pledge their “faith and credit” to repay a debt without voter approval in six circumstances. N.C. Const. art. V, § 4(2). Relying upon that authority, the General Assembly has excused several types of GO bond issuances from the voter approval requirement. See G.S. 159-49. In practice, however, local governments typically only utilize two exceptions—those for “two-thirds” bonds and refunding bonds—to issue GO bonds without voter approval.
- Two-Thirds Bonds
My colleague Kara Millonzi has previously written about the ability of local governments to issue “two-thirds” bonds without voter approval. Relying upon this exception, counties and municipalities may issue new GO bonds without voter approval—for a variety of purposes—in an amount not exceeding “two-thirds of the amount by which the . . . outstanding indebtedness [of the unit] has been reduced during the next preceding fiscal year.” G.S. 159-49(2). The term “indebtedness” has been interpreted to include only GO bond debt and GO bond anticipation note debt.[1] In other words, a county or municipality can issue new GO bonds—without voter approval—in an amount not exceeding two-thirds of the amount by which it reduced the principal amount of its GO bond indebtedness in the prior fiscal year. With several exceptions, the purpose for which a unit originally incurred such GO bond or GO bond anticipation note debt does not restrict the purposes for which it may issue two-thirds bonds.[2] .
- Refunding Bonds
A local government also need not obtain voter approval prior to issuing GO bonds to refinance (i.e., retire) existing GO bonds and bond anticipation notes. See N.C. Const. art. V, § 4(2)(a); G.S. 159-49(1); G.S. 159-48(a)(5). Local governments commonly refinance outstanding GO bonds when they can realize a savings on their total interest costs.
Required Form of Ballot Question for GO Bond Referendum
When a county or municipality places a GO bond issuance before its voters, North Carolina law prescribes the form of the question that appears upon a voter’s ballot. See G.S. 159-61(d). In particular, the question must be in “substantially the following words”:
Shall the order authorizing $ ______ bonds plus interest for (briefly stating the purpose) and providing that additional taxes may be levied in an amount sufficient to pay the principal of and interest on the bonds be approved?
[___] YES [___] NO
A local government, typically in consultation with bond counsel, drafts the exact purpose that will appear in the ballot question—and the governing board approves that language when fixing the date of the referendum. See G.S. 159-61(b). Generally speaking, a local government only has legal authority to expend the proceeds of a GO bond issuance for the purposes stated in the specific bond order and ballot question—and as a result, the purposes of a bond issuance are typically stated broadly, with little deviation from the text of the permissible purposes for a GO bond issuance set out in state law. See, e.g., Wake County Board of Commissioners, Resolution Calling for a Bond Referendum (July 11, 2022), § 4.
- Prior and Future Forms of the Ballot Question
The statutory form of the current ballot question has been in effect for almost ten years, applying to bond orders introduced on or after September 1, 2013. See S.L. 2013-200, §§ 4-5. For the previous forty years, however, the form of ballot question did not reference “interest” or note that “additional taxes may be levied in an amount sufficient to pay the principal of and interest on the bonds.” See S.L. 1973, c. 494, § 9(b). Instead, it stated only:
Shall the order authorizing $ ______ bonds for (briefly stating the purpose) be approved?
[___] YES [___] NO
A bill currently pending in the General Assembly—S99, entitled “Bond Referendum Transparency”—proposes further revisions to the form of the ballot question. S99, in its present form, would alter the question to require a local government to disclose (1) “[t]he estimated cumulative cost over the life of the bond, using the highest interest rate charged for similar debt over the last (maximum bond issuance term)” and (2) “[t]he amount of property tax liability increase for each one hundred thousand dollars . . . of property tax value to service the cumulative cost over the life of the bond provided above.”[3]
Required Disclosures for GO Bonds
S99 would seek to build in some respects upon new disclosure requirements that the General Assembly imposed in 2022 upon the GO bond issuance process. See S.L. 2022-53, §§ 1-4.
For bond orders introduced after October 1, 2022, the finance officer of a unit proposing to issue GO bonds must file—and post online—a statement of disclosure with the Local Government Commission (“LGC”) and the clerk to the unit’s board that contains (1) an estimate of the total amount of interest that will be paid on the bonds over their expected term, if issued, and a summary of the assumptions upon which that estimate is based; (2) an estimate of the increase in property tax rate, if any, necessary to service the proposed debt; and (3) the amount of “two-thirds” bonds that the unit may issue in the current fiscal year. See G.S. 159-55.1(a) and (c). If the finance officer determines that a property tax increase may be necessary to service the proposed debt, they also must now include that information when they publish the bond order, as introduced and as adopted. See G.S. 159-56 and 159-58. As the title of S.L. 2022-53 (An Act to Increase the Transparency of Bond Referenda By Requiring Additional Disclosures By Units of Local Government) suggests, these reforms aimed to increase the amount of information provided to citizens voting in a bond referendum.
Local GO Bond Referenda in North Carolina Over the Last Decade – Types and Success Rate
From November 2012 through November 2022, North Carolina voters approved 202 of the 213 (94.8%) county and municipal GO bond referenda put before them. Out of a proposed issuance volume of approximately $14.592 billion, voters authorized counties and municipalities to issue up to $14.376 billion in GO debt (or 98.5% of all GO bond authority sought). The data can be accessed here: GO Bond Election Referenda – 11-12 – 11-22.
In that time period, some types of GO bond referenda were much more common than others—and bond proposals for public schools strongly overshadowed all other types of proposed issuances The vast differential between counties and municipalities in the amount of GO indebtedness proposed can be explained by the fact that only counties—not municipalities—may issue GO bonds for public schools (G.S. 159-48(c)(4)) and community colleges (G.S. 159-48(c)(1)). By volume, these two types of referenda accounted for more than 68% of all general obligation bonds proposed during the last decade.
The five most common types of referenda, by number of separate proposals, and the average issuance proposed per each type of referendum are reflected below.
From November 2012 through November 2022, only 11 GO bond referenda failed across the state—two in counties and nine in municipalities. Only one GO bond referendum out of thirty-three held for public schools failed to obtain voter approval during that time period.
The failed referenda for parks and recreation bonds proposed to finance a variety of projects—in Hendersonville (Nov. 2013), Goldsboro (Nov. 2014), Harrisburg (Nov. 2017), Mint Hill (Nov. 2018), Cape Carteret (Nov. 2020), and Mount Holly (Nov. 2021). The five other failed referenda proposed to authorize GO bond issuances (1) by the City of Wilmington to finance the acquisition and construction of a minor league baseball stadium, (2) by Onslow County to pay for public school facilities, (3) by the Village of Bald Head Island to finance a broadband network, (4) by the Town of Mint Hill to build a cultural arts center, and (5) by Union County to build and equip a 4-H pavilion.
What Do These Numbers Suggest?
The data from the last ten years clearly demonstrates that North Carolina voters almost always approve local bond referenda. The much more difficult question—and one that can’t be answered without additional empirical analysis—is why.
A wide variety of variables might affect the success or failure of a bond referendum. For example, the types of voters that actually vote in a local election could affect the outcome of an election. So, too, a voter’s knowledge about the potential effects of a bond issuance on a local government’s finances or on the voter’s household finances (e.g., a higher property tax bill) might affect his or her decision to support or oppose a referendum. It would seem rational for an individual to oppose a potential increase in his or her property taxes—but at the same time, a voter may support such a trade-off if he or she has sufficient knowledge of and supports the projects to be financed. Even if a voter understands all of these potential effects or the scope of the proposed project, it still is difficult to determine how salient these understandings might be to their decision to vote.
Academics have explored some of these topics in other states, but it is worth additional study in North Carolina. As the provisions of S.L. 2022-53 (and potential revisions to the form of ballot question) are implemented in the coming years, it will be interesting to track whether we see any decrease in the relatively high approval rate of GO bond referenda seen over the past decade.
Another interesting question—and one difficult to analyze—is how frequently North Carolina’s local governments elect not to proceed with a bond referendum due to concerns that a referendum will fail. Although GO bonds typically bear the lowest interest rate of debt available to a local government, the process to issue GO bonds is lengthy and procedurally challenging—and a local government that fails to obtain voter approval will have wasted time and public funds to secure approval of something that might be more quickly and economically financed by another type of debt (e.g., installment financing). In addition, bringing forward an unpopular proposal may have political consequences for elected officials. Gathering data on the types of projects that communities elect not to finance with GO bonds due to referendum concerns also merits additional study.
________________________________________________________________________________________________________________________________________________________________________________
