Opinion by
Judge Craig,Lily Penn Food Stores, Inc. and Joan Arnone, individually and on behalf of all Area 1, Zone 21 milk *599consumers, petition for review of the Pennsylvania Milk Marketing Board’s order A-837, which increased minimum resale prices for whole, lowfat and skim milk by twelve cents per gallon.2
The price increase was precipitated by a request for hearings to review minimum prices filed on April 16, 1980 by the Suburban Milk Dealers Association (Association), representing four milk dealers located in Zone 2.3 The board convened the hearings on September 8,1980; the dealer’s primary witness was their accountant, Carl Herbein.
Lily Penn, which owns and operates a chain of dairy convenience stores in Area 1 under the trade name Cumberland Farms, presented several witnesses. The board also heard testimony from three of its staff members: an audit supervisor, the economic director and director of the bureau of statistics, and its executive secretary.
Following the parties’ submission of proposed findings, the board issued order A-837 on January 28, 1981, with an effective date of February 4. Petitioners Lily-Penn and Mrs. Arnone obtained a stay from this court pending determination of their appeal.
The pertinent section of the Milk Marketing Law4 is Section 801, which states in part:
The board shall base all prices upon all conditions affecting the milk industry in each milk *600marketing area, including the amount necessary to yield a reasonable return to the producer, which return shall not be less than the cost of production and a reasonable profit to the producer, of the quantity of milk necessary to supply the consumer demand for fluid milk plus a reasonable reserve supply as determined by the board, and a reasonable return to the milk dealer or handler. However, where the board determines that the market for Pennsylvania produced milk is threatened it may establish producer prices designed to market the milk. In ascertaining .such returns, the board shall utilize a cross-section representative of the average or normally efficient producers and dealers or handlers in the area and shall consider the cost of containers according to size and type.
All provisions of all price-fixing orders of the board shall be presumed to be valid, and the burden of proving any invalidity of any provisions thereof shall be upon the person asserting the same.
The Gross-Section Used
The petitioners’ submit that order A-837 is not supported by substantial evidence, is arbitrary, unreasonable and not in accordance with law.
However, the pivotal issue is whether the board used a “cross-section representative of the average of normally efficient... dealers or handlers in the area” in ascertaining the dealer’s reasonable return, as mandated by Section 801. Embraced within this issue is the petitioners’ assertion that the board-failed to consider “all conditions affecting the milk industry” in the area.
*601In its order, the board concluded that the four member dealers of the Association were a representative cross-section of dealers in Zone 2 for 1979. The board (and the Association, as intervenor in this appeal) defend that decision on the grounds that: (1) the plants of all four dealers are physically located within Zone 2; (2) they are full service dealers handling a complete line of products and selling milk to home delivery customers, stores, restaurants, institutions and schools; and (3) the cross-section accounts for approximately 39% of the net sales dollars for Zone 2 in 1979.
Because the testimony of the expert witnesses was conflicting on many points, the board, in its findings and conclusions, evaluated the evidence and explained its reasons for accepting as more credible certain statements over others.
Although we may not infringe upon the board’s exclusive function as arbiter of credibility barring complete absence of evidentiary support, Lily-Penn Food Stores, Inc. v. Milk Marketing Board, 42 Pa. Commonwealth Ct. 92, 400 A.2d 661 (1979), our review of the record leads us to conclude that the board erred as a matter of law in holding that the four dealers were a representative cross-section of the average or normally efficient dealers or handlers in the area.
The record establishes, and the board so found, that the four dealers accounted for only 28% of the volume of milk sold in Zone 2 by weight. The board’s own exhibits demonstrate that the dealers used in the representative examination comprise only one-fourth of the sixteen licensed dealers whose plants are physically located within Zone 2. Twenty-one additional dealers with plants located outside the zone are listed as selling milk within the zone.
A significant amount of the evidence related to the changing characteristics of milk sales to consumers, such as the current predominance of non-returnable *602plastic gallon containers, the ascendant consumer practice of purchasing milk in supermarkets, and the decrease in home delivery and institutional sales volume.
Several exhibits indicated that sales by dealers to Area 1 stores accounted for anywhere from 66% to 75% of all sales, and 75% to 81% of total wholesale sales of packaged milk. Stores in Area 1, Zone 2 purchased almost 90% of gallon-size containers, 67% of half-gallon size and 97 % of gallon twin-packs sold by state milk handlers in 1979. Of the four dealers whose operations were studied by the board, only one had some supermarket chain business, and the dealers’ accountant testified that the percentage was much smaller than that of conventional dealers which serve the supermarkets on a continuing basis.
The tables further reveal that the plastic, nonreturnable gallon container accounted for the overwhelming majority of gallon sales by handlers in Zone 1 of Area 1. The board’s economic director testified that at least eight plants selling in Zone 2 have “in-line blow-molding equipment,” integrating the filling operation with in-house manufacturing of the plastic container.5 None of the four dealers have this equipment.
The economic director agreed with the opinion of a Lily-Penn witness that a very high percentage of milk sold out of supermarkets in Zone 2 is sold by plants processing between 100,000 to 150,000 quarts per day. The average daily production of the four subject dealers is less than 25,000 quart equivalents.
*603A final point, significant to our decision, is the economic director’s testimony that a substantial amount of inter-zone movement of packaged milk exists between both zones of Area 1, and between Area 1 and out-of-state. The witness stated that at least 29.2% of the milk sold in Zone 2 was processed in Zone 1, and 11.6% was processed out of state. More interestingly, a board report entered as an exhibit indicated that fully 55.4% of the milk received from producers by Zone 2 dealers is distributed outside the state, with 20% distributed in Pennsylvania areas other than Zone 2.
The proximity of Zone 2 to Zone 1 in Area 1, to Pennsylvania’s Area 4, to Delaware, and to New Jersey admittedly provides a greater sphere of sales activity than might normally exist for the board to consider. However, the interstate and inter-zone movement of milk, along with the changing characteristics of milk marketing and consumption trends apparent from the testimony, clearly indicate that the four dealers were not representative for purposes of ascertaining reasonable returns in setting prices for milk in Zone 2.
In complying with the statutory directive to use a representative cross-section, the board must be guided by the primary language of Section 801, which mandates board consideration of “all conditions affecting the milk industry in [the] area.” That language reflects legislative recognition of the varied elements affecting prices, along with imminent transition and growth in the industry. As we stated in City of Pittsburgh v. Milk Marketing Board, 1 Pa. Commonwealth Ct. 300, 275 A.2d 115 (1971), “[the] law does not remain static. It is a living thing, and when conditions change, the legislative intent must be read in light of changed circumstances.” 1 Pa. Commonwealth Ct. at 311, 275 A.2d at 121.
*604Where, as here, the operations of dealers or handlers located outside the marketing zone so integrally affect the sale and distribution of milk within the Zone, we believe that the language of the section should not be so strictly construed as to limit board consideration of dealers only to those physically situated within the subject area.
Our decision that the board may use peripheral businesses for cross-section purposes does not put too much of an onus on the board or those petitioning for a change in milk prices. Dealers or handlers marketing milk in an area subject to Pennsylvania law, whether located there or out of state, are directly affected by price-setting orders, and are potential “interested parties” under Section 801. Participation of a sampling of them in the proceedings is clearly warranted.
We will reverse order A-837 and remand to the board for new hearings, with a direction to utilize a cross-section of dealers or handlers, including retail stores,6 more representative of the industry.
The Cost Statistics Used
The petitioners also challenge the validity of order A-837 on the ground that the board failed to use the “proper statistics” in support of its conclusions. *605G.S.F. Corp. et al. v. Milk Marketing Board, 1 Pa. Commonwealth Ct. 216, 218, 273 A.2d 756, 757 (1971).
Two of petitioners’ witnesses presented evidence from cost studies of the operations of various milk dealers and stores, none of which were located in Zone 2; their findings incorporated an evaluation of individual cost factors (raw product cost, size of deliveries, type of package, cost of utilities, mode of transportation, administration, personnel involvement, and rate of return) in calculating the prospective dealer and in-store handling costs for each product sold.
The board found that the unit cost evidence could not be the basis for setting Zone 2 minimum prices because of certain disparities between the operations of the six dealers used for the cost studies and those of the four cross-section dealers.7
The figures presented by the Association’s accountant, which were based wholly on the 1979 composite statements of the four member dealers, were accepted by the board in support of its order. Because the figures were prepared in accordance with the board’s Uniform System of Accounts, and the dealers studied were representative of the Zone 2 dealers, the board found the composite statements to be “an accurate starting point for determining resale prices in Area 1 Zone 2.”
The board’s decision to accept the Association’s evidence over that of the petitioners is solely a credibility determination, which may not be disturbed absent an abuse of discretion. Lily-Penn Food Stores, Inc. v. Milk Marketing Board, 42 Pa. Commonwealth Ct. 92, 400 A.2d 661 (1979).
*606However, in light of onr conclusion that the four dealers studied by the board were not a. representative cross-section, we find merit in the petitioners’ assertion that the financial data utilized by the board was legally insufficient because the composite statements did not reflect the lower operational costs of any dealers with a large volume of supermarket deliveries.
Order A-837 establishes disparate prices for out-of-store, home-delivery, school and institutional sales. Keeping in mind the statutory mandate to consider all conditions affecting the milk industry, we believe that board appraisal of the different costs resulting from these methods of distribution is justified, if not mandated.
Section 704 instructs licensed milk dealers to establish and maintain records in accordance with the Uniform System of Accounts “in order to facilitate the cost studies provided for in section 801.” (Emphasis supplied.)
Section 801 directs the board to “consider the cost of containers according to size and type” in ascertaining a reasonable rate of return. The board’s use of unit costs in determining minimum out-of-store prices in past orders8 demonstrates that recognition of the variable factors affecting dealer costs is essential to an accurate and fair determination of profit.
Board consideration of cost elements, derived from the operations of a carefully chosen cross-section of dealers would not, as the Association claims, be unduly burdensome. Unit cost information is readily available to the board; section 701 requires all licensed milk dealers to keep records on (1) the cost of the raw prod-*607net, (2) the cost of containers used, by size and type, (3) the quantity of milk sold, with details as to grade, use, location and container size, (4) the transportation costs and distances for the movement of milk sold, (5) the wastage or loss of milk or butterfat, and (6) the “spread or handling expenses and profit or loss, represented by the difference between the price paid and the price received for all milk and milk products.”
Consequently, the board should consider evidence of the unit costs of representative dealers or handlers in redetermining the minimum resale prices for Zone 2.
A supplemental argument, raised by the petitioners in response to our holding in City of Pittsburgh v. Milk Marketing Board, 60 Pa. Commonwealth Ct. 422, 433 A.2d 561 (1981), questions the board’s failure to require the dealers’ accountant to produce any documentation to support his cost projections for 1980. In light of our decision to remand for new hearings, we need not address this issue. The petitioners will have ample opportunity to request the production of individualized operating statements or other pertinent books and records for purposes of cross-examination.
The Rate of Return Used
The board found that there was “no sufficient evidence by qualified witnesses to justify a change” in the 3% rate of return on net sales, which was established as reasonable for Zone 2 dealers in the order preceding A-837. In order to achieve that return, the board increased minimum resale prices by 2.91 cents per quart equivalent, or 12 cents per gallon.
In Milk Control Commission v. United Retail Grocers, 361 Pa. 221, 64 A.2d 818 (1949), the commission had refused to change an existing differential between *608minimum wholesale and retail prices; the Supreme Court held:
[OJrderly procedure requires that previous orders of the Commission, unappeaied from, fixing a minimum price in the case of producers or markup in the case of dealers, including stores, be assumed to be fair, just and reasonable in the absence of proof of a change in factual circumstances. [Citation omitted.] A reexamination is contemplated when conditions change, but the burden is on complainants in such case to show facts or circumstances that warrant such a conclusion. (Emphasis supplied.)
361 Pa. at 224, 64 A.2d at 820.
The commission had rejected the testimony in that case as insufficient to warrant a change in the differential because it was “so fragmentary, so speculative and so indefinite, and was based so largely upon opinion and guess, as to make it unworthy of acceptance. ’ ’ 361 Pa. at 222, 64 A.2d at 819.
The petitioners here presented two expert witnesses to testify on the subject of profits. Dr. Stammer, an associate professor in economics and agriculture, based his testimony on a U.S. Department of Agriculture Survey of 30 “conventional” dealers, which was introduced as an exhibit. The survey, covering a period from 1972-1977, indicated an average annual return of between 2.4% and 2.8% on sales before taxes for the dealers studied. Dr. Stammer stated that, taking into consideration the lower risk element in a price-controlled area such as Pennsylvania, he felt that a reasonable rate of return should be in the range of 2.5% to 2.8% on the dealers’ total product line.
The dealers’ average net operating income for 1979, as determined by the board’s audit supervisor, was 2.4% of “controlled” net sales, 5.3% of “non-*609controlled” net sales, and an overall operating income of 3.2% of net sales. 9
Unlike the milk commission in United Retail Grocers, the board here not only failed to explore the veracity of the proffered evidence, but also omitted any explanation for its conclusion that there was no change in the circumstances to justify a revision of the established rate of return.
Absent enumeration of the findings underlying the board’s decision, we cannot presume the continued reasonableness of a 3% rate of return for the purpose of setting minimum prices in Zone 2; any review of the determination is impossible. City of Pittsburgh v. Milk Marketing Board, 35 Pa. Commonwealth Ct. 400, 386 A.2d 624 (1978).
Lowfat and Skim Milk
The petitioners’ final objection to order A-837 is grounded on the minimum prices for lowfat and skim milk.10 The board retained the previously established differential of 8 cents per gallon between the price of lowfat and whole milk and 12 cents per gallon between the price of skim and whole milk, so that the minimum price for both defatted products rose by the same increment of 12 cents per gallon which the board applied to whole milk.
Claiming that the lowfat and skim milk prices should have been reduced, the petitioners point to the *610raw product cost savings which result from the lower butterfat content of these products.
The butterfat differential, set by federal order, expresses the difference in the value per pound of butterfat and skim in raw milk. The per gallon raw product cost savings is computed by multiplying the butterfat differential by a conversion factor which incorporates the difference in butterfat content of the product from whole milk and the number of gallons in 100 pounds of milk.11
Thus, with a butterfat differential of $1.60 per pound when the hearings began in September of 1980, the raw product cost difference was approximately 31 cents per gallon for 1% lowfat milk and 38 cents for skim milk. As of the effective date of order A-837, the difference became 32.6 cents for 1% lowfat and almost 40 cents for skim milk.12
The Association, in contesting the petitioners’ allegation of lower raw product cost for lowfat and skim milk, argues that the additional costs incurred by dealers in extracting the excess butterfat supports the lower differentials between the prices. Without supporting evidence, the Association’s brief claims that *611lowfat products-have low sales volume, and that sale on the open market of the excess cream generated by the production is often at a price less than the cost of the raw milk product, resulting in a net loss.
In addition, and somewhat inconsistently, the Association argues that if prices are lowered, the sales volume of those products will increase, with a corresponding decrease in the sales of standard milk. The Association contends that prices for whole milk, school and institutional milk would have to be increased in order to compensate for the “inevitable” loss of revenue arising from enlarged lowfat differentials. Nor was any evidence presented before the board to support these allegations.
To substantiate their claim, the petitioners offered the testimony of expert witness Stammer; his lengthy and detailed exhibits revealed calculations of the net raw product cost savings per gallon of lowfat and skim milk, using federal order butterfat differentials and average market prices for cream disposal, and taking into account shrinkage and transportation costs.
Dr. Stammer also computed, for dealer cost purposes, the amount of surplus cream generated by the separation process, explaining:
The first step we do, when we separate out the cream, we want to be left with 100 pounds of one percent milk. Basically, we start 105.625 pounds, and we separate out 5.625 pounds of cream, and we would be left with 100 pounds of one percent milk.13
*612He concluded that disposal costs for the excess cream would average from one to three mills per gallon (.001 to .003 cents per gallon), and that marketing difficulties were unlikely in light of the regular purchase of surplus cream, at market prices, by an area butter manufacturer.
Hr. Stammer also examined the separating process incrementally, computing machine time, labor, overhead and depreciation costs involved in producing a gallon of lowfat milk. He determined that the extra expense of processing was approximately half a mill per gallon (.0005 cents per gallon).
Based on all the evidence, the witness stated that the best way to compute the price differentials would *613be to multiply the conversion factors for 2%, 1% and skim milk14 by the current federal order butterfat differential, published monthly, and to subtract a penny from that figure to account for any processing or disposal costs.
The board found that “no relevant evidence in the record . . . would justify a change in the existing differential for lowfat and skim milk.” In its discussion of the evidence, the board discounted Dr. Stammer’s testimony on the basis of a single miscalculation. The discrepancy resulted from the board’s discovery that the separation process generated 6 pounds of cream in producing 100 pounds of 1% milk, rather than 5.625 pounds.15
In rejecting the witness’ testimony, the board failed to observe that the miscalculation arose solely with respect to Dr. Stammer’s approximation of the offsetting cost of disposing of the surplus cream. The *614error did not in any way reduce the relevancy or weight of the evidence of raw product cost savings per gallon,16 nor was it so significant as to render immaterial the otherwise comprehensive testimony, as the board apparently concluded. Without any other evidence to controvert the legitimacy of the petitioners’ challenge to the lowfat differentials, we must conclude that the board abused its discretion in dismissing the testimony.
The redetermination hearings will provide the parties with the opportunity to present supplemental evidence in support of or in response to their relative positions on the issue.
Conclusion
We hold that the board failed to use a representative cross-section of the average or normally efficient dealers serving Zone 2 in ascertaining a reasonable rate of return, and abused its discretion in rejecting the only material evidence on the issue of lowfat and skim milk price margins. Order A-837 will be reversed and the matter remanded to the board for redetermination of the minimum prices for Zone 2, with instructions to consider relevant unit cost evidence and to support its findings and conclusions with a discussion of the applicable testimony and determinations of relative weight or credibility.
Order.
Now, November 25, 1981, order A-837 of the Pennsylvania Milk Marketing Board, dated January 28, 1981, is reversed, and the matter is remanded for further hearings consistent with this opinion, with prices as fixed in order A-821 to remain in effect until disposition by the board.
Area 1, the Southeastern Milk Marketing Area, was divided into two zones in 1977. Zone 2, the suburban Philadelphia zone, encompasses Chester County and those portions of Bucks and Montgomery counties lying north of the Pennsylvania turnpike.
The order does not change the price which milk dealers or handlers must pay to producers — the dairy farmers — or their cooperatives. Minimum producer prices remain as set under the regional directives of the Federal Milk Marketing order; the board, in its orders, incorporates by reference the provisions of the federal order.
Bechtel Dairies, Inc., Royesford, Pa.; Meadow Brook Farms, Inc., Pottstown, Pa.; Rosenberger’s Dairies, Inc., Hatfield, Pa.; and Sunny Slope Dairies, Inc., Spring City, Pa.
Act of April 28, 1937, P.L. 417, as amended, 31 P.S. §700j-101— §700j-1203.
The witness further testified, over strenuous objection by counsel for Lehigh Valley Farms (a participant in the hearings), and counsel for the Association, that, in his opinion, the data would indicate that more .than fifty percent of the milk sold in gallon quantities out of supermarkets in Area 1, Zone 2 was processed in plants which have their own in-line blow molding equipment.
The Act broadly delineates those businesses which the board may utilize in ascertaining a representative cross-section. Section 103 defines “milk dealer” or “handler” as “any person, including any store or subdealer or subhandler . . . who purchases or receives or handles on consignment or otherwise milk within the Commonwealth, for sale, shipment, storage, processing or manufacture, within or without the Commonwealth, whether on behalf of himself or others, or both.”
Thus, cost evidence from supermarkets and convenience stores is certainly relevant to board consideration of price modifications which would result in increased revenue for those businesses. As noted by the dissenting board member in her opinion, no supermarkets participated in the hearings preceding order A-837.
The board stated that its rejection of the evidence was based on the fact that the six dealers studied were located in five states other .than Pennsylvania; and that all of the plants produced over 100,000 quarts per day, had large deliveries to supermarkets, utilized in-line blow molding equipment, and did not produce paper gallons.
See, e.g., the findings in support of board order A-793, September, 1975, and the findings in support of board order A-821, July, 1978, which established the prices for then newly-created Area 1, Zone 2, and is currently operative pending the disposition of this appeal.
The dealers’ accountant’s figures differed slightly from those of the board auditor. Mr. Herbein’s combined statement of operations showed net operating income of 2.2% of “controlled” net sales, 4.5% of “non-controlled” net sales, with an overall operating income of 2.8%.
The standard butterfat content of whole milk is 3.25%, although the amount of butterfat in raw milk varies seasonally from 3.5% to 3.8%. Lowfat milk generally has a butterfat content of anywhere from .05% to 3.25% while skim milk has less than .05% butterfat.
Assuming that the product is 1% lowfat milk, the difference in butterfat from whole milk is 2.25%. One hundred pounds (cwt.) of lowfat milk thus contains 2.25 pounds less butterfat than whole milk. With 11.628 gallons of milk per cwt., the conversion factor for lowfat milk is 2.25 -r 11.628 = .193.
The conversion factor for .05% skim milk is
2.75 -i- 11.628 = .236.
The conversion factor for 2% lowfat milk is
1.25 -i- 11.628 = .107.
September, 1980
1% lowfat $1.60 x .193 == 30.9 cents per gallon
skim $1.60 x .236 = 37.7 cents per gallon
February, 1981
1% lowfat $1.69 x .193 = 32.6 cents per gallon
skim $1.69 x .236 = 39.8 cents per gallon
Dr. Stammer’s calculations, based on tbe July 1980 federal order butterfat differential of $1.60 and the average market prices for Class II cream, as reported in the Dairy Market News, are as follows:
Gross Raw Product Cost Differences
Butterfat differential per cwt. (2.25 x $1.60) $3.60
Gross differential per gallon (3.60 -t- 11.628) .3096
*612 Net Average Cost (Or Profit) In Sale of Cream Generated
Step 1: In reducing fat content of milk from 3.25% to 1%, dealer generates 5.625 lbs. of 40% cream for each cwt. of 1% milk.
Federal Order cost:
2.25 lbs. fat (at $1.60) $3.60
5.625 lbs. of skim (at .0624) .351
3.951
Step 2: Net average market value of cream generated :
a) Avg. market price ($2.25 lbs. of butterfat at $l,7901/lb) $4.0277
b) Less shrinkage and transportation (shrinkage at 1.25% of volume; transportation at $1.00/cwt. of product) - .0764
3.9513
Step3: Net cost (profit) on disposal of cream:
Federal Order cost (Step 1), minus net market value (Step 2) :
Per cwt. (3.9510 - 3.9513) .000 (rounded)
Per gallon (-4- 11.628) .000
Step4: Net raw product cost saving per gallon:
gross difference ($.3096) ± net disposal cost
(0) $ ,3096/gal.
See note 11.
The board, in its discussion, expressed it differently, stating that the fat content would he 1.2% if only 5.625 pounds of cream were removed:
[Dr. Stammer’s] testimony, however, was based upon the erroneous assumption .that when using a separator which provides forty per cent (40%) cream and sixty per cent (60%) low-fat product from 100 pounds of 3.25% milk, the addition of 5.625 pounds of milk containing 3.25 per cent (3.25%) butterfat to 100 pounds of product also containing 3.25 per cent (3.25%) butterfat, upon separating out 5.625 pounds of forty per cent (40%) cream and sixty per cent (60%) low fat product, the 100 pounds of low fat product that remains contains 1% butterfat. Mathematical calculation will show that the 100 pounds of low fat product remaining will have a butterfat content of 1.2 per cent (1.2%) which is a significant amount of error when viewed in terms of Federal Order accountability. (See Testimony 10/1/80, p. 1326).
In view of the situation described above, Dr. Stammer’s testimony cannot be the basis for setting low-fat differentials in Area 1, Zone 2.
See, e.ff., the calculations of raw product cost savings in note 12.