The ABA’s recent Ethics Opinion No. 518 has sparked a discussion about how, if at all, it may restrict the use of mediators’ proposals. My estimable colleague Jeff Kichaven has argued that the Opinion, and other ethical rules or standards, prohibit all but one form of a mediator’s proposal.[1} Because I fear his argument may cause mediators to refrain (on dubious grounds) from using a time-honored practice that has successfully settled many cases, I wanted to elaborate further my defense of that practice, which I have begun to advance elsewhere.[2]
ABA Ethics Opinion No. 518
Let’s begin with a brief description of ABA Opinion 518. It makes two main points. First, it notes the requirements of Model Rule 2.4(b) that a lawyer-neutral should explain to unrepresented parties (or others the neutral knows or reasonably should know do not understand the neutral’s role) that he or she is not representing them as a lawyer represents a client. The neutral also should avoid communicating “in a manner that might be taken as rendering legal advice” or suggesting that the neutral’s role “is to protect or advance a party’s legal interests or to help the party to attain a particular desired result.”
Second, the Opinion notes that Model Rule 8.4(c), which prohibits a lawyer from engaging in conduct involving dishonesty, fraud, deceit or misrepresentation, also applies to a lawyer who acts as a neutral, not just one who represents a client. It then elaborates further on how that rule applies to the conduct of a lawyer-neutral.
Notably, Opinion 518 does not mention or discuss mediators’ proposals at all. The jumping-off point for Kichaven’s critique is the Opinion’s statement regarding Model Rule 2.4(b) that a mediator should not state “that a proposed settlement is in a party’s best interest.” The Opinion also states that a mediator “may offer an opinion as to how a tribunal is likely to rule on an issue, but the lawyer-mediator should not state or imply that a settlement is in the party’s best interest because a tribunal is likely to rule adversely to the party.” (Emphasis added in part.)
Kichaven’s Preferred Method: The Pre-Agreed Mediator’s Proposal
Kichaven has argued that only one form of mediator’s proposal is now ethically permitted—which he calls the “vetted mediator’s proposal.”[3] When the parties are at an impasse, the mediator floats a number to both sides’ counsel, separately or possibly in a joint meeting, to seek their agreement on a number they will recommend to their clients. Counsel either accepts that number or negotiates a different number that both counsel can recommend. The mediator then “capstones the negotiation by putting it into the form of a mediator’s proposal. The lawyers next present that number to their clients with the full weight of the mediator’s authority behind it.”[4]
This cherry-on-top approach is not really a “mediator’s proposal” so much as a “mediator’s blessing” of a number already agreed to by counsel. That is not to denigrate it: such a blessing may well be useful in persuading other constituencies to jump on board, such as insurance carriers or recalcitrant client representatives. But the mediator is only “proposing” what has been tentatively agreed to already by counsel.
The “Classic” Non-Pre-Agreed Mediator’s Proposal
The classic double-blind mediator’s proposal, which Kichaven says is now ethically forbidden, differs from his preferred method in one main respect: the number the mediator proposes is not one to which counsel have agreed in advance. It works like this.
Near the conclusion of the mediation, the mediator asks the parties if they would like a mediator’s proposal or (as often occurs) the parties ask him or her to make one. Either way, no proposal is made unless the parties welcome it. If they do, the mediator selects a relatively narrow range of numbers and asks each side privately whether it would seriously consider settling at one or more numbers within that range. If both sides say yes, the mediator then proposes a single number to both sides within the range. After a suitable deliberation period, each side then confidentially tells the mediator whether it accepts. If both sides accept, the mediator announces agreement at that number. If one side or both do not accept, neither side is told whether the other agreed; the mediator says only that there is no agreement. Both sides then revert to their prior negotiating positions.
This “double blind” feature encourages flexibility by preventing a party’s willingness to agree to the new proposed number from becoming a new floor or ceiling if the other side declines to move. No one but the mediator knows that the party was willing to move off the prior number if the other side does not agree, and both retain their prior positions.
Although Kichaven calls his method a “vetted” mediator’s proposal, the classic non-pre-agreed proposal also has been vetted with the parties to a large extent—and is not, as Kichaven has tried to deride it, an “unvetted, pulled-from-the-ear number.”[5] The number proposed is within a range the parties have already agreed they would seriously consider. The number chosen within that range (and indeed the range itself) is carefully selected after a long day of discussions with each side, during which the mediator has engaged with the parties on the merits, floated numbers, gauged reactions, and listened carefully to the parties’ hopes and needs. As Norman Feit has explained in this publication, “[a]n effective mediator gleans that the proposed level is viable only through the process, many ex parte discussions, and a good deal of body language.”[6] For this reason, Kichaven’s preferred method might more accurately be called a “pre-agreed” mediator’s proposal, because the method he criticizes is also “vetted” in meaningful ways, directly and indirectly.
Kichaven has offered several reasons why mediators’ proposals other than a pre-agreed one are not ethically permitted, one of which derives from Opinion 518 and others of which do not. First, he argues that, unless it selects the midpoint between the parties’ last bid and ask, the non-pre-agreed mediator’s proposal abandons the mediator’s requisite neutrality. Second, he argues that a non-pre-agreed mediator’s proposal inherently states or implies the mediator’s judgment that the number is in both parties’ best interest, contrary to the dictates of Opinion 518. Third, he argues that a non-pre-agreed mediator’s proposal usurps the role that counsel should play in the negotiation and undermines party self-determination. Each of these reasons is unpersuasive.
Neutrality Does Not Require a Midpoint Proposal
Kichaven has argued that “[t]he only mediator’s proposal that is theoretically neutral is the midpoint between the last offer and last demand.”[7] As a result, he claims, “[o]nce a mediator chooses a non-midpoint number, neutrality is gone.”[8] This argument is misguided.
As I have previously written, and as mediators frequently tell parties, there is nothing magic about the midpoint. If a plaintiff says it will move down to 30 if the defendant moves up to 10, that doesn’t mean the plaintiff will settle for 20. Nor does it mean the defendant will settle for 20 if it agrees to the bracket. Of course, the parties will frequently consult the midpoint of a bracket to see what it may signal, and those signals may help move the mediation along, but the midpoint is simply a temporary relationship between two moving numbers.
Indeed, the respective rates of change between the parties’ demands and offers may well suggest a settlement at something other than the midpoint. Suppose the plaintiff has moved down to 45 in three increments of 10 (75 to 65, 65 to 55, and 55 to 45), and the defendant has moved up to 30 in three increments of 5 (15 to 20, 20 to 25, and 25 to 30). The midpoint between the last demand of 45 and the last offer of 30 is 37.5. But the recent movement of the parties suggests they may have been pointing instead to 35—i.e., if plaintiff were again to come down 10 and defendant were again to move up 5. Of course, there may be valid reasons why either number (or some other) might be the right one to propose, but the mediator has by no means lost neutrality if he or she proposes a settlement at the non-midpoint figure of 35 as the one most likely to settle the case.
To similar effect is Feit’s critique, where he argues that Kichaven’s “syllogism assumes that only the mid-point is the ‘neutral’ and ‘fair’ outcome.” Yet “the mediator has no way of knowing how each side truly values the case, and where each seeks resolution vis-à-vis the mid-point, which itself often changes dramatically as the gap closes.” Ultimately, “a proposal is no more or less neutral when not at the precise mid-point, so long as the mediator clearly expresses the proposal in terms of a realistic resolution both sides may find acceptable for their own respective reasons.”
A Mediator’s Proposal Does Not Express or Imply a Best-Interest Judgment
In his critique that derives from Opinion 518, Kichaven has argued that a non-pre-agreed mediator’s proposal shows that the mediator must believe the proposal is “equally good for both sides.”[9] (Emphasis added). To believe that, the “mediator must first believe the proposal is ‘good’ for each side.’” He then argues: “If the mediator believes the number is the best a party can get in the negotiation, and better than the alternative of going to trial—in other words, ‘good’—that is a best-interest judgment” that a mediator is forbidden to make. (Emphasis added.)
The argument wrongly assumes that a mediator is advising the parties that it is in their best interests to settle at the selected number—something that, of course, Opinion 518 forbids a lawyer-mediator to say. The classic mediator’s proposal does not say that. In its usual form, the proposed number represents only what the mediator thinks is most likely to get agreement by both sides—and, indeed, only as things stand today. (For example, if a party decides to keep litigating, it might win that summary judgment motion or that motion for class certification, at which point the parties’ relative bargaining power might change, with a corresponding change in the most-likely-agreeable number.) Whether it is in a party’s best interest to accept that number is something on which the mediator expresses no view. That question is left to the parties’ counsel alone to answer and advise their clients about.
Now, the dictates of Opinion 518 do suggest a helpful practice tip. Going forward, mediators should make clear to the parties, if they haven’t before, what their proposed number represents and does not represent. Stating that the proposed number is simply the one most likely to be accepted avoids any argument that a mediator’s proposal may imply an assertion that it is in the parties’ best interest to accept the proposed number. When the intended meaning of the proposed number is clearly stated, nothing in Opinion 518 forbids it.
Other commentators have also rejected the conclusion that Opinion 518 forbids a mediator’s proposal. As Feit has argued, “[i]f vetted and communicated properly, the mediator’s proposal makes no comment on its fairness to each side, but merely reflects the mediator’s perception of a number that both sides may accept based on their own analysis and rationale.”
Likewise, Noah Hanft has noted that the exact import of the mediator’s proposal, whether the number is based on the “chances of acceptance” or otherwise, “should be discussed explicitly before any proposal is made,” but that, when this is done, “mediators can continue to use proposals ethically and effectively.”[10]
The Classic Non-Pre-Agreed Mediator’s Proposal Does Not Usurp Counsel’s Function or Undermine Party Self-Determination
As Kichaven’s argument has developed, he has appeared to shift away from the non-midpoint and best-interest arguments to ones based more on the obligations of counsel and the need for party self-determination. These arguments, which are not based on Opinion 518 at all, are no more persuasive.
First, Kichaven argues that the pre-agreed approach is superior because the proposed number results from the negotiations of counsel, who are exercising their obligations under Model Rule 1.1 to act competently and under Model Rule 2.1 to exercise independent professional judgment. Such a proposal “validates, and does not replace, the lawyers’ independent judgment about the best available deal.”[11] Done Right at 1. In his view, if the mediator selects a number that counsel have not agreed to in advance, this improperly “substitutes the mediator’s judgment for the lawyer’s work.”[12] Id. at 2.
But the classic mediator’s proposal does not usurp the role of the parties’ counsel or undermine their ability competently to exercise independent judgment. Counsel still must exercise judgment when advising their clients whether to accept the proposal. Indeed, they are better able to exercise that judgment when they have the valuable information that a classic mediator’s proposal provides.
Second, Kichaven argues that a non-pre-agreed mediator’s proposal undermines party self-determination in violation of ABA/AAA/ACR Model Standards of Conduct for Mediators 1(B). When a non-pre-agreed number is proposed, Kichaven argues, “[i]t is the mediator and not the parties who decides where the case should settle.”[13] Mediation After at 2. Relatedly, he argues that the mediator’s proposal “is a species of adjudication” without proper notice and opportunity to be heard and that, although the parties are free to accept or reject it, “the critical decision—the selection of the number itself—has already been made without the adversarial testing competent representation and independent judgment require.”[14] Done Right at 2.
Again, however, in making a mediator’s proposal, the mediator does not “decid[e] where the case should settle.” The “critical decision” is still made by the parties alone, if and when they accept the proposal after receiving the advice of their counsel. Only then does the number the mediator “selects” have lasting meaning. On the other hand, if a party rejects the selected number, which it is fully free to do, then the number has no further significance. It becomes simply another proposal that was floated and not accepted. Nothing was “decided” or “adjudicated,” and the parties retain full control over where the matter settles, if at all.
One final caveat is that the analysis here assumes both sides at the mediation are represented by counsel, who are exercising independent judgment. Court-annexed and community mediation programs, however, handle pro se litigants regularly. In such cases, special concerns arise, some of which Opinion 518 directly addresses. For example, “[u]nless the parties are sophisticated consumers of mediation services, it is prudent for the lawyer-mediator not only to inform all parties that the lawyer-mediator does not represent them but also to afford them an opportunity to discuss what this means.”[15] Op. 518 at 2. How mediators’ proposals might ethically be presented in such a setting is beyond the scope of this article.
Kichaven’s Argument is Harmful and Proves Too Much
If Kichaven likes the pre-agreed method, and doesn’t like the non-pre-agreed one, that’s his prerogative. He can use or not use whatever methods he chooses. He also can advocate to others the advantages he sees in following his favored approach. But the suggestion that the classic method is unethical is, in my view, not only wrong (for the reasons discussed above) but also harmful.
First, if accepted, his argument would deprive the parties of important information that the mediator is uniquely able to provide and that the parties want to know. Mediators unquestionably have valuable information about the other side that the parties’ counsel simply do not know. After all, mediators have been alone in the room with the other side for much of the day. Ideally, the parties in those sessions have been more candid with the mediator than they ever had been with the other side. For that reason, the lawyers simply do not know what the mediator knows about what outcomes may be realistically achievable at that time. Equally important, this is information the parties want to know. Remember, mediators make a proposal like this only when both sides want them to.
Therefore, the perverse result of Kichaven’s argument is that, when the mediator uniquely has information that the lawyers do not know (and cannot otherwise get), that is extremely valuable to the lawyers in advising their clients, and that the parties want to know, the mediator can’t give them that information. It is difficult to see what legitimate interest is served by such a result.
Second, Kichaven’s argument is also harmful because it would deprive mediators and the parties they serve of a method that can achieve settlement when his preferred method cannot. Getting both sides’ counsel to agree on a number they will recommend to their clients is a lovely closing technique when the parties are sufficiently close and settlement-minded to allow it, but that happy result is not always obtainable. All too often, the parties and their counsel at day’s end remain entrenched at numbers that leave too large a gap for the mediator to secure such an agreement from counsel. If mediators were prohibited in that circumstance from taking the leap of proposing a number, many settlements that could otherwise have been achieved at the mediation would simply not be made.
In addition to being harmful, Kichaven’s argument proves too much. If suggesting a possible landing place in the negotiation is legal advice that is the responsibility of the parties’ counsel alone to provide, then why is that problem solved by making the recommendation after the lawyers have agreed on a number? If a mediator’s proposal necessarily says or implies what’s in the client’s best interests, the mediator would have no business making the recommendation at any time—even as the cherry on top. In short, the effort to save one form of mediator’s proposal, while condemning another on ethical grounds, simply does not make logical sense.
Conclusion
In short, the classic mediator’s proposal remains a valuable and viable option notwithstanding Opinion 518, but mediators should make sure the parties understand that the proposed number is simply what the mediator thinks is most likely to be agreed to by both sides, and that the mediator is not advising what settlement number will best serve a party’s interest. When that is clear, nothing in Opinion 518 forbids such a proposal, whether or not pre-agreement by counsel has been obtained. Sadly, unfounded ethical objections to the classic mediator’s proposal may have the unfortunate effect of causing mediators to avoid a method that can settle many cases when pre-agreement by counsel cannot be obtained—and that gives the parties valuable information that they have requested and cannot obtain on their own.
-----------------------------------------
[][1] Jeff Kichaven, “Mediator’s proposals – done right,” Daily Journal, Mar. 20, 2026; Jeff Kichaven, “The Mediator’s Proposal after ABA Opinion 518: Best Practices Just Became Ethical Requirements,” ABA Litigation Section Newsletter, Feb. 6, 2026; Jeff Kichaven, “Mediation After ABA Opinion 518: What are a lawyer’s responsibilities?,” Daily Journal, Dec. 3, 2025.
[2} “Walking the Tightrope: Mediator Ethics After ABA Opinion 518,” Cal. Lawyers Ass’n Webinar, April 29, 2026 (Helm and Kichaven panelists); Mark Helm, “In Defense of Mediators’ Proposals: Clarifying ABA Opinion 518,” Daily Journal, Feb. 5, 2026.
[3] Jeff Kichaven, “Mediator’s proposals – done right,” Daily Journal, Mar. 20, 2026, at 2 (“Done Right”).
[4] Id.
[5] Jeff Kichaven, “Mediation After ABA Opinion 518: What are a lawyer’s responsibilities?” Daily Journal, Dec. 3, 2025, at 2 (“Mediation After”).
[6] Norman Feit, “Avoiding Misleading Mediator Communications,” Mediation Magazine, Mar. 9, 2026.
[7] Mediation After, at 1.
[8] Id.
[9] Mediation After, at 2.
[[1]0] Noah Hanft, “Mediator’s Proposals After ABA Opinion 518: Ethics, Effectiveness, and the Reality of Breaking Impasse,” Lexology, Feb. 26, 2026.
[11] Done Right, at 1.
[12] Id., at 2.
[13] Mediation After, at 2.
[14] Done Right, at 2.
[15] ABA Ethics Opinion No. 518, at 2.