Nolan, J.

This case represents a test of certain procedures taken pursuant to the tax classification amendment, art. 112 of the Amendments to the Massachusetts Constitution, and its enabling legislation, St. 1979, c. 797. At issue is the propriety of the Superior Court judge’s findings and rulings regarding (1) certification by the Commissioner of Revenue (Commissioner) that the city of Fitchburg (city) was qualified to implement differential taxation of property by use classification for fiscal years 1981 and 1982; (2) implementation by the city of that program; and (3) failure of the city in 1981 to apply free cash available at the close of fiscal 1980 to reduce its property tax levy. We affirm the judgment of the Superior Court with the one exception noted in part 5, infra.

1. Overview. A brief description of the property tax system, and especially classification, will prove useful in understanding our discussion of the issues in this case. 3 In recent years this court has upheld the constitutional mandate that the Legislature impose “proportional and reasonable assessments, rates and taxes.” Part II, c. 1, § 1, art. 4, of the Massachusetts Constitution. See Sudbury v. Commissioner of Corps. & Taxation, 366 Mass. 558, 563 (1974); Coomey v. Assessors of Sandwich, 367 Mass. 836, 837 (1975). Shortly after this court’s decision in Bettigole v. Assessors of Springfield, 343 Mass. 223 (1961), in which the Constitution of the Commonwealth was construed to require equality in tax rates on all classes of property, the Legislature began preparation of an amendment allowing for classification, or different tax rates for different classes of property. *755 Ultimately, in 1978, the voters ratified a classification amendment. Art. 112 of the Amendments to the Massachusetts Constitution. Legislation was enacted by St. 1978, c. 580, § 38, which inserted c. 59A (Classification Act). Statute 1979, c. 797 (Enabling Act), repealed and amended c. 580, § 38, with the exception of c. 59A. 4 See generally Opinion of the Justices, 378 Mass. 802 (1979).

The amended Classification Act permits municipalities to determine, within limits, the tax burden for four classes of property. 5 Before classification of real property is implemented, the Commissioner is required to certify that “the assessments on the real property . . . are at full and fair cash valuation.” G. L. c. 59, § 2A(c). Once the Commissioner determines that a community is assessing property at full value, she “shall determine a minimum residential factor for each city and town.” G. L. c. 58, § 1A. This factor, to be not less than 65 %, nor more than 100 %, must prevent the tax burden on any other class of property from being increased by more than 50 %. The community may adopt a residential factor, not less than the minimum residential factor prescribed by the Commissioner. G. L. c. 40, § 56. At the same time, the board of assessors of the community must determine the percentages of the local tax levy to be borne by each class of real property. G. L. c. 40, § 56.

While the responsibility for assessments and valuation is ordinarily committed to local assessors, the Commissioner has substantial authority and powers with respect to these matters. G. L. c. 58, § 1A. Assessors must follow her lawful commands. See Sudbury v. Commissioner of Corps. & Taxation, 366 Mass. 558, 563-564 (1974); Commonwealth v. Andover, 378 Mass. 370, 377-378 (1979). She is required to administer all laws which the Department of Revenue must enforce. G. L. c. 14, § 3. She may further the attainment of the goal of uniformity in “valuation, classification *756 and assessment” throughout the Commonwealth, G. L. c. 58, §§ 1A, 4, by issuing guidelines to assist local assessors in the performance of their duties. G. L. c. 58, §§ 1, 3. To assist her in carrying out her responsibilities, the Commissioner has a staff which the Superior Court judge found to be “without the resources to accomplish its many functions in the manner and to the degree of accuracy which . . . [is] desirable.”

2. Factual background — certification of Fitchburg. After St. 1979, c. 797, was enacted, the Commissioner issued a set of guidelines. 6 Pursuant to these guidelines, the Commissioner conducted a certification review of Fitch-burg. The review consisted of a statistical analysis of residential and land parcel values. 7 The Commissioner derived a median assessment/sales ratio 8 for some such parcels and a coefficient of dispersion (COD) about that median. 9

*757 The Commissioner’s study derived ratios and COD’s for only two classes of property. For Rl (single family residential) 10 she found a median ratio of 80% and a COD of 11.2%. For land she found a median ratio of 55% and a COD of 63.7%. Determining that the data submitted for her land study was inadequate, the Commissioner discarded those results and relied on Rl statistics alone.

In order to raise assessment for Rl and other parcels to acceptable levels, the Commissioner developed a factor to be applied to Rl and other parcels. 11 The factor was developed by taking the mean ratio for Rl parcels 12 derived from the *758 1980 equalized value study (EQV study) 13 and finding a multiplier that would increase that mean ratio to 100 % . 14 Using the mean method from the EQV study for R1 properties, the Commissioner calculated the assessment/sales ratio as 89.6% (as distinguished from the 80% median ratio derived from the nine-month study). The Commissioner chose a multiplier (factor) of 1.12 to raise the EQV mean for R1 properties to 100%. 15 Fitchburg chose a system of factors differing from that prescribed by the Commissioner. Having separately assessed values for land and buildings, the city applied a 1.10 factor for R1 land and a 1.13 factor for R1 structures in Fitchburg. The R2, R3, and R4 properties were factored at 1.06. 16 Commercial and industrial properties were factored at 1.00 and land was factored at 1.10. The Commissioner accepted Fitchburg’s factoring program and certified that the city could implement tax classification.

At the conclusion of fiscal 1980, Fitchburg had free cash in the amount of $1,610,330. Had this money been applied to 1981 taxes, such taxes would have been reduced by $7.28/$l,000, from $73.20/$l,000 to $65.92/$l,000. 17 Of *759 the free cash available, Fitchburg appropriated and expended $1,222'479 to defray ordinary expenses during fiscal 1981.

3. Procedural background. This case, here on cross appeals from a judgment in the Superior Court, is the result of a ten-taxpayer action seeking injunctive relief under G. L. c. 40, § 53, and declaratory relief under G. L. c. 231A, which was consolidated with ten separate taxpayer actions for the recovery of back taxes under G. L. c. 60, § 98. The plaintiffs (taxpayers) are taxpayers in the city of Fitchburg. The defendants are the Commissioner and various officials of Fitchburg (city defendants).

This case was preceded by Litton Business Syss., Inc. v. Commissioner of Revenue, 383 Mass. 619 (1981). Litton presented essentially the same question as the instant case, but was dismissed for want of subject matter jurisdiction. Id. at 623-624. During the pendency of the Litton appeal in this court, the ten actions under G. L. c. 60, § 98, were filed. Four days after Litton was remanded, the instant action for injunctive and declaratory relief was filed in the county court. The case was transferred to this court pursuant to G. L. c. 211, § 4A. Injunctive relief was denied and the case was then transferred to the Superior Court for further proceedings.

The consolidated actions were tried between September 15, and September 29, 1981. Shortly thereafter, a Superior Court judge entered an order for partial summary judgment. Mass. R. Civ. P. 54 (b), 365 Mass. 820 (1974).

The judge denied the taxpayers’ motion for injunctive relief by which they sought to restrain the issuance of classified real property tax bills in fiscal 1982. The judge noted that, although there was a reasonable likelihood that the plaintiffs would be able to prove various errors, inconsistencies and at least one flaw in the tax program as adopted *760 by Fitchburg and approved by the Commissioner, they failed to show that they were threatened by “sufficient irreparable” harm to warrant interference with the city’s collection of taxes. The judge ruled further that on “the present record . . . the Plaintiffs are not entitled to any relief under either G. L. c. 40, Section 53 or G. L. c. 60, Section 98 as to the tax classification aspects of the case,” and thus he denied relief on these claims. The order did grant relief to the taxpayers on the issue of the city’s failure to apply its free cash to abate tax assessments, but the judge deferred to the Appellate Tax Board to ascertain the limits of its own jurisdiction on this issue. The judge retained jurisdiction in the event that the Appellate Tax Board decided that it had no jurisdiction over this matter.

In November, 1981, the judge issued his memorandum of decision in which he fleshed out the reasoning behind his order and, in addition, he made several declarations pursuant to G. L. c. 231A. Appeals and cross appeals were filed, and this court granted a motion brought jointly by the city and the Commissioner for direct appellate review.

In brief, the taxpayers argue that the Superior Court judge erred by (1) not invalidating the implementation of classification for 1981 and 1982, after having found that the city was not assessing property as of January, 1980, at its full and fair cash value; (2) upholding the Commissioner’s 1981 guidelines for the implementation of classification; (3) concluding that the Commissioner’s guidelines had not been violated and that her methodology adequately tested whether the city was assessing at full and fair value; (4) ruling that illegalities in the certification of the city’s plan were excused by time or resource limitations, or by a need for a uniform methodology; (5) concluding that the plaintiffs were not injured and by denying the declaratory, injunctive, and monetary relief the plaintiffs sought and; (6) concluding that G. L. c. 60, § 98, could not be utilized by the taxpayers to seek relief in the free cash matter absent an appeal to, and a denial of jurisdiction by, the Appellate Tax Board.

The Commissioner, in addition to seeking to uphold the conclusion that the guidelines were valid and that the city *761 was properly certified, seeks to have this court reject certain findings made in the Superior Court.

The city seeks to uphold the judgment denying injunctive relief, but argues that the judge’s rulings with respect to the free cash issue were erroneous.

4. Guidelines on face and as applied. Under our present property tax system, property within each class must be assessed proportionately and uniformly, and at full and fair cash value. G. L. c. 59, § 2A (c). See Opinion of the Justices, 378 Mass. 802, 806 (1979); Commonwealth v. Andover, 378 Mass. 370, 372 (1979). We note at the outset that full and fair cash values can only be approximated. Sudbury v. Commissioner of Corps. & Taxation, 366 Mass. 558, 562 (1974). Cf. Roberts v. State Tax Comm’n, 360 Mass. 724, 728 (1972). See also Newton v. Commissioner of Revenue, 384 Mass. 115, 121-122 (1981). The taxpayers argue that the Commissioner’s guidelines 18 are not even reasonably calculated to achieve full and fair cash valuation and therefore it was error for the judge to fail to declare them void on their face. We disagree.

The Commissioner is authorized to prepare guidelines to assist assessors in the performance of their duties. G. L. c. 58, § 3. The guidelines, as issued, are subject to varying interpretations. 19 In short, it is only as to residential property that assessors are given any real guidelines. 20 The guideline promulgated for land is open to various interpretations. 21 Further, there are no guidelines for testing uni *762 formity of assessment for commercial or industrial property or for real property as a whole. That the guidelines did not specifically speak to all real property is of little importance, for the Commissioner’s staff was instructed to verify the accuracy and uniformity of assessments of industrial and commercial properties by other means. 22 The judge recognized, however, that the guidelines, “on their face, permit an unconstitutional degree of underevaluation of specific classes of property.” Under the Commissioner’s interpretation of the guidelines, that the variation between residential and land ratios must be less than 20 %, they appear to permit the median ratio for land to be at 72% of full and fair cash value. (90 % less 20 % of 90 % =72%). If the median ratio for residences were 100%, these guidelines would permit the median ratio for land to be 80 %. Such a variation is contrary to our Constitution. See Leto v. Assessors of Wilmington, 348 Mass. 144, 150 (1964).

Reasoning that “the power to tax is . . . the power to destroy,” the judge applied to the guidelines a heightened level of scrutiny akin to that adopted by this court in reviewing health and safety regulations. See, e.g., West Broadway Task Force, Inc. v. Commissioner of the Dept. of Community Affairs, 363 Mass. 745, 752 (1973). The Commissioner argues that these guidelines, being “almost experimental” in nature, deserve judicial deference. Cf. American Hoechest Corp. v. Department of Pub. Utils., 379 Mass. 408, 412-413 (1980). We need not enter into this debate, as our conclusion does not depend upon its resolution.

Applying a strict level of review, the Superior Court judge determined that the guidelines pass muster. He reasoned that since the guidelines do not prescribe improper assessment, he would have to turn to what the Commissioner actually did pursuant to her guidelines. That is, if she required full and fair cash valuation pursuant to a ra *763 tional scheme before certifying Fitchburg, the “plaintiffs cannot be heard to complain that the guidelines, as written, could conceivably allow improper assessment practices.” The Commissioner, through her guidelines, informed the city of assessment limits which were acceptable to her. These limits were illegal as to the land classification. This declaration is not, however, dispositive of this case. Since the disputed guidelines were merely an aid to cities and to the Commissioner to facilitate assessment at full and fair cash value, that they may have allowed for illegal results does not end the matter. Rather, we must turn to the validity of the Commissioner’s certification procedures, as applied, to determine whether the plaintiffs in this case were injured.

We turn first to the question whether the Commissioner was legally bound to adhere to her own guidelines. We note that these guidelines were promulgated to inform communities and their residents as to how the Commissioner viewed compliance with certain constitutional and statutory requisites. As such, we agree with the Superior Court judge that “the unexplained deviation from the guidelines in the case of Fitchburg may undermine public confidence in the [Department of Revenue’s] integrity and impartiality. See United States v. Leahey, 434 F.2d 7, 10 (1st Cir. 1970).” Confidence in government is involved here, and we thus hold that, in these circumstances, the Commissioner had a duty to conform to her guidelines, else the spectre of different assessment values in different communities becomes all too real.

The plaintiffs point to three principal departures from the guidelines: (1) the decision to certify Fitchburg was based on an assessment/sales ratio study using sales over the first nine months of 1979; 23 (2) median assessment/sales ratios and COD’s were calculated only for R1 and for land; and (3) the failure to conduct appraisals of commercial and industrial properties.

*764 The nine-month period was chosen for Fitchburg because the city had recently been reviewed for certification under the predecessor to the Enabling Act. Given the fact that the Commissioner had the entire Commonwealth to review, her judgment that those communities most likely to be certified should be subjected only to a nine-month study was neither arbitrary nor capricious. We leave that decision to the discretion of the Commissioner. See Newton v. Commissioner of Revenue, 384 Mass. 115, 121-122 (1981). In this case her choice of a nine-month study for Fitchburg was rational and we uphold it.

The second shortcoming that the taxpayers discern is the limitation of the study to Rl parcels and to land. 24 As to the residential class, the taxpayers argue that the Commissioner’s guidelines called for an evaluation of all residential sales — that is, R2, R3, and R4, in addition to Rl. The Commissioner argues that garnering information solely as to Rl parcels was not a departure from her guidelines, but, rather, was justified by two exigencies. First, as Rl parcels represent approximately 50% of all parcels in Fitchburg, there were very few sales in the other R subclasses. Consequently, data generated from R2, R3, and R4 studies would be unreliable. Second, Rl parcels are the easiest to appraise. Thus, she continues, performing a study of Rl parcels only was within her discretion. See Newton v. Commissioner of Revenue, supra at 121-122.

The Superior Court judge found, however, that this was not valid reasoning for omitting R2, R3, and R4 parcels, because her guidelines clearly stated that she was going to calculate a median ratio and COD for residential property as a whole. She clearly should have included R2, R3, and R4 data in her study, regardless of the low numbers of those parcels involved.

*765 We agree. But we also agree with the judge’s conclusion that while the Commissioner’s reliance on R1 as the measurement for all R parcels was wrong, she made a good faith effort, given the resources and time available to her, to develop a reasonable sales/assessment ratio and COD for all R parcels. We note, however, that in the future the Commissioner should make a reasonable effort to develop data for all classes of property.

Finally, the taxpayers argue that the Commissioner disregarded her guidelines by failing to appraise commercial and industrial property. Once again the Commissioner did deviate from her guidelines. She did, however, have data as to these two classes of property from an earlier study. We agree with the Superior Court judge that, given her lack of resources and time, the Commissioner was not overstepping legal bounds by considering these earlier studies in determining whether “general assessing practices in Fitchburg were such that she might certify the city for tax classification.”

Thus, while it is clear that the Commissioner failed to adhere to her guidelines, exigent circumstances and her efforts, attempted in good faith, exculpate her. 25 Her conclusion was that Fitchburg was not assessing at full and fair cash value, but that the city could be certified to implement classification if it employed an approved factoring program.

Factoring, as a method to increase a community’s assessment/sales ratio, is not attacked in this case. The taxpayers argue, however, that the factoring program employed by the city and approved by the Commissioner was illegal.

*766 Rather than using the data generated by her nine-month study, the Commissioner turned to the EQV study 26 to generate the factor. The taxpayers argue that the EQV study was conducted, at least as far as Fitchburg is concerned, “in an exceedingly cavalier manner.” The Superior Court judge analyzed the Commissioner’s methodology and found that the 1980 EQV study was “accomplished in a sloppy and irresponsible manner.” 27 Furthermore, mean, not median, assessment/sales ratios were utilized.

Nevertheless, the Commissioner argues that reliance on the 1980 EQV study to derive the factor was appropriate. She points to this court’s approval of the use of equalized value as a mechanism for testing a community’s full and fair cash value for purposes of Proposition 2Vz. Newton v. Commissioner of Revenue, supra. She argues, further, that the 1980 EQV study was the best information available to her and that she was, therefore, entitled to reasonable reliance on it. We disagree. A study which is “sloppy and irresponsible” has not been carried out “in absolute good faith and to the best of the abilities of the public officers charged with making valuations.” Bettigole v. Assessors of Springfield, 343 Mass. 223, 231 (1961). This study, resulting as it did in assessment of property in Fitchburg at less than full and fair cash value, should not have been relied upon by the Commissioner. 28

*767 The Commissioner derived a factor for Fitchburg. Fitch-burg chose to use a more complex factoring system. This program was accepted by the Commissioner and she thus certified that the city might implement tax classification.

The Superior Court judge ruled “that the factoring program, as implemented by the Assessors and approved by the Commissioner, is improper and must be declared so.” We have examined the record, and find no reason to disturb the judge’s finding that the factoring process actually used by Fitchburg “must be declared illegal.”

5. Declaration of rights. We adopt the Superior Court judge’s declaration of rights with one exception. We hold that the Commissioner was not entitled to rely on the 1980 EQV study for Fitchburg. 29 As to the remainder of the judge’s declaration we quote: “the court declares that the procedure actually followed by the Commissioner and the Assessors implementing the factoring program for the city of Fitchburg was illegal and improper in that it did not cause each sub-class of property in that city to be factored to 100% of full and fair cash value. Moreover, the court declares that assessments in Fitchburg were not, in fact, at full and fair cash value as at January 1, 1981 and that, if in the future such data comes to the attention of the Commissioner in a timely manner, she cannot simply ignore it in the discharge of her functions. Finally, the court declares that it may properly question assessment procedures adopted by the Commissioner which allow for a coefficient of dispersion in excess of 10% for Rl parcels, which rely on assess *768 ment studies covering periods more than two years and six months prior to the assessment date for which the studies are performed, any procedures which permit factoring of assessment data that has previously been twice factored, and any procedures limited to evaluation of the R1 subclass standing alone so long as the language of the Bureau of Local Assessment Technical Information Release 80-403, par. 3 continues to read ‘residential’ and not ‘single family residential.’” 30

Thus, we rule that the Commissioner did not properly certify Fitchburg for classification. Foreseéing this result, the taxpayers claim that they are entitled to an injunction under G. L. c. 40, § 53, against issuance of fiscal 1982’s second-half tax bills on a classified basis. The purpose of § 53 is to restrain a community from raising money by means not authorized by law. See Dealtry v. Selectmen of Watertown, 279 Mass. 22, 27 (1932).

As an initial matter, the city argues that equitable interference with the collection of the taxes under discussion is not timely. The city would have us treat the property taxes assessed for the fiscal year as a single equity. It argues that the entire tax is due and payable when committed to the tax collector. G. L. c. 60, § 3A. G. L. c. 59, § 57. Cf. Boston v. Du Wors, 340 Mass. 402, 403-404 (1960). Therefore, the city continues, Fitchburg was “about to raise . . . money” (G. L. c. 40, § 53) only before the annual tax was committed to the tax collector and therefore the court may not enjoin the 1982 second-half tax bills. We agree.

Assessors establish a single assessment list (G. L. c. 59, §§ 11, 21, 38), assess a single annual tax (G. L. c. 59, § 23), and commit a single tax list and warrant to the collector (G. L. c. 59, § 23). The entire tax is due and payable when committed to the tax collector. Boston v. Du Wors, supra at 402, 403.

*769 A suit in equity will not lie to restrain a tax collector from proceeding to collect his tax. Warr v. Collector of Taxes of Taunton, 234 Mass. 279, 282-283 (1920). California Village Corp. v. East Longmeadow, 4 Mass. App. Ct. 128 (1976). In Gallo v. Division of Water Pollution Control, 374 Mass. 278, 287-288 (1978), this court observed that the “[plaintiff] may have been able to challenge the validity of the assessment in a suit for declaratory relief brought prior to the commitment of the assessment to the tax collector. . . . However, it appears . . . that the assessment has been committed to the tax collector for collection. Therefore, [plaintiffs’] only avenue of relief lies with the administrative remedies.”

We hold that relief is not available pursuant to G. L. c. 40, § 53, once the assessment is committed to the tax collector pursuant to G. L. c. 59, § 53. 31 Cf. Bettigole v. Assessors of Springfield, 343 Mass. 223 (1961) (enforcement of levy enjoined, but only prior to commitment of tax list to collector). See Boston v. Second Realty Corp., 9 Mass. App. Ct. 282, 284 n.4 (1980). A community is about to raise money only before the tax list is committed to the tax collector. The taxpayers may not now obtain relief pursuant to G. L. c. 40, § 53.

The taxpayers next claim that they are entitled to recovery of back taxes under G. L. c. 60, § 98, in the amount of the difference between the amounts they paid, for fiscal 1981 and 1982, and the amounts they would have paid had classification not been implemented. The Superior Court judge denied relief to the taxpayers pursuant to § 98, as to the tax classification aspect of this case because, according to his analysis, the taxpayers incurred no injury; that is, they did not pay higher taxes than they otherwise would *770 have because of the impropriety of the method of factoring utilized. 32

The taxpayers ably argue that to reach his conclusion, the judge must have utilized figures from the 1980 EQV study showing industrial and commercial property mean assessment/sales ratios at 100% . However, the judge found the “actual” mean ratio under the 1980 EQV study to be 118.3% for commercial property and 117.2% for industrial property. As such, it appears that the taxpayers may have been injured by the factoring system as adopted. Notwithstanding an apparent injury, however, the taxpayers cannot avail themselves of relief under § 98 for errors in assessment. We have recently made it clear that “an action under c. 60, § 98, cannot be maintained unless the tax is wholly void.” Sears, Roebuck & Co. v. Somerville, 363 Mass. 756, 757-758 (1973). That there were irregularities and even illegalities in the tax assessments at issue here does not render Fitchburg’s tax void. For this court to declare a tax to be void, a community would have had to engage in a “widespread scheme of intentional discrimination.” Id. at 759. The Superior Court judge declined to find the city’s tax “wholly void,” and our reading of the record supports his conclusion. For any injury which the taxpayers may have suffered due to improper assessments, they are limited to administrative relief. See G. L. c. 59, § 59.

6. Free cash. Statute 1979, c. 151, § 12A, required that in fiscal years 1980 and 1981, if a community had any free (extra) cash available at the closing of a fiscal year such money should be used as a continuing appropriation to reduce the property tax levy for the next year. The taxpayers alleged that Fitchburg had free cash in the amount of $1,610,330 at the close of fiscal 1980 which, if properly applied, would have reduced the taxpayers’ tax rate. The city *771 admitted the existence of the free cash but disputed the amount that the taxpayers claimed their taxes would have been reduced. The judge found that St. 1979, c. 151, § 12A, had been violated.

The taxpayers seek to recover their overpayment for back taxes, under G. L. c. 60, § 98, due them because of the free cash that Fitchburg did not properly apply. The city argues that relief pursuant to § 98 is available only if the entire tax in question is “wholly void.” Sears Roebuck & Co.v. Somerville, supra. The judge declared that the free cash “ought have been applied in accordance with the statutory command.” We have already stated our approval of the judge’s conclusion. Litton Business Syss., Inc. v. Commissioner of Revenue, 383 Mass. 619, 624 (1981). The judge declared that the taxpayers were entitled to relief pursuant to the abatement procedures in G. L. c. 59, § 59, if they had properly pursued that remedy. The judge alluded, further, to a question whether the Appellate Tax Board has jurisdiction over the free cash issue. He committed that question to the board but retained jurisdiction of the free cash issue if the board ruled that it did not have jurisdiction. 33

7. Conclusion. The taxpayers’motion for injunctive relief is denied. The taxpayers are not entitled to relief in these proceedings under G. L. c. 40, § 53, or G. L. c. 60, § 98, as to the tax classification aspect of the case. As to the free cash issue, we adopt the judgment of the Superior Court, and commit that question to the Appellate Tax Board in conformity with the judgment of the Superior Court. The judgment is to be modified to conform the declaration of rights to part 5 of this opinion. In all other respects the judgment is affirmed.

So ordered.

Notes

See generally Hines, 1979 Classification Legislation, 24 B.B.J. (June, 1980).

This chapter was repealed by St. 1980, c. 261, § 16.

The four classes defined in G. L. c. 59, § 2A (b), are residential, open space, commercial, and industrial.

The guidelines, used in this case but no longer in use, may be found in Bureau of Local Assessment Technical Information Release No. 80-401. In pertinent part, the guidelines stated: “The certification review will consist of a statistical analysis of arm’s-length residential and land sales to determine the median ratio between assessed and market value and the variation from the median ratio in these classes. A median ratio within ten percent of full value and an average variation from such ratio of no more than fifteen percent is acceptable for residential property. The variation between residential and land ratios must be less than twenty percent. Additionally, the bureau will conduct appraisals of representative commercial and industrial properties for comparison with their assessed values to verify the accuracy and uniformity in these classes. Finally, the Bureau will review the revaluation program and general assessment practices to determine the reliability of the valuation base.”

The analysis covered the period of January through September, 1979. These nine months were chosen in conformity with the guidelines which used that period for communities which had been reviewed for certification under St. 1978, c. 580. Fitchburg had been so reviewed.

An “assessment/sales ratio” is a comparison of the assessed value of a parcel of property to its sales price, computed by dividing the assessed valuation by the sales price and expressing the quotient as a percentage. For example, the assessment/sales ratio of a parcel valued at $80,000 and selling for $100,000 is 80,000/100,000 or 80%.

The median is a measure of central tendency of a group of assessment/ sales ratios. It is the middle ratio of the group when that group is arranged *757 in order of magnitude. For example, in the group 10%, 20%, 80%, 90 %, 100 %, the median is 80 %.

The coefficient of dispersion (COD) is the average variation from the median ratio, expressed as a percentage of that median. See Oldman & Schoettle, State and Local Taxes and Finance 265-266 (Foundation Press 1974). The COD is a measure of uniformity within a group of ratios. The lower the COD, the more uniform are the assessments, as it measures the average (mean) deviation of the ratios of the group from the median. The COD may be expressed as follows:

COD = Sum of absolute values of differences between each ratio and median -s- number of ratios in group -r median

The Superior Court judge gave the following example to elucidate this matter: “By way of example, consider the following series of nine assessment/sales ratios: 70%, 70%, 100%, 100%, 100%, 100%, 115%, 130%, 130%. The coefficient of dispersion of the group, as defined by the Commissioner, is 15%, computed as follows: (1) The median ratio is 100%. (2) The sum of the absolute values of the differences of the individual ratios from the median is 135. (3) There are nine ratios in the group. (4) Thus, the coefficient of dispersion is (135/9/100, which equals .15, or 15%According to the guidelines, a median assessment/sales ratio within ten percent of full value (i.e. 90% through 110%) and a COD of 15 % or less were considered acceptable for residential property.

The Commissioner, for purposes of her evaluation, subdivided the residential class into numerous subclasses: Rl (single family dwelling); R2 (two family); R3 (three family); R4 (four family); and CD (condominiums) . In her Fitchburg study, the Commissioner analyzed only Rl and land.

Factoring is a method of achieving full and fair cash value. In substance, a multiplier is found which will increase the assessment/sales ratio to 100%.

The mean is the average of the ratios.

Equalized value is a figure determined at two-year intervals by the Commissioner pursuant to G. L. c. 58, §§ 9-10C. The equalized value study is supposed to be based on a review of reports submitted by local assessors and individual property owners, supplemented by field reports and assessments by the Commissioner’s staff.

We note that the factor was derived from the 1980 EQV study (not the nine-month study performed to test assessment levels in Fitchburg) which relied on the mean, not the median, ratio. Use of the mean method yields a different result from that yielded by use of the median method.

Thus 89.6% x 1.12 = 100%.

The mean ratios from the 1980 EQV study for these subclasses were: 112-96.5%, R3-100.0%, and R4-96.5%. Thus the factored mean ratios for these properties were R2-102.3%, R3-106.0%, and R4-102.3%.

In their answer and motion to dismiss, the city of Fitchburg defendants state that the reduction in the tax rate would have been $6.28/$l,000. In their brief to this court, however, these defendants stated, with respect to the free cash issue, that they “are generally in agreement with the factual statement of the Taxpayers and incorporate same herein.” In their statement of facts, the taxpayers state that the reduction in the tax rate *759 would have been $7.28/$l,000. We decline to resolve this aspect of the dispute because the record before us is not sufficient for us to do so and because such a determination is better left in the hands of the Appellate Tax Board.

See guidelines supra at note 6.

The Superior Court judge termed them “opaque, even to a skilled assessor.”

Assessment/sales ratios should fall between 90% and 110% and COD’s should be 15% or less.

The guidelines state, in pertinent part: “[T]he variation between residential and land ratios must be less than twenty percent.” This may mean either that the COD for land shall not be greater than 20% more than the COD for residential property, or that the difference between the median assessment/sales ratios for residential and land must be less than 20 % of the residential/land median. Under the first interpretation, there *762 is no limit on the median assessment/sales ratio for land; under the second, there is no limit on the COD for land.

See note 6, supra.

See note 7, supra.

In fact the Commissioner discarded the results of her study of land parcels because her data were not sufficient to obtain statistically significant results. The taxpayers do not specifically object to this action.

We emphatically state that we do not approve of the Commissioner’s going outside of her guidelines, absent such exigent circumstances as a lack of time and staff. While we do not take it upon ourselves to advise the executive here, we note that the Commissioner is free to promulgate guidelines to which she and her staff are able to adhere.

“Equalized value” is a figure determined biennially by the Commissioner pursuant to G. L. c. 58, §§ 9-10C. The valuation is determined from reports submitted by assessors and property owners and from field visits and appraisals by the Commissioner’s staff.

In addition, the judge stated: “[Ojf the ten sub-classes which ought [to] have been studied in Fitchburg as part of the 1980 EQV Study, six sub-classes received arbitrary measures of assessment level and a seventh had the assessment level derived from calculation ‘rounded off’ to drop four percentage points. This is hardly an acceptable approach to measuring general assessment practices in Fitchburg and no cry of budgetary constraints excuses it.”

The Superior Court judge, while pointing out the inadequacies in the Commissioner’s methodology and the resultant inequities in assessments, ruled that the taxpayers could not complain that the Commissioner relied on the EQV study. He found that the taxpayers had standing pursuant to G. L. c. 58, §§ 10A-10C, to attack the EQV study and, as they failed to *767 do so, their present complaint “simply comes too late.” He reasoned that, since the Commissioner relies on the EQV study to certify communities, taxpayers in those communities can show the requisite injury to challenge inadequate EQV studies. Indeed, stated the judge, “were these plaintiffs to be denied any forum in which they might challenge the manner of conducting the Equalized Valuation Study (now that we know the important role that study plays in tax classification decisions), it might pose substantial constitutional problems.” We do not think the taxpayers’ complaint is untimely. Since the plaintiffs have proven the requisite illegality in the Commissioner’s actions, we make a declaration so stating. G. L. c. 231A.

See note 28, supra.

We note that the Commissioner is free to revise her guidelines if she finds them unsatisfactory.

Equitable relief is available before such commitment. Pursuant to G. L. c. 59, § 43, assessors must value and assess property on tax lists which are open to the public. It is before these lists are committed to the tax collector that equitable relief may be pursued under G. L. c. 40, § 53.

In brief, the judge explained that while the assessments for land would rise if all assessments were at full and fair cash value, those in R2, R3, and R4 would decline by an even greater margin. Thus, total tax revenues for all R parcels would decline, resulting in a greater tax burden for the plaintiffs who are owners of commercial and industrial parcels.

The parties neither argued the jurisdictional question nor did the judge state why the board might not have jurisdiction. In such circumstances, we need not reach nor discuss that issue.