(concurring in part and dissenting in part). I concur with the majority that Pontiac Osteopathic Hospital is entitled to a remand for the trial court’s issuance of a certificate of need. Because the other health-care facilities are entitled to similar relief, I dissent from the rest of the majority opinion.
This Court originally held that the failure of the Department of Public Health to adopt and follow a state medical facilities plan (smfp), which had been expressly required by the Legislature, was error and that the department was therefore without power to process, much less deny, several hospitals’ applications for certificates of need. West Bloomfield Hosp v Certificate of Need Bd, 208 Mich App 393; 397; 528 NW2d 744 (1995). The Michigan Supreme Court reversed our decision on the narrow basis that an agency’s failure to adopt rules requisite to the processing of an application does not automatically make an agency’s decision invalid or subject to reversal. West Bloomfield Hosp v Certificate of Need Bd, 452 Mich 515, 524; 550 NW2d 223 (1996).
*527The Supreme Court, pointing to the case of American Farm Lines v Black Ball Freight Service, 397 US 532; 90 S Ct 1288; 25 L Ed 2d 547 (1970),1 held that the failure to have rules could be excusable if the missing rules were merely procedural and their absence would not create substantia] prejudice to the complaining party. I believe American Farm Lines holds, at its broadest, that a federal administrative agency may “relax or modify its procedural rules” when the “ends of justice require it,” id. at 533, but in no event can it be read to excuse the failure of the administrative agency to adopt rules if those rules are substantive, i.e., provide the decisional basis, for the agency’s ruling. Indeed, I believe, American Farm Lines must be read in this fashion. To do otherwise, would grant administrative agencies the right to adjudicate parties’ rights without due process for those parties. Certainly, our Supreme Court would not order such a thing because it would be contrary to the most fundamental understandings of the Michigan and federal constitutions. Needless to say, there is no case that gives such authority.
The question, then, before us is: Were the missing rules substantive or procedural in their import? I believe they can only be considered as substantive. The smfp was to inform the public of the things *528needed to be considered in deciding whether a certificate of need was justified. An applicant could look at the SMFP and determine what facts had to be shown to get a certificate of need and plan the case accordingly. This is not, by any reasonable understanding of the concepts, a procedural matter. A fair review of the record shows that the decisional basis of the department was the acute-care-bed-need methodology (acbnm),2 which, as the department described it, stood in lieu of the smfp. 452 Mich 522. If the acbnm was substantive, how can the smfp, for which it was a surrogate, not be?
Moreover, even if I were to assume that the department’s failure to adopt a state medical facilities plan was merely a procedural problem and not substantive, we must still look to the second part of the American Farm, Lines holding to determine if the parties suffered substantial prejudice from the department’s failure. On this record, I find that they did. One of the criteria the applicants had to establish was future need. MCL 333.22131(l)(n); MSA 14.15(22131)(l)(n). However, department witnesses conceded that the acbnm was not predictive.3 Simply *529stated, this means it could not do the job it was supposed to do. Because it was not predictive of future hospital need, it was no more than a snapshot, at best, of the current market. No matter how good that snapshot may have been, it cannot serve as a prediction. Accordingly, to use it as the test was arbitrary and unreasonable. Under such circumstances, using the acbnm snapshot methodology as a decisional standard caused the applicants substantial prejudice.
Because the department’s failure to adopt an smfp was not a procedural deficiency, and because it caused the applicants substantial prejudice, it is not necessary to reach the remaining issues in this case.
For the foregoing reasons, I would remand to the trial court for issuance of certificates of need to all the appellant health-care providers.
At the outset, it seems unusual that we are referred to American Farm Lines. The last time the Michigan Supreme Court cited American Farm Lines was in Michigan Farm Bureau v Bureau of Workmen’s Compensation, 408 Mich 141, 150-151, n 4; 289 NW2d 699 (1980). There, the Court noted the higher standards that are required of Michigan agencies, as contrasted with federal agencies, regarding an agency’s need to promulgate, as rules, their regulations, statements, standards, policies, and so forth. In any case, misgivings noted, I will, as we have been directed to do, analyze under American Farm Lines as I believe our Supreme Court currently understands it.
Phyllis Eaton, a staff analyst with the department’s planning section, testified that the department would rely only upon the acbnm and was precluded from relying upon any methodology other than the ACBNM.
Dr. Harvey Day, the chief of the Bureau of Health Facilities for the department, testified that the acbnm is not predictive of future market shares because it is only descriptive of retrospective market. The significance of this limitation is made clear by the following portion of the appellants’ joint brief, which summarized a hypothetical situation that expert James Lifton presented:
Assume that there were four hospitals in a particular subarea that were full of patients. The acbnm black box would measure, in the aggregate, the market share that had been captured by ail four hospitals and would calculate that their market share or volume of *529business justified all of the existing beds. Then if three of the hospitals burned down and were closed, the acbnm would only analyze the discharge data from the one remaining hospital in the subarea since the acbnm only measures the market (discharges) in the subarea from existing subarea hospitals. The acbnm would show that there were sufficient discharges being handled by the one hospital to justify only its beds, since that is all the capacity that the one hospital could handle. The acbnm would not show a need for the three hospitals that previously existed, or even one of the three that previously existed. ... It only looks at the market share of existing hospitals. ... It does not measure need for new beds, as required by Sections 22131(l)(c) and 33132 [sic]. It cannot predict whether a hospital will be full once it is built and it cannot predict the impact a full hospital will have on other hospitals.