Can the Arkansas legislature change a constitutional amendment that voters passed by initiative, or do voters have to approve any change?
Subject
Whether the Arkansas General Assembly, under Amendment 7 to the Arkansas Constitution, has the power to amend a voter-approved initiated constitutional amendment by a two-thirds vote of each house, or whether the 1951 Edgmon decision that prohibited such amendments still controls.
Plain-English summary
State Senator Jim Dotson asked the AG a deceptively simple question: can the legislature, by a 2/3 vote, change a constitutional amendment that the people put into the constitution by initiative? Amendment 7 (the 1920 amendment that gave Arkansans the initiative and referendum power) says no measure approved by a vote of the people may be amended or repealed by the General Assembly except by a two-thirds vote of each house. The word "measures" expressly includes constitutional amendments. Read literally, the answer is yes.
But in 1951, the Arkansas Supreme Court in Arkansas Game & Fish Commission v. Edgmon refused to read it literally. The court called the literal reading "inconceivable" and held that the General Assembly cannot amend an initiated constitutional amendment, even by 2/3 vote.
AG Tim Griffin disagrees with Edgmon. His opinion argues:
- The plain text of Amendment 7 says what it says. Constitutional provisions get their "obvious and common meaning."
- The absurdity doctrine doesn't apply, because the literal reading isn't absurd. There are perfectly reasonable reasons to give the legislature this power: fixing technical defects in a passed initiative, dealing with unworkable language, etc. Arkansas's first three constitutions allowed legislative amendment without voter ratification, so it isn't an unprecedented arrangement.
- Stare decisis doesn't save Edgmon. The decision strayed from the constitutional text, and Arkansas precedent says courts will not follow precedent that "rewrites" plain language. Constitutional precedent gets less deference than statutory precedent because errors in interpreting the constitution are much harder to correct.
The AG's bottom line: if the question reaches the Arkansas Supreme Court today, he believes it would overrule Edgmon and let the General Assembly amend initiated constitutional amendments by 2/3 vote.
This is the AG's reading. Edgmon is still on the books and binding on lower courts unless and until the Arkansas Supreme Court overrules it. The AG is signaling that he would defend a 2/3 legislative amendment of an initiated constitutional amendment if the legislature passed one and a court challenged it.
What this means for you
State legislators
If you're considering a bill or joint resolution to amend a voter-approved constitutional amendment by a 2/3 vote, the AG has telegraphed that his office reads Amendment 7 to allow it and will defend that reading in court. Be aware:
- Edgmon is still binding precedent. A circuit court, court of appeals, or Arkansas Supreme Court panel that takes the case at first instance is bound to apply Edgmon unless the supreme court itself overrules it.
- A 2/3 amendment of an initiated constitutional amendment will draw an immediate constitutional challenge. Expect plaintiffs to seek a preliminary injunction.
- The AG's office (under Tim Griffin, who issued this opinion) will defend the amendment. Whether his successors will hold the same view is up to them.
- The strongest test cases would be amendments to initiated constitutional amendments that have minor textual or technical defects, where the absurdity doctrine and stare-decisis arguments are weakest against the legislature.
Ballot-initiative drafters and sponsors
If you're drafting an initiated constitutional amendment, the Edgmon protection that you may have assumed (that voters lock in the text and the legislature can't touch it) is no longer reliable. Two practical responses:
- Draft tightly so technical defects are minimized. The "fix-it" rationale is one of the strongest arguments for legislative amendment.
- Include language in the amendment itself constraining future amendment, if you can do so within Amendment 7's framework. The AG's opinion does not address whether a self-limiting amendment could prevent legislative amendment.
Election-law attorneys
This is a significant change in the AG's litigating position. Edgmon has been the assumed background rule for decades. If you advise initiative campaigns, government officials, or legislators, you should brief them that the AG's office will not defend Edgmon and may affirmatively argue against it.
Citizens who voted on a ballot initiative
The AG's reading does not, by itself, change anything. It tells you how the AG would handle litigation if the legislature tried to amend a constitutional amendment you helped pass. The legislature would still have to pass a 2/3 amendment in each house. If that happens, expect litigation. If you want to preserve the current state of the law (where Edgmon protects voter-approved amendments), pay attention to legislative votes that change voter-passed measures.
Constitutional-law scholars and judges
The opinion is a textualist critique of Edgmon, citing Scalia & Garner's Reading Law, Kimble v. Marvel, and Cooley's treatise. It applies the modern absurdity doctrine and Arkansas's recent precedent (Coble, Moore, Ward) on stare decisis and palpable error. Whether the supreme court agrees is another matter.
Common questions
Q: Has the legislature ever tried to amend an initiated constitutional amendment in modern times?
The opinion does not say. It addresses the legal question in the abstract. Edgmon itself involved a 1949 act of the General Assembly that conflicted with the 1944 initiated constitutional amendment creating the Game and Fish Commission.
Q: Does this mean the legislature can ignore voter-approved measures whenever it wants?
No. Amendment 7 still requires a 2/3 vote in each house, which is a high bar. Ordinary legislation only needs a simple majority. The AG's opinion does not weaken the supermajority requirement; it argues that the supermajority is the only requirement, and that the Edgmon gloss adding "and only for statutes, not constitutional amendments" was wrong.
Q: What's the practical difference between a constitutional amendment and a statute, then?
In Arkansas, both initiated constitutional amendments and initiated statutes can be amended by the legislature under Amendment 7's terms. Constitutional amendments still require a constitutional amendment to repeal in the strict sense, but the AG's reading would let a 2/3 legislative vote effect that change. The supremacy difference between constitutional amendments and statutes remains in the sense that constitutional text controls in conflicts with statutes.
Q: Could a future AG opinion or a court ruling reverse this?
Yes, both. AG opinions are not binding on courts; the next AG could withdraw or contradict this opinion. The Arkansas Supreme Court is the only authority that can definitively settle the question, and Edgmon remains the law until that court overrules it.
Q: Is the General Assembly limited to amending vs. repealing?
Amendment 7's text says "amended or repealed" by a 2/3 vote. The AG's opinion focuses on amendment, but the same logic would extend to repeal. A 2/3 vote could, on the AG's reading, repeal an initiated constitutional amendment outright.
Q: What about stare decisis protecting Edgmon?
The AG argues stare decisis is weakest for constitutional precedent (Kimble v. Marvel) and where the precedent strayed from text (Moore v. Moore). Both factors apply to Edgmon in his view. The Arkansas Supreme Court could disagree and follow Edgmon on stare-decisis grounds even if the original reasoning was flawed.
Background and statutory framework
Amendment 7, ratified in 1920, gave Arkansans the initiative and referendum. Article 5, section 1 of the Arkansas Constitution as amended provides: "[N]o measure approved by a vote of the people shall be amended or repealed by the General Assembly … except upon a yea and nay vote on roll call of two-thirds of all the members elected to each house of the General Assembly." The text of the amendment expressly defines "measures" to include "constitutional amendments."
In 1951, the Arkansas Supreme Court decided Arkansas Game & Fish Commission v. Edgmon, 218 Ark. 207. The case involved a 1949 statute that conflicted with the 1944 initiated constitutional amendment creating the Game and Fish Commission. The court acknowledged that, "literally," Amendment 7 would let the legislature change a constitutional provision by a 2/3 vote. The court refused to read it literally, calling that reading "inconceivable" and reasoning that if the people had intended that meaning, they would have used "more emphatic terms." The court held the General Assembly cannot amend an initiated constitutional amendment.
The AG's opinion applies more recent doctrine. Arkansas now reads constitutional provisions by their "obvious and common meaning" (Johnson v. Wright, 2022 Ark. 57; Ark. Dep't of Educ. v. Jackson, 2023 Ark. 140) and applies a narrow absurdity doctrine that requires both a no-reasonable-person-could-intend-this absurdity and a reparable-by-changing-a-word fix (State v. Coble, 2016 Ark. 114). The opinion also relies on Moore v. Moore, 2016 Ark. 105, for the rule that courts will not follow precedent that strays from plain text and "rewrites" the law.
The opinion's stare-decisis analysis draws on Kimble v. Marvel Entertainment (U.S. Supreme Court 2015) for the principle that constitutional precedent receives less weight, and on Jernigan v. Niblock (260 Ark. 406, 1976) and Rice v. Palmer (78 Ark. 432, 1906) for the Arkansas rule that courts must follow the constitution over conflicting precedent.
The AG also notes Arkansas's pre-Amendment 7 history: the 1836, 1861, and 1864 constitutions all allowed legislative amendment of constitutional provisions without separate ratification by the people, so the Amendment 7 reading he advocates is not historically unprecedented in Arkansas.
Citations
- Ark. Const. amend. 7 (initiative and referendum, including 2/3 amendment provision)
- Ark. Const. of 1836, 1861, 1864 (historical legislative-amendment provisions)
- Arkansas Game & Fish Comm'n v. Edgmon, 218 Ark. 207, 235 S.W.2d 554 (1951) (holding that legislature cannot amend initiated constitutional amendments; opinion argues this was wrongly decided)
- Johnson v. Wright, 2022 Ark. 57, 640 S.W.3d 401 (constitutional provisions get obvious-and-common meaning)
- Ark. Dep't of Educ. v. Jackson, 2023 Ark. 140, 675 S.W.3d 416 (rules of construction may not defeat clear and certain meaning)
- State v. Coble, 2016 Ark. 114, 487 S.W.3d 370 (narrow absurdity doctrine)
- Moore v. Moore, 2016 Ark. 105, 486 S.W.3d 766 (palpable-error departure from precedent)
- Ward v. State, 2015 Ark. 62, 455 S.W.3d 830 (palpable-error standard)
- In re Guardianship of W.L., 2015 Ark. 289, 467 S.W.3d 129 (return to clear language when precedent rewrites)
- Kimble v. Marvel Ent., LLC, 576 U.S. 446 (2015) (constitutional stare decisis is weaker)
- Jernigan v. Niblock, 260 Ark. 406, 540 S.W.2d 593 (1976) (constitution above precedent)
- Rice v. Palmer, 78 Ark. 432, 96 S.W. 396 (1906) (court is sworn to support the constitution)
- Little Rock & Ft. Smith Ry. Co. v. R.W. Worthen, 46 Ark. 312 (1885) (constitution is higher law)
Source
Original opinion text
Opinion No. 2024-024
November 15, 2024
The Honorable Jim Dotson
State Senator
Post Office Box 651
Bentonville, Arkansas 72712
Dear Senator Dotson:
I am writing in response to your request about the interpretation of Amendment 7 to the Arkansas
Constitution. You ask:
Can the General Assembly under Amendment 7 amend an initiated constitutional
amendment by a 2/3 majority?
RESPONSE
The plain language of Amendment 7 gives the General Assembly power to amend initiated
constitutional amendments. In 1951, the Arkansas Supreme Court reasoned that—read literally—
Amendment 7 does just that. But the Court departed from the text because it did not believe the
people could have intended to give the General Assembly that power. In my opinion, that decision
employed an erroneous form of reasoning and was wrongly decided. If this issue were raised today,
I believe the Supreme Court would overturn that case and hold that the plain language controls.
DISCUSSION
Constitutional provisions are given their "obvious and common meaning." "[A]rticle 5, section 1
of the Arkansas Constitution" is no exception: "Neither rules of construction nor rules of
interpretation may be used to defeat [its] clear and certain meaning …."
As relevant to this opinion, Amendment 7, which amended Article 5, § 1, provides that "[n]o
measure approved by a vote of the people shall be amended or repealed by the General Assembly
… except upon a yea and nay vote on roll call of two-thirds of all the members elected to each
house of the General Assembly." The word "measures" includes "constitutional amendment[s]."
There is no other restriction in Amendment 7 on the General Assembly's ability to amend a
constitutional amendment.
Thus, under the plain language of Amendment 7, the General Assembly can amend initiated
constitutional amendments.
In fact, the Arkansas Supreme Court reasoned, in Arkansas Game & Fish Comm'n v. Edgmon,
that, read "literally," Amendment 7 would allow the General Assembly to "change[] by legislative
action" the "meaning of a constitutional provision." But then it refused to read it literally; the
Edgmon Court believed that the plain meaning of Amendment 7 was "inconceivable," saying that
if the people had intended the plain meaning they should have made it even plainer. In other
words, the Court could not believe that the text meant what it said. Thus, it held—without reference
to the text—that the General Assembly was prohibited from amending constitutional amendments
by a two-thirds vote based on what the Court believed was the extratextual "intent" and "purpose"
of Amendment 7.
Edgmon did not mention the absurdity doctrine in its analysis, but it may have been why the Court
declined to follow the text. The doctrine, however, is inapplicable. It only applies when two
elements exist: (1) the absurdity must be something "no reasonable person could intend"—a
judge's belief that "the framers of the instrument could not intend what they say" is not enough—
and (2) "[t]he absurdity must be reparable by changing or supplying a particular word or phrase
whose inclusion or omission was obviously a technical or ministerial error." This narrow
definition is necessary to keep judges from substituting their personal policy preferences for the
plain text of the law. Since Edgmon, the Arkansas Supreme Court has adopted this particular
view of the absurdity doctrine. Neither prong of the doctrine is met here.
Amendment 7's grant of power to the General Assembly is not so absurd that no person could have
intended it. For example, it could be that this power was intended to allow the General Assembly
to change the Constitution by a supermajority vote if a ratified amendment failed to include
necessary provisions or if, in application, it became clear that the amendment was unworkable.
Not to mention, Arkansas already had experience with a similar arrangement: our State's first three
Constitutions allowed legislative amendments without ratification by the people.
Further, even if there were an absurdity, there's no "easy fix" to Amendment 7. There is no word
or phrase that was "obviously" left out or incorrect for the courts to "chang[e] or supply[]" with
scalpel-like precision.
Therefore, the absurdity doctrine provides no authority for a court to reinterpret Amendment 7's
clear grant of power to the General Assembly.
Although, in my opinion, the Supreme Court would likely hold that Edgmon was wrongly decided,
one might argue that stare decisis will now defeat the plain language of Amendment 7. Applying
stare decisis, courts will sometimes stick to decisions that the Court itself believes were wrongly
decided. But stare decisis is inapplicable here for two reasons.
First, courts will not follow precedent if the legal analysis has "palpable error." Thus, if precedent
strays from the "plain language" and "rewrit[es] it to achieve a contrary result," courts "are
compelled to serve justice by returning to the … clear language." As explained, Edgmon strayed
from the constitutional text.
Second, stare decisis carries less weight when interpreting the Constitution. That's so for at least
two reasons:
1. The Arkansas Constitution is our State's highest law; judicial precedent is not. To ignore
the meaning of the Arkansas Constitution in favor of stare decisis "would be to violate the
instrument [courts] are sworn to support"—a "gross[] derelict[ion] in duty."
2. If a decision misinterprets the Constitution, it is difficult to alter the Constitution to align
it with the correct interpretation. This contrasts with incorrectly interpreted legislation,
which is easier than the Constitution to alter because the legislature only needs to pass new
legislation to "correct any mistake it sees."
For these reasons, it is my opinion that, if the Supreme Court were faced with this question, it
would most likely hold that stare decisis is inapplicable.
Amendment 7's plain language gives the General Assembly power to amend initiated
constitutional amendments. In my opinion, if the Supreme Court were faced with your question
today, it would most likely hold that Edgmon was wrongly decided, employed a form of reasoning
that is now rejected, and should thus be overturned.
Deputy Attorney General Noah P. Watson prepared this opinion, which I hereby approve.
Sincerely,
TIM GRIFFIN
Attorney General