Harmless error and administrative appeals


A paragraph in a recent SJC decision (Mulready v. Maine Board of Real Estate Appraisers, 2009 ME 135 (12/29/09)) got me pondering.  It is an 80C decision, naturally to whetting my insatiable my administrative appeal appetite.

The decision itself is straightforward – it affirms a Superior Court decision that, in turn, affirmed a decision of the Board of Real Estate Appraisers (BREA) to issue a warning and order the appellant to pay hearing costs pursuant to 10 M.R.S. §§ 8003(5-A)(A)(2), (B)(2).

The section of the decision that has me musing is the following:

…the BREA erred in finding that Mulready should have evaluated 80 Exchange as a leased fee.

However, “[w]e will not disturb a judgment if an error is harmless.”  Tolliver v. Dep’t of Transp., 2008 ME 83, ¶ 39, 948 A.2d 1223, 1235. A preserved error is harmless “if it is highly probable that the error did not affect the judgment.” Id. The BREA’s error on this issue was harmless because it was not identified during the deliberations as a basis for finding a violation or determining a sanction, and because there was sufficient evidence, as discussed below, for the BREA to conclude that Mulready had violated Standards Rule 2-1(b). See id.

Mulready, 2009 ME 135, ¶¶ 28-20.  Tolliver is not an administrative appeal but involved a jury finding.

Thus, the SJC upheld an agency decision on harmless error grounds because (1) the deliberations did not indicate that the agency relied on the error; and (2) there was enough other material in the record to support the sanction.

The first question I asked myself is whether you can ever have harmless error in the administrative context.  A court isn’t supposed to substitute its judgment for the agency’s – that’s why, for example, the court can’t uphold an agency decision on alternate grounds:

‘in dealing with a determination or judgment which an administrative agency alone is authorized to make, (a court) must judge the propriety of such action solely by the grounds invoked by the agency.  If those grounds are inadequate or improper, the court is powerless to affirm the administrative action by substituting what it considers to be a more adequate or proper basis.  To do so would propel the court into the domain . . . set aside exclusively for the administrative agency.’  Securities & Exchange Commission v. Chenery Corporation, 332 U.S. 194, 196 … (1947).

In Re Maine Motor Rate Bureau, 357 A.2d 518, 526-527 (Me. 1976).

So, if a board lists – “invokes” – 100 reasons for deciding something, and one of those reasons is wrong, then it would seem to me that the court would have to remand and can't affirm, even if every single one of those other 99 reasons could sustain the decision.  To do so would be do substitute the court's reasoning for the agency's. 

On the other hand, if a board doesn’t say that it’s basing its decision on a particular fact or conclusion – if it’s not “invoking” – then why can’t the court affirm on harmless error grounds? Is merely stating a finding or conclusion in a written agency decision "invoking” that finding or conclusion as a basis for the decision, so that the court must remand?  Some agency decisions have gazillions of findings that go on for many pages, made on many preliminary and tangential matters.  It would wreak even more havoc than the normal delays if the itsiest little mistake in a written decision required a remand to clarify, causing yet more delay.

Hence, the SJC’s position that not every finding and conclusion in a written decision needs to be correct in order to affirm the decision makes sense.

The next question I asked is about the test the SJC applied for harmless error – looking, subjectively, at what the board said in its deliberations, and then, objectively, at the record for supportability absent the finding at issue.

On the subjective front, things get a little complicated because of the “deliberative process privilege.”  There is law for the proposition that courts shouldn’t look to statements made in deliberations; it’s only the written decision that counts.  See 1 M.R.S. § 407 (“Every agency shall make a written record of every decision involving the conditional approval or denial of an application, license, certificate or any other type of permit.  The agency shall set forth in the record the reason or reasons for its decision and make finding of the fact, in writing, sufficient to appraise the applicant and any interested member of the public of the basis for the decision”); 5 M.R.S. § 9061 (agency must make findings of fact and state them on the record or in writing, with a copy of the “decision” sent to the parties).  A primary reason for providing a privilege as to the deliberations preceding the written determination is “to protect the decisionmaking process by insuring that officials are judged by what they actually do, not by what they considered before deciding.”  R. Weaver & J. Jones, The Deliberative Process Privilege, 54 Mo. L. Rev. 279, 289 & n. 49 (1989), citing Russell v. Department of the Air Force, 682 F.2d 1045, 1048  (D.C. Cir. 1982);Jordan v. United States Dep't of Justice, 591 F.2d 753, 772-73 (D.C. Cir. 1978), overruled on other grounds, Crooker v. Bureau of Alcohol, Tobacco & Firearms, 670 F.2d 1051 (D.C. Cir. 1981); Marzen v. U.S. Dep't of Health & Human Servs, 632 F. Supp. 785, 813 (N.D. Ill. 1986), aff'd, 825 F.2d 1148 (7th Cir. 1987).  See also American Federation of Government Employees, AFL-CIO v. U.S. Department of Health & Human Services, 63 F. Supp. 2d 104 (D. Mass. 1999) (to prevent agency officials from being “judged not on the basis of their final decisions, but for matters they considered before making up their minds”), citing City of Virginia Beach v. U.S. Dep't of Commerce, 995 F.2d 1247, 1252-53 (4th Cir.1993).

Looking at deliberations is also a dangerous proposition as a practical matter because the statement of one board member can hardly represent the whole.  How would you conclude whether a majority felt one way or the other – count up statements by individuals?  This messiness is one reason why we require written decisions in the first place.

On the other hand, taken to the extreme, not looking at deliberations ever, for any reason, wouldn’t make sense either.  If, for example, all members of a board announced in deliberations that they were rejecting an application because the applicant was black, or a woman, or a Muslim, etc., then however pristine the subsequent written decision articulating other grounds for denial, I’m thinking that the deliberations might be relevant to an appeal.

Also, completely precluding examination of deliberations in all contexts could exacerbate a situation one sees sometimes when board members say one thing in deliberations, ask their counsel to write up a decision for them, and mysteriously the bad bases articulated in the deliberations are purged and replaced.  Even if this change is not as obvious as the exclusion of rampant expressions of prejudice, any kind of serious deviation between the articulation of grounds in deliberations and the written decision leads one to question who is actually making the decision – the elected or delegated officials, or their lawyer? (particularly if there isn’t a second vote to approve the written decision).

So again I’m on board, thinking yes, it could make sense to look at deliberations in some situations for some purposes.  And looking at them for clues to whether a finding was material could logically be one of them, in the proper context, understanding the limitations and potential pitfalls in examining this evidence.  If no one on a board makes any mention of the erroneous finding — which appears to be the situation in this case, while such silence might not be determinative of the question whether the error was harmless, it also isn't totally irrelevant — just as the announcement by a board member of the importance of the finding to him or her would have some relevance in showing that the error was not harmless.  Beyond these two ends of the spectrum, given the nature of deliberations, the usefulness of this evidence would appear diluted.

Similarly relevant as an end-of-the spectrum factor is the other point cited by the Court – if the remainder of record doesn’t support the decision, there can be no affirmance.  To be immaterial, at a minimum, you need to hack out the bad stuff and see if a logical, cogent decision, supported by the facts, remains.  If it doesn't, there can be no affirmance.

But even if a board doesn't mention a point in deliberations, and even if the record can support the ruling even with the error excised, that may not end the matter.  I’m also thinking that you need to look at the nature of the a finding or conclusion, it's tangentialness.  Once you hack out the bad finding, what remains – is it like removing an imperceptible hangnail, or has a major organ now gone missing?  And in the end, doesn’t that analysis sound much like the harmless error test in jury cases?

So, finally we reach my fundamental question – should the harmless error test be the same in administrative appeals as in appeals from jury cases (with the difference, of course, that you don’t have the jury’s deliberative statements to review)?  An error is deemed harmless in the jury context if it is ‘highly probable” that the jury didn’t rely on it.  Does that test make sense in the administrative context, when the board is both factfinder and legal interpreter?  Like sausages and politics, for various reasons, we don’t like looking under the rug at jury decisions – juries aren’t supposed to issue written decisions setting forth their logic.  In contrast, as noted above, we require agencies to do so.  And unlike judges, if an agency’s reasoning is not sustainable but alternate reasoning can sustain the conclusion, we, as noted, don’t say that’s ok – because the role of the court’s review of an agency decision is different than it is for a trial ruling.  Because an agency is to some extent both the jury and the judge, should that change the harmless error test in any way?

This entry is already far too long, so I’m going to stop there.  And keep pondering.