General Dynamics Corp. v. Board of Assessors of Quincy , 388 Mass. 24 ( 1983 )

388 Mass. 24 (1983) 444 N.E.2d 1266

GENERAL DYNAMICS CORPORATION
vs.
BOARD OF ASSESSORS OF QUINCY & another.^[1]

Supreme Judicial Court of Massachusetts, Suffolk.

September 15, 1982. January 21, 1983.

Present: WILKINS, LIACOS, ABRAMS, NOLAN, & LYNCH, JJ.

John M. Lynch (Wayne E. Hartwell with him) for the defendants.

James K. Brown (Philip Burling & Donald R. Ware with him) for the plaintiff.

James A. Aloisi, Jr., Assistant Attorney General, for Attorney General, amicus curiae, submitted a brief.

WILKINS, J.

This case involves the fair cash value on January 1, 1978, for the purposes of local taxation, of that portion of the General Dynamics Corporation's shipyard, known as the Quincy shipyard, which is within the city of Quincy.^[2] There is a further issue concerning the timeliness of General Dynamics's applications for abatement for the 1977 fiscal year. The city appeals from a judgment of the Superior Court ordering the payment of substantial abatements in favor of General Dynamics. We granted the city's request for direct appellate review and affirm the judgment.

The parties agreed that the highest and best use of the property was as a shipyard, an integrated industrial and business entity engaged in the fabrication and repair of ships — in short, a special purpose property. The judge concluded, and the parties agreed, that the shipyard could not be valued by considering sales of comparable properties, because there was no such sale, or by capitalizing income, because the hypothetical rental income of the individual buildings would not measure the value of the property as a shipyard. The judge rejected the city's argument that the depreciated replacement or reproduction cost (DRC) of the property should be used. He found that "[t]he physical state of the Shipyard is such that [a purchaser] would not consider useful any calculus of cost to replace or reproduce any building." The judge therefore concluded that each of the "normal" methods of valuation was inappropriate in these special circumstances because each focused on individual buildings (and associated land) rather than on the value of the shipyard as a unit. He determined the shipyard's fair cash value, consistent with a method advocated by General Dynamics, on the basis of the value of the business conducted at the shipyard reduced by (a) the value of "intangibles," (b) the value of nontaxable equipment, and (c) the portion of the shipyard allocable to Braintree. The city argues that the judge was required as a matter of law to use the DRC method to arrive at the fair cash value of the shipyard, and that the method he chose was improper. It further contends that, even if this method may be used, the judge misapplied it on the facts before him.^[3]

1. Neither party nor the Attorney General in his amicus curiae brief challenges the propriety of this local tax assessment matter coming to us through the Superior Court rather than from the Appellate Tax Board (board) following a hearing and board decision. It is important to note the significance of this departure from the usual practice of having a decision by the board on a challenge to local assessments. We know that both parties wanted the matter to be decided in the Superior Court rather than before the board, that the board acquiesced in that respect by staying all proceedings on General Dynamics's timely appeals to the board, and that a judge of the Superior Court, after hearing and without objection, accepted jurisdiction of the case "in all its aspects."

The question whether to accept a tax matter for judicial determination in lieu of an agency determination is a matter of discretion. See Sydney v. Commissioner of Corps. & Taxation, 371 Mass. 289, 293-294 (1976); East Chop Tennis Club v. Massachusetts Comm'n Against Discrimination, 364 Mass. 444, 449-451 (1973) (comparing tax cases with other administrative matters). Certainly, at the outset, it was apparent that this case presented novel questions of law. But there were factual issues as well. As is clearer with the advantage of hindsight, the factual issues were substantial and may even have been predominant.

A decision by one judge of the Superior Court on how to arrive at the fair cash value of property cannot have the same Statewide impact as a decision of the board. By its decisions, the board establishes a body of applicable precedent that serves as a guide for assessors and taxpayers in all municipalities. We have recognized, however, that there are situations that justify the Superior Court's taking jurisdiction. See Sidney v. Commissioner of Corps. & Taxation, supra at 294-295. Some factors we have considered warranted the exercise of jurisdiction by the Superior Court: the issues were important, at least to the parties, and there were novel legal issues regarding the proper valuation of the shipyard. The valuation problem may be recurrent, although that is not certain. There were, however, factors weighing against exercising jurisdiction over the case. The unique nature of the shipyard made it unclear whether any decision would have public significance, affecting persons other than the litigants. Nor, as we have said, does the case reduce "to an issue of law without dispute as to the facts." Id. at 295.

Not only was there doubt whether the decision of the Superior Court judge would serve as a significant guide in other cases, there is also the possibility that our opinion may not furnish as firm a guide as would an opinion of this court following a decision of the board. This court has deferred in numerous instances to the judgment of an administrative agency charged with the implementation of a legislative mandate. If there is "substantial evidence to support [an agency's] decision, we defer to the [agency's] judgment as to what evidence to accept and which method or methods of valuation to rely on." Boston Edison Co. v. Assessors of Watertown, 387 Mass. 298, 302 (1982). See New Boston Garden Corp. v. Assessors of Boston, 383 Mass. 456, 466-467 (1981). We grant no parallel deference to the "expertise" of a trial judge. Questions of law, properly preserved, are fully open on appeal from the trial court. We do, of course, defer to a trial judge's findings of fact where he has had the opportunity to observe and evaluate witnesses. Mass. R. Civ. P. 52(a), 365 Mass. 816 (1974). But even in reviewing factual determinations, the "clearly erroneous" standard of review of a judge's findings (Mass. R. Civ. P. 52 [a]) and the "substantial evidence" standard of review of an agency determination of fact (G.L.c. 30A, § 1 [6]) are expressed differently and arguably are not the same.^[4] The "substantial evidence" test may call for narrower review than the "clearly erroneous" test. See K.C. Davis, Administrative Law of the Seventies § 29.00, at 646-647 (1976); L.L. Jaffe, Judicial Control of Administrative Action 615-616 (1965). Where agency factual determinations are mixed with questions of law, our review is likely to be narrower than that of similar trial court determinations. See, e.g., French v. Assessors of Boston, 383 Mass. 481, 482-483 (1981) (upholding Appellate Tax Board's determination that single family residences are a "class"). The parties have urged that we use the "clearly erroneous" test, and we shall do so. Distinctions between the two standards of review may be significant in subsequent cases. Our conclusion, however, that the judge's findings of fact were not clearly erroneous in determining the fair cash value of the shipyard, thus satisfying the arguably broader scope of review, makes any distinction unimportant in this case.

From what we have said it should be apparent that we look with disfavor on the Superior Court's exercising jurisdiction over cases of this character. The question of the appropriate method of valuing special purpose property for local taxation purposes is one for which the Appellate Tax Board is more appropriately suited, for reasons we have just explained. The city, however, does not argue the point before us. The case has been tried and a decision rendered. We, therefore, decline to say, on our own motion, that the decision to accept jurisdiction of this case, based on the circumstances at the time the decision was made, was an abuse of discretion. Thus we turn to the issues argued on appeal.

2. We consider first the city's argument that the trial judge had no alternative but to base his determination of the fair cash value of the shipyard on the depreciated replacement or reproduction cost (DRC) of the shipyard's buildings and other taxable property.^[5] In Correia v. New Bedford Redevelopment Auth., 375 Mass. 360, 362 (1978), we noted the three traditional ways that experts in the real estate field calculate fair market value: the market study method (i.e., the use of recent sales of comparable property), the income capitalization method (capitalizing net income), and the DRC method. The DRC method is disfavored because of the difficulty of adequately discounting for obsolescence as well as for physical deterioration. Id. at 364. Unlike the income capitalization method and the market study method, the DRC method does not directly incorporate the influences of the real estate market on the value of the property. In the Correia case, however, the other two methods did not yield an accurate valuation, and we held that, in the circumstances, the judge did not abuse his discretion in allowing the admission of DRC evidence. Id. at 367. Similarly, in Foxboro Assocs. v. Assessors of Foxborough, 385 Mass. 679, 687 (1982), we held that the Appellate Tax Board did not err in relying on DRC evidence in arriving at its determination of the fair cash value of a harness raceway.

In the case before us, it is agreed that neither the income capitalization method nor the market study method could reasonably be applied, and the judge so found. Thus the initial question is whether the judge had no alternative but to use a DRC method of valuation. In our Correia opinion, we did not treat these three methods as exclusive. Id. at 362-363. "We have recognized that courts, in permitting the use of valuation data other than those factors ordinarily bearing on market price, ``may be doing no more than recognizing that more complex and resourceful methods of ascertaining value must be used where the property is unusual or specialized in character and where ordinary methods will produce a miscarriage of justice.'" Id. at 363, quoting from Newton Girl Scout Council, Inc. v. Massachusetts Turnpike Auth., 335 Mass. 189, 195 (1956). We conclude that the judge was not compelled to use the DRC method, at least where, as we shall next show, his finding that such a method of valuation would be inappropriate was not clearly erroneous.

3. The judge's finding that the DRC of the property would not be an appropriate measure of the shipyard's fair cash value was not clearly erroneous. He found that the physical state of the shipyard was such that a purchaser "would not consider useful any calculus of cost to replace or reproduce any building."^[6] In Commonwealth v. Massachusetts Turnpike Auth., 352 Mass. 143, 149 (1967), we stated that "where special purpose structures are very greatly out of date, are no longer well fitted to their particular use, and would not be reproduced by any prudent owner ... evidence of adjusted reproduction cost will be irrelevant, for it is difficult, even for an expert, to estimate suitable allowances for physical depreciation and obsolescence of such an obsolete structure."

There was evidence that the layout and construction of the shipyard would not be reproduced in its present form. Some structures were designed for shipbuilding methods now abandoned. Parts of the shipyard were constructed during World Wars I and II when immediate rather than long-term benefits were desired. In addition, the judge's finding was not clearly erroneous that in the circumstances a prospective purchaser would not rely on the DRC of the yard but would only consider the realty's value as a shipyard.^[7]

We should not be understood to say that the judge would have been wrong, as a matter of law, if he had determined to use the DRC of the buildings at the shipyard and had made appropriate adjustments, largely judgmental, properly to reflect obsolescence. We simply conclude that his finding that the DRC method was inappropriate for use in these circumstances was not clearly erroneous. We come then to the question whether the judge committed any error of law in adopting the method of valuation he did use.

4. The judge was warranted on the evidence in using the method of valuation he employed and, in his decision to use that method, he committed no error of law. He found that "the hypothetical seller and purchaser would adopt the following evaluative method, which I further find to be the most credible way of determining the Shipyard's fair market value ...:

"First, ascertain the value of the business conducted on the entire Shipyard (Braintree and Quincy portions).

"Second, pursuant to the principle that income from a business conducted upon land depends on non-land factors, and therefore does not determine the value of the land, Amory v. Commonwealth, 321 Mass. 240, 258 (1947), calculate and subtract the value of the so-called ``intangibles', i.e., the elements that prescind from the physical assets and go to establish the Shipyard's market value as a business entity.

"Third, subtract the value of (a) the cranes ... and (b) other non-taxable equipment (located in Quincy and Braintree).

"Fourth, multiply the resultant by a decimal equivalent to the proportion which the Quincy-sited part of the Shipyard bears to the entire Shipyard area.

"22. The end product is a figure which represents the amount that a willing purchaser would pay a willing seller (each without compulsion to do business) for the taxable real estate assets located in Quincy."

The first step in this process was a determination of the value of the business of the entire shipyard. In doing this, the judge concluded that projected earnings of the shipyard over a reasonable period after January 1, 1978, must be estimated. Then, in a way analogous to "the concept employed when valuing property by the capitalization-of-income method, ... the rate of return which the capital market ... expects from capital invested in shipbuilding/ship repair enterprises" must be ascertained. Based on General Dynamics's projected earnings for 1978-1982, the judge found that average projected earnings of $13,581,000 was reasonable.^[8] He then capitalized the annual earnings by a rate that would reflect the market's assessment of the risk and future prospects of the enterprise, recognizing that such a rate, or its reciprocal, the price-earnings ratio, was not susceptible of mathematical precision. On all the evidence, he selected a price-earnings ratio of 5.9.

The judge then turned to the value of the intangible assets that must be deducted from the value of the shipyard as a going concern. These are factors of a business's value other than land, plant, and equipment, such as working capital, going-concern value, and the condition of the work force. One witness testified that the intangible assets represented 50% of the value of the business. In the opinion of another witness a reduction of 20%, without regard to other factors, was warranted simply because of the value of the work force. The judge rejected these specific opinions and credited testimony that "no conventional wisdom exists for making this type of necessarily judgmental allocation." He settled on a discount of 40%.

The judge then subtracted the value of the cranes, which he concluded were not subject to local taxation, and the value of other equipment not subject to local taxation. Finally, he made an allocation of the value of the property between Quincy and Braintree and arrived at a fair cash value of $25,918,801 for the Quincy land, buildings and other taxable property. This figure was reduced by 75% to $6,479,700, to reflect the disproportionate assessments in effect in Quincy during the relevant years.^[9]

The city argues that the judge erred, as a matter of law, in adopting the valuation method he used. This argument, going beyond the city's claim that the judge was obliged to use the DRC method of valuation, says that, in any event, General Dynamics cannot prevail because it advanced, and the judge accepted, a legally impermissible valuation method. We recently noted, in Foxboro Assocs. v. Assessors of Foxborough, 385 Mass. 679, 691 (1982), that "special purpose properties ... pose unusual problems of valuation and ... no particular method of valuation may be entirely satisfactory." The problem of valuing such property may have to be solved by "recognizing that more complex and resourceful methods of ascertaining value must be used ... where ordinary methods will produce a miscarriage of justice." Newton Girl Scout Council, Inc. v. Massachusetts Turnpike Auth., 335 Mass. 189, 195 (1956). "One must remember that whatever valuation method is used, the goal is to determine what a willing buyer would pay a willing seller for the property." Boston Edison Co. v. Assessors of Watertown, 387 Mass. 298, 305 (1982).

Where conventional methods of valuation are inapplicable or inappropriate in the circumstances, the task of valuation cannot be simply abandoned. See Boston Gas Co. v. Assessors of Boston, 334 Mass. 549, 580 (1956). The judge adopted a method of valuation of the shipyard that, according to expert testimony, was an accepted method used in the marketplace for valuing a going concern. Admittedly, choosing the capitalization rate and discounting the value of the intangible assets of the shipyard as a going concern involved judgment. Variations in the judge's conclusions about these elements would have changed the end result. But we cannot fairly say that the method used was unlawful or that the findings made in the process were clearly erroneous. The expert testimony and other evidence warranted those findings.^[10]

The city correctly points to decisions of this court that have stated that property should not be valued on the basis of the profitability of business operations conducted on the premises. See Piemonte v. New Boston Garden Corp., 377 Mass. 719, 732 (1979); Amory v. Commonwealth, 321 Mass. 240, 258 (1947). See also 5 Nichols, Eminent Domain § 19.3[1], at 19-49 to 19-55 (rev. 3d ed. 1981). In various special circumstances, however, we have recognized and have accepted the business income of property as a relevant consideration in arriving at fair cash value. See Boston Edison Co. v. Assessors of Watertown, 387 Mass. 298, 305 (1982); Boston Gas Co. v. Assessors of Boston, supra at 580; Assessors of Quincy v. Boston Consol. Gas Co., 309 Mass. 60, 67 (1941); Revere v. Revere Constr. Co., 285 Mass. 243, 248-250 (1934).^[11] Business income itself, however, cannot normally be considered as a controlling, or even relevant, factor because it is the property, and not the business, that must be valued.

The judge in this case, of course, started with a projection of the average annual business earnings of the shipyard over a five-year period. This method of valuation, however, used business earnings only in the interim step of arriving at a going-concern value for the shipyard. Applying the discount factor for intangible assets was his way of correcting for the valid objections to the use of business profits. Such an approach cannot always be justified, particularly if a taxpayer has no earnings at all. We conclude, however, that in this case the judge's approach was permissible.^[12]

The city challenges the use of the hypothetical income of the shipyard based on earnings available for distribution to owners of the enterprise. It claims that capitalizing the earnings figure measures only a fraction of the shipyard, the value of the equity interest, and it ignores the value of assets represented by the interests of creditors. This challenge was not presented below. General Dynamics's witnesses did not treat their method of valuation as designed to measure only the equity interest in the shipyard. One of General Dynamics's witnesses testified on cross-examination that he did not know the amount of General Dynamics's long-term debt and did not undertake to find out what it was. He apparently thought it unimportant to his conclusions. The city did not pursue the point with him.

If the point that the city now advances had any validity, it should have been raised at trial so that General Dynamics could rebut it. See Royal Indem. Co. v. Blakely, 372 Mass. 86, 88 (1977). The argument is basically that the experts' opinions should not be believed because they omitted a relevant matter for consideration. These experts stated the reasons for their conclusions about the correct method by which to value a business, and the judge was entitled to give them such weight as he did. On the record, the judge's findings in this respect are not clearly erroneous nor his legal conclusions wrong for the reason advanced by the city.

5. We come then to the city's specific challenges to the way in which the judge applied the method he selected to value the shipyard's taxable property.

The city argues that the judge erred in his finding that the income to be capitalized was after-tax income, and in his conclusion that income taxes would have a combined "effective rate" of 50%. These conclusions, although not compelled, were warranted on the evidence and were not clearly erroneous.

The city challenges the judge's conclusion that the going-concern value of the shipyard should be reduced by 40% to reflect the value of the intangible assets included in the going-concern value. Some enterprises may have no value attributable to intangibles, and no deduction would be warranted. This is likely to be true of a company on the verge of a seemingly unavoidable bankruptcy. But that determination is a question of fact on the record. Once one accepts the propriety of the use of the method of valuation used in this case, the factual determination can be upset only if it is clearly erroneous. On the record, this determination of fact was not.

The city argues that the 40% reduction for the value of intangible assets should have been applied to the value of the cranes and other equipment not subject to taxation (and thereby increasing the value of the taxable property). This argument lacks merit because it fails to recognize that the 40% reduction was designed to eliminate the value of intangible assets from the going-concern value, not to reduce the value of all assets.

The judge valued the tax-exempt cranes by the DRC method, and he valued other tax-exempt equipment on the basis of book values. The city questions these approaches to the value of the tax-exempt equipment because the judge rejected DRC as the method of valuing the shipyard's buildings. The city argues, in essence, that if the DRC method is appropriate for valuing the cranes, then it is appropriate for valuing all real estate improvements. The city does not, however, urge that the judge, having adopted the method he did, should have used some other method of determining the value of the tax-exempt property. Superficially, if use of DRC is not appropriate in arriving at the fair cash value of the shipyard's buildings, use of DRC for the cranes might also be inappropriate. But the city does not present any challenge to the value the judge placed on the cranes and other tax-exempt equipment. On the record, the valuations adopted by the judge were not clearly erroneous. The judge tended to accept evidence favorable to the city rather than evidence favorable to General Dynamics.

6. We come finally to the city's claim that General Dynamics's applications for abatement with respect to fiscal year 1977 were not seasonably filed with the city's assessors. A taxpayer has the right to apply for an abatement "on or before the thirtieth day after the date on which the bill or notice was so sent." G.L.c. 59, § 59, as amended through St. 1974, c. 831, § 4. Failure to file a timely application is a jurisdictional bar to the granting of an application for abatement. New Bedford Gas § Edison Light Co. v. Assessors of Dartmouth, 368 Mass. 745, 747-748 (1975).

The judge concluded that the applications were filed within the statutory period and, even if they were not, the city was estopped to deny the timeliness of the filings. This issue was presented on a stipulation as to facts, and, because we are in as good a position as the trial judge to pass on the matter, we do not consider his factual determination simply to see if his findings were clearly erroneous. We reach our own determination. We conclude that Quincy's 1977 fiscal year real estate tax bills were not mailed (i.e., sent) before November 1, 1976, and that the filing of the applications for abatement on November 30, 1976, was, therefore, timely.

We summarize the parties' stipulation. The city's treasurer-collector would have testified that all real estate tax bills for fiscal year 1977 were stamped on the collector's postage meter and mailed on Friday, October 29, 1976, and that he personally delivered them to the Quincy post office (presumably on that day). On November 1, 1976, the treasurer-collector posted a notice that the 1977 fiscal year tax bills "were mailed on or before November 1, 1976 and are due and payable on or before December 1, 1976." In the second week of November, a representative of General Dynamics collected approximately two dozen applications for abatement from the office of the clerk of the board of assessors. "[E]ach blank application had been pre-stamped by the Board of Assessors as ``Due December 1st.'"^[13] The assessors received 113 abatement applications on November 30, 1976, and 83 more on December 1, 1976. They granted tax abatements as to 68 of these parcels. The city did not collect interest from, nor levy interest charges on, numerous taxpayers who paid taxes on November 30 and December 1. There was no stipulation as to the date on which General Dynamics's tax bills, or any other tax bills, were processed by the Quincy post office.

From these facts, the inference can and should fairly be drawn that the city officials concerned with mailing the fiscal year 1977 real estate tax bills regarded November 1 as the relevant date. We give great weight to what those officials did shortly after the delivery of the tax bills to the post office for mailing. We will not attribute to them the intention of misleading taxpayers, of granting tax abatements on applications not seasonably filed, or of failing to collect nonwaivable interest on taxes not paid on time (see G.L.c. 59, § 57). Rather we construe their actions as admissions to the effect that the fiscal year 1977 real estate tax bills were sent, for the purposes of G.L.c. 59, § 59, on November 1. Because of our factual conclusion, we need not consider (a) whether the treasurer-collector's statements properly could be considered to contradict what General Dynamics characterizes as "official public records," (b) whether the city should be estopped to deny the timeliness of the abatement applications, or (c) whether the adverse consequences the city sought to apply only to General Dynamics violate General Dynamics's right to equal protection of the laws. We conclude that General Dynamics's abatement applications were filed within the statutory time limit.

Judgment affirmed.

NOTES

[1] The collector-treasurer of the city of Quincy. We shall refer to the defendants collectively as the city.

[2] Although General Dynamics challenged the assessments for the fiscal years 1977-1980, inclusive, the parties agreed that the fair cash value on January 1, 1978, would be treated as the proper fair cash value for each year. Therefore, the testimony was directed to values on January 1, 1978. Although the city assessed twenty-five separate parcels, the parties stipulated that only one valuation was needed of the entire taxable property.

It was agreed that 72.95% of the shipyard is in Quincy, with the balance in Braintree, and that whatever value was determined for the property should recognize that fact. The shipyard occupies 5,705,717 square feet of land in Quincy and 2,115,278 square feet of land in Braintree.

The parties also agreed that the ratio of assessments to valuation in Quincy was 1:4 and that the fair cash value would have to be reduced to reflect the disproportionate assessment practices of the city. Thus the proper assessed value would be 25% of the fair cash value.

One dispute at trial concerned whether cranes were subject to local taxation. The judge ruled that they were not, and the city has not appealed that determination.

[3] The city's assessment for the applicable tax years was based on a fair cash value of the shipyard's Quincy property, including the disputed cranes, of approximately $82,900,000. The city's evidence, based on a combination of depreciated reproduction and replacement costs, indicated a fair cash value of $103,085,200 or, excluding the disputed cranes, a fair cash value of $82,161,350. Although General Dynamics disagreed with the use of depreciated reproduction or replacement costs to value the shipyard, it presented rebuttal evidence that the shipyard's value using that method (excluding the cranes) was $33,879,262.

General Dynamics presented testimony from two experts who valued the shipyard by capitalizing its earnings, and reducing that figure to eliminate the value of intangible assets and nontaxable personal property and adjusting it to reflect only the property in Quincy. One expert witness arrived at a fair cash value of $16,114,000. The other expert witness concluded that the fair cash value of taxable property in Quincy was $15,325,000. The city presented no direct testimony bearing on the appropriate factors to consider if this method of valuation were to be used.

The judge, following the method advocated by General Dynamics but reaching different initial conclusions, arrived at a fair market value of General Dynamics's taxable property in Quincy on January 1, 1978, of $25,918,801.

[4] Substantial evidence is "such evidence as a reasonable mind might accept as adequate to support a conclusion." G.L.c. 30A, § 1 (6), inserted by St. 1954, c. 681, § 1. In deciding whether substantial evidence supports an agency determination, the court "upon consideration of the entire record" (G.L.c. 30A, § 14 [7], as appearing in St. 1973, c. 1114, § 3) must determine the substantiality of evidence by taking into account whatever in the record fairly detracts from its weight. See Cohen v. Board of Registration in Pharmacy, 350 Mass. 246, 253 (1966), relying on Universal Camera Corp. v. NLRB, 340 U.S. 474, 488 (1951), construing similar language in the Federal Administrative Procedure Act.

We have accepted the interpretation of the standard of "clearly erroneous" in Mass. R. Civ. P. 52 (a), adopted under the parallel Federal rule. See Freyermuth v. Lutfy, 376 Mass. 612, 615 (1978). A finding of fact is "clearly erroneous" and should be set aside "when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed." Id., quoting from United States v. United States Gypsum Co., 333 U.S. 364, 395 (1948).

[5] The depreciated reproduction cost method estimates the cost, less depreciation, of exactly duplicating the building based on current costs. The depreciated replacement cost method estimates the current cost, less depreciation, of constructing a building of equivalent utility. See Boston Edison Co. v. Assessors of Watertown, 387 Mass. 298, 300 n. 2 (1982).

[6] The judge made further general findings bearing on the appropriateness of what we have called the three traditional methods of valuation. "16. On the basis of the evidence introduced at the trial, whether oral or written, I find that none of these valuation methods alone — however useful any or all of them might be in valuing other realty — is appropriate for deriving the fair market value of the Shipyard. That is, no purchaser of the Shipyard realty would use any of these methods alone to establish the price he ought to pay.... [N]or would he regard as feasible any calculation of the buildings' respective rental values. He would not use the comparable-sales method because he would be interested only in the value of the realty qua shipyard, and no comparable sales of a shipyard are available to consider.

"....

"18. I find that each of the ``normal' valuation methods is inapplicable to the Shipyard because the Shipyard is a property of such unusual and specialized character that none of the methods would fairly establish its value. In doing so, I have in mind that the theoretical purchaser is likely to be a corporation similar to [General Dynamics], and that the number of such buyers would be small, see Boston Gas Co. v. Assessors of Boston, 334 Mass. 549, 572, 584 (1956), and that, as I have earlier indicated, a purchaser would only consider the realty's value as a shipyard.

"19. The basic flaw common to each of the ``normal' methods is its focus on the individual building (including the land on which it stands), rather than on the premises' symbiotic relation to all the other Shipyard realty. Put another way, none of the ``normal' methods sufficiently recognizes that it is the Shipyard which the City seeks to value, not the individual parcels. Because, as the parties have properly stipulated, the highest and best use of the entire land area is as a shipyard, one must evaluate every building only in that context. I do not accept as credible, testimony suggesting the possibility of valuing the Shipyard by considering the hypothetical rent derivable from a hypothetical lease of the entire Shipyard."

[7] It is, of course, possible that, in different circumstances, the depreciated replacement cost of a shipyard would be most relevant in determining what a willing buyer would pay a willing seller for the shipyard. If that cost were less than the value of the property as a shipyard, a buyer might reasonably conclude to construct its own shipyard.

[8] The judge stated: "Bearing in mind that a prospective purchaser of the Shipyard's realty would consider the investment only if he could profitably operate a shipbuilding/ship repair business on the premises, it is apparent that such a purchaser would first consider a reasonable projection of the business' earnings for several years after the pertinent date (January 1, 1978). He would average out all these figures to produce an ``annual' earnings figure, or, more accurately, a reasonable prediction of subsequent earnings."

[9] This description of the process used by the judge is a general one and does not include various interim calculations that are not challenged in this appeal.

[10] The city offered no evidence from experts of its own that would have supported the use of either a capitalization rate or a discount for intangible assets that was more favorable to it. It offered no evidence directly challenging the use of the valuation method advocated by General Dynamics and adopted by the judge. In general, the city's experts were told to use the DRC method and did so. The city presented virtually no evidence that the DRC was conceptually preferable to the valuation method used. One of its experts testified without any explanation that the DRC method was the appropriate method, but the judge was justified in rejecting that summary conclusion.

[11] In Northampton Nursing Home, Inc. v. Assessors of Northampton, 383 Mass. 884 (1981), we summarily affirmed a decision of the Appellate Tax Board which determined the fair cash value of a nursing home by capitalizing its gross income, reduced by operating expenses and "income attributable to personal property, management, etc." The board rejected the DRC method of valuation in preference to the method it used. It did so because of "the nature of the nursing home business" and because the use of income from the taxpayer's business was "an appropriate method of valuation in certain instances where special use property is involved, (Encyclopedia of Real Estate Appraising, [3d ed.], Prentice-Hall, pp. 1005-1022 [1978])." The assessors' challenge in that appeal was limited to claiming that (a) the board was required as a matter of law to use the home's DRC, and (b) earnings could be considered only if nursing homes were rented or leased.

[12] We recognize that the valuation method that the judge used presents practical problems. Because of fluctuations in earnings, the method may not be appropriate in other circumstances. Also, the considerations and calculations involved are not common to the usual valuation practices that assessors use. The cost of obtaining experts and of assembling data can be considerable, but the same may be said of the DRC method of valuation as well. Data essential to this method of valuation are largely within the taxpayer's control and are not immediately available to assessors. But that information must be furnished in response to a reasonable request from the assessors. See G.L.c. 59, § 38D. In the case of special purpose property, and even in the valuation of property based on net rental income, such problems are often unavoidable. The requirement that the fair cash value of property be determined by proper means must prevail over practical considerations that would favor the use of an inappropriate but administratively convenient method of valuation.

[13] Although the parties stipulated that the abatement application forms had stamped on them "Due December 1st," the sample form annexed to the stipulation has, by stamp, "December, 1, 1976" placed over the date "October 1, [blank]" in the phrase "Must be filed with the Board of Assessors on or before October 1, [blank]."