LEVIN H. CAMPBELL, Circuit Judge.

The Massachusetts Department of Correction (MDOC) petitions for review of a decision by the Law Enforcement Assistance Administration (LEAA) declining to award MDOC a corrections training grant under the LEAA's discretionary funds program, 42 U.S.C. § 3736(a)(2).

LEAA funding of state and local law enforcement activities falls into two categories. Under 42 U.S.C. § 3736, 85 percent of LEAA's funds are allocated to the states by population for block grants. The remaining 15 percent are available for distribution "in the discretion of the Administration" to eligible applicants "according to the criteria and on the terms and conditions the Administration determines consistent with this chapter." 42 U.S.C. § 3736(a)(2). Acting under the latter, discretionary, provision LEAA released a program announcement on September 27, 1976, soliciting applications for, among other things, a "Corrections Training Program." This brief announcement described the program LEAA envisioned and stated that the agency contemplated five grants ranging from $30,000 to $250,000 for one year. The same document indicated that applications would be evaluated through "advisory reviews by panels of experts."

On March 5, 1977, MDOC submitted an application seeking about $250,000 to provide Interpersonal Communications Skills Training (IPC) to prison guards and other "line-departmental staff." Seventeen other applications were received by LEAA. Pursuant to LEAA Instruction ORO I 4560.1, LEAA's central office in Washington assembled a panel of experts to review the submissions and recommend action to the agency. The panel met for approximately three hours on the morning of May 3, 1977. The panelists prepared a scoresheet for each application, indicating how well each did on seven criteria drawn from the original program announcement. After an initial tabulation of the scores, eight applications whose average score totalled 55 or more (on a 100-point scale) were discussed in more detail.

It is not clear from the record what transpired next. The panel may have recommended that only the top two applications (Vermont and South Dakota) be accepted. Other evidence indicates that the chairman, William C. Mooney, may have decided to rate, chairman Mooney solicited comments rate, chairman Moody solicited comments on each application from the various panelists. On May 6 he sent a memorandum to LEAA Assistant Administrator J. Robert Grimes describing the panel review, recommending that Vermont and South Dakota be funded, and giving "brief summaries of the major weaknesses of the remaining 16 grant applications." Grimes, in turn, sent Mooney's memorandum to each of the regional LEAA administrators with a cover letter, dated May 13, stating that LEAA concurred in the panel's recommendations. On May 19 George K. Campbell, Regional Administrator for the New England area, sent a letter to Frank Hall, Commissioner of MDOC, informing him that Massachusetts' application would not be funded. As the reason for this decision, Campbell quoted verbatim from Mooney summary of the "major weaknesses" of the Massachusetts application:

"The application did not reflect coordination with the Massachusetts Criminal Justice Training Council, which is authorized by statute to provide training. The training methodologies were not innovative, the approach to comply with Part E requirements is more properly funded with block grant funds, and Massachusetts has participated in three years of NIC sponsored training programs."

On June 8, 1977 Hall wrote Campbell to request a hearing on the grant denial under the provisions of 42 U.S.C. § 3758(b) and 28 CFR § 18.33. Pursuant to its regulations, 28 CFR § 18.31, the Administration responded by directing LEAA's general counsel's office to conduct an adjudicative investigation. Charles Lauer, Deputy General Counsel, prepared a report which went to Hall on September 9. Lauer concluded that the agency decision had been made in accordance with the preannounced guidelines and that it had a rational basis. Admitting that one of the four reasons given by Campbell — the Part E reason — was unfounded, Lauer nevertheless maintained that the record amply supported the other three grounds, "especially the key concern that the application did not propose an innovative delivery technology." Unimpressed with this explanation, Hall continued to press his claim for a full hearing.

Six months later, on March 10, 1978, an adjudicatory hearing was conducted before Charles Rinkevich, a hearing examiner for LEAA. After taking testimony from witnesses presented by both sides and considering the parties' exhibits and submissions, Rinkevich issued a 13-page report, dated May 22, recommending that LEAA's rejection of the MDOC application be set aside and that Massachusetts receive a fresh review of its proposal. Rinkevich concluded that none of the four reasons cited by LEAA in its original decision letter was supported by substantial evidence. Moreover, he decided that the review process was not conducted in accordance with departmental guidelines, and that these procedural defects contributed to the erroneous findings of the panel.

Rinkevich submitted his report to James Gregg, LEAA Assistant Administrator for Planning and Management, who made the final agency decision. LEAA's counsel, Charles Lauer, also submitted a brief to Gregg, setting forth the Administration's exceptions to the examiner's report. On September 15, 1978 Gregg wrote MDOC's attorney, Robert Claridge, informing him that LEAA had decided to uphold the denial of Massachusetts' application, despite the examiner's recommendation. Gregg gave four reasons for his decision. First, he said, the LEAA procedural guidelines supposedly violated by the panel were not imposed by statute or official regulation, and thus were not legally binding on the agency. Second, such departures from agency guidelines as occurred did not render the process unfair, and the resulting decision was not arbitrary, capricious, or an abuse of discretion. Third, the record contained substantial evidence to support the panel's decision, based on its findings that MDOC's proposal was not innovative. Gregg's fourth reason responded specifically to the allegations of procedural unfairness and indicated Gregg's conclusion that these asserted flaws did not "invalidate the ultimate findings and recommendations of the panel."

I.

The present petition for review was filed by Massachusetts in response to the Gregg letter. Section 3759 of 42 U.S.C. provides that any applicant "dissatisfied with the Administration's final action with respect to the approval of its application" or any applicant "dissatisfied with the Administration's final action under . . . section 3758 of this title [the hearing provision]" may petition the circuit court of appeals for the circuit in which the applicant is located for review of the action. § 3759(a). The court may affirm or set aside LEAA actions, in whole or in part, § 3759(c), or "for good cause shown," may remand to LEAA for the taking of further evidence, § 3759(b). "[T]he determinations and the findings of fact by the administration, if supported by substantial evidence, shall be conclusive." § 3759(b).

The substantial evidence standard requires an agency action to be supported by "'such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.'" Appleyard's Motor Transportation Co., Inc. v. Interstate Commerce Commission, 592 F.2d 8, 9 (1st Cir. 1979) ( quoting Universal Camera Corp. v. National Labor Relations Board, 340 U.S. 474, 477, 71 S.Ct. 456, 95 L.Ed. 456 (1951)). See also Bunny Bear Inc. v. Peterson, 473 F.2d 1002 (1st Cir. 1973). Here, we think the evidence adduced at the adjudicatory hearing was adequate to sustain the final denial by LEAA of MDOC's grant application. By the same token, LEAA's final action was not "arbitrary, capricious, [or] an abuse of discretion." 5 U.S.C. § 706(2)(A) (the Administrative Procedure Act). The evidence compiled at the adjudicatory hearing indicates that LEAA made an informed and not unreasonable decision to deny Massachusetts a discretionary grant. It is not our function to second guess such a decision.

Under the governing statute, 42 U.S.C. § 3736, LEAA is to dispense its discretionary funds "according to the criteria and on the terms and conditions the Administration determines consistent with this chapter." Petitioner interprets this to mean that LEAA must make its grant awards based on criteria announced in advance. Respondent does not quarrel with this reading, and we accept it for present purposes. The issue, therefore, is whether the decision is consistent with the announced criteria. We think it was, even though LEAA, during the course of its administrative proceedings, eventually concluded that three of the four "reasons" asserted for the panel action were either erroneous, arbitrary, or irrelevant to the established criteria for the program. Those three were that MDOC lacked Part E training, that the state had previously taken part in National Institute of Corrections training, and that the proposal had not been coordinated with the Massachusetts Criminal Justice Training Council. In his final decision, Mr. Gregg, the LEAA Assistant Administrator, relied solely on the finding that "the training methodologies described in the [MDOC] application were not innovative (as required by established criteria for funding)."

Innovation plainly was one of the criteria announced for selection of projects under this program. The grant announcement, contained in the September 27, 1976 LEAA Discretionary Grant Guideline Manual, described the "results sought" as " demonstration training programs for correctional officers" (emphasis added) and stated that: "Skills will be improved through innovative training delivery methodologies, including the use of software and audio-visual aids within the correctional system" (emphasis added). Further, the "Criteria for Selection of Grants," described in § 15(g) of the manual, states that "the nature of the training methodologies, including the use of technology," would be a factor.

There is also substantial evidence from which LEAA could find that Massachusetts' proposal was not innovative in the sense intended by the guidelines. MDOC proposed to provide something called Interpersonal Communications Skills Training (IPC) to its "line-departmental staff." IPC apparently is a relatively recent development in corrections training, first implemented in the Atlanta federal prison in 1969. However, in the last ten years, it has been used in 38 states as well as the federal prison system. Although IPC would be "innovative" for Massachusetts, it would not be innovative in the sense of serving as a model suitable for a "demonstration" project for transfer to other regions. So at least, the agency could reasonably conclude. A decision that funds under this program would be best utilized if reserved for projects with a potential impact in more than one state was within LEAA's discretion. The evidence fully supports the finding that Massachusetts' proposal did not fulfill this criterion.

Petitioner responds that the record does not support a finding that lack of innovativeness, all by itself, accounted for the panel's rejection of Massachusetts' application. Because the decision must have reflected, at least to some extent, the influence of the "extraneous factors" given as reasons for the denial, MDOC would have us find that the decision was not made "according to the criteria . . . the Administration determine[d]" as required by law. While this argument has some force, it would require more of LEAA's decision-making processes than we think is justified in the circumstances. Whatever was the case at the panel review level, LEAA's final decision did not rely on the discredited factors. It relied exclusively on lack of innovativeness, and that determination is supported by substantial evidence in the record. The record does not indicate that the panelists were acting out of bad faith or malice, or that any panelist had a personal bias against Massachusetts or any other state. One can understand MDOC's frustration when it discovered that some of the panelists mentioned reasons for opposing Massachusetts' application that were unrelated to the guidelines, but we do not believe these beginning errors sufficiently infected the entire process, as reviewed by the agency, to warrant setting aside a decision entrusted to LEAA's discretion.

This is not a case where none of the purported reasons for LEAA's decision was substantiated. Were none supported, the inference would be inescapable that the decision was not based on appropriate criteria. Here LEAA decided not to award a discretionary grant because the applicant failed to meet an established criterion of the program, and the record of the adjudicatory hearing contains substantial evidence to support that determination. The earlier findings and conclusions would be significant only were they to show that LEAA's action was unreasonable in light of the evidence available.

II.

Petitioner says that LEAA gave "inadequate treatment" to the hearing examiner's finding of serious procedural defects in the panel review process. The agency was of course not bound to adopt its hearing examiner's views. We understand MDOC's complaint to mean that the Administration abused its discretion by validating the decision despite evidence of serious procedural shortcomings. LEAA concedes that the agency's guidelines on the conduct of panel reviews were not observed to the letter. The issue is whether the admitted procedural deficiencies were so grave that the administrator's decision to disregard them was arbitrary, capricious, or a clear error of judgment. See Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971).

Of the eleven procedural deficiencies originally alleged by MDOC, the hearing examiner found a substantial basis for five: (1) failure of the panel chairman to challenge extreme scores; (2) failure of the chairman to make an adequate opening statement; (3) failure of the panel to give applicants a chance to rebut negative inferences about their applications; (4) failure of the chairman to challenge panelists' stated rationales for their scores; and (5) failure of the panel to select more applications for approval. Each of these asserted defects rested on the agency's non-compliance with an internal agency memorandum governing conduct of the panel review process. There could be no claim of non-compliance with the governing statute or official regulations of the agency — these did not specify any formal procedures for making the grants in question. Nor could petitioner assert a lack of due process. It had no statutory or other entitlement to the government benefit at stake. See Board of Regents v. Roth, 408 U.S. 564, 92 S.Ct. 2701, 33 L.Ed.2d 548 (1972). MDOC's claim rests entirely on the internal agency guidelines.

We would be troubled if LEAA had violated a procedural requirement imposed by law or official regulation that was meant to protect the rights of private parties. See School Committee of Monson v. Anrig, 520 F.2d 577 (1st Cir. 1975); Note, Violations By Agencies of Their Own Regulations, 87 Harv.L.Rev. 629 (1974). In Anrig, the court overturned a decision by the Massachusetts Department of Education to reject a local school system's application for a federal education grant. The court found that the state had failed to follow procedures specifically commanded by the governing statute and regulations. Further, the legislative history of the statute indicated that Congress had intended to circumscribe the states' authority in dispensing funds. Under these circumstances, the court held, the failure to follow the appropriate procedures, "burdensome [though] these procedures may be," required that the grant denial be vacated. Id. at 581.

The circumstances of this case differ significantly from those of Anrig. As the Anrig panel stated, "The case would be different if the provisions at issue were intended primarily as internal organizational guidelines, not designed to protect or actually relied on by affected parties." Id. at 580 n. 4. See also United States v. Caceres, 440 U.S. 741, 754 n. 18, 99 S.Ct. 1465, 1473, 59 L.Ed.2d 733 (1979); Chasse v. Chasen, 595 F.2d 59, 65 (1st Cir. 1979) (Campbell, J., concurring); Zabala Clemente v. United States, 567 F.2d 1140, 1144-45 (1st Cir. 1977), cert. denied, 435 U.S. 1006, 98 S.Ct. 1876, 56 L.Ed.2d 388 (1978). The procedures assertedly violated here were imposed only by the terms of an internal agency memorandum, LEAA Instruction ORO I 4560.1, dated April 4, 1977. This Instruction stated its purpose as follows:

"This instruction establishes guidelines for Office of Regional Operations (ORO) management of the Panel Review Process. Functions and responsibilities are defined. The process will conform to all requirements pertaining to grants and will be conducted in an orderly and expeditious manner."

The guidelines were not published in the Federal Register or Code of Federal Regulations, nor were they disseminated to applicants or the public at large. Petitioner does not contend that it even knew about, much less relied on, the terms of the Instruction in making its application. See Bauza v. Morales Carrion, 578 F.2d 447, 453 (1st Cir. 1978). It is not evident that the guidelines were intended to protect the interests of applicants, rather than simply to serve the administrative convenience of LEAA. The bulk of the Instruction is devoted to listing the responsibilities of each member of the LEAA hierarchy with regard to the panel review process, including such particulars as setting the dates of meetings and making travel arrangements for the panelists. We therefore agree with LEAA that these procedural guidelines did not carry the force of law.

Furthermore, the panel's deviations were not so serious that LEAA's decision to overlook them can be characterized as arbitrary or capricious. Failure of the panel chairman to make an opening statement was surely within the category of harmless error. Discrepancies in the scoring probably helped as much as hurt Massachusetts. Even if we accepted the hearing examiner's questionable conclusion that the guidelines required LEAA to furnish Massachusetts an opportunity to rebut the extraneous factors introduced at the panel meeting, we could not say that the state was fatally prejudiced under the circumstances, since the ultimate decision to deny MDOC's application was made without reference to those factors. Finally, the decision to recommend only two applications for funding, instead of the five originally contemplated, does not seem to have affected Massachusetts, since the MDOC proposal finished seventh in the panel rankings.

In short, there is substantial support for LEAA's conclusion that "failure of the panel to strictly adhere to the review process set forth in the Instruction" did not deprive MDOC's application of "fair and adequate consideration." No discrimination as between applicants was established: all were evaluated in essentially the same manner. Any prejudice to Massachusetts from the agency's failure to follow its own internal guidelines was so minimal that it was not unreasonable or arbitrary for LEAA to uphold the decision in spite of such deficiencies.

In sum, the action of LEAA in upholding the panel recommendation not to provide discretionary funds for Massachusetts was — if not a model of commendable agency behavior — at least free from major substantive or procedural defects. Given the limited nature of this court's review function under § 3759, and the large discretion vested in the agency in selecting grant recipients under this program, we deny Massachusetts' petition to set aside the LEAA decision.

Petition for review denied.