The basic issue on this appeal is whether the PSC’s decision and order denying Robertson’s application to remove the interroute restrictions on engaging in a single-line service between Monroe and Madison-Milwaukee is supported by substantial evidence in light of the entire record as submitted.

The granting or denial of a certificate of authority by the PSC is done in the exercise of legislative discretion by the commission. Its power is granted by sec. 194.23 (1), Stats., which provides in substance that no person shall operate a motor vehicle as a common motor carrier except in accordance with the terms and conditions of a certificate issued by the commission. The commission has the power “as the public interest may require, upon a finding of public convenience and necessity, to issue or refuse any such certificate or amendment . . . .” However, before granting a certificate “the commission shall take into consideration existing transportation facilities in the territory proposed to be served . . . .” On review under ch. 227, when the sufficiency of the evidence is challenged, this court in its review must apply the “substantial evidence” test. Sec. 227.20 (1). And, the court must give due weight to the experience, technical competency, specialized knowledge and discretionary authority of the agency. Sec. 227.20 (2).

*658It is argued the traditional scope of this test has been enlarged in recent cases. Consequently, it is necessary for us to briefly review the more important cases. We call attention to the case of Stacy v. Ashland County Dept. of Public Welfare (1968) (handed down concurrently with this opinion), ante, p. 595, 159 N. W. 2d 630, which also discusses this problem. The basic case is Gateway City Transfer Co. v. Public Service Comm. (1948), 253 Wis. 397, 34 N. W. 2d 238. That case pointed out that in reviewing administrative decisions, “substantial evidence” did not include the idea of this court weighing the evidence to determine if a burden of proof was met or whether a view was supported by the preponderance of the evidence. Such tests are not applicable to administrative findings and decisions. We equated substantial evidence with that quantity and quality of evidence which a reasonable man could accept as adequate to support a conclusion. And, in this process, sec. 227.20 (1) (d), Stats., providing that the decision of an agency may be reversed if unsupported by substantial evidence in view of the entire record as submitted does not permit this court to pass on credibility or to reverse an administrative decision because it is against the great weight and clear preponderance of the evidence, if there is substantial evidence to sustain it.

Substantial evidence is not equated with preponderance of the evidence. There may be cases where two conflicting views may each be sustained by substantial evidence. In such a case, it is for the agency to determine which view of the evidence it wishes to accept. Likewise, there are cases where only one view can be supported by substantial evidence and the determination depends upon the credibility of witnesses.

The view that this court is guided by different standards in reviewing an agency’s finding than it is in reviewing a finding of a trial court or a jury is followed in Motor Transport Co. v. Public Service Comm. (1953), *659263 Wis. 31, 56 N. W. 2d 548. We do not think this view is changed by Copland v. Department of Taxation (1962), 16 Wis. 2d 543, 114 N. W. 2d 858, in which case this court held there was no substantial evidence to support the finding of the commission of the value of goodwill. This was not because of this court’s concept of what the value of goodwill was but rather because the method of evaluating or arriving at that value was unreasonable. What is substantial or what a reasonable man might consider to be adequate support of a conclusion lies within the domain of the reviewing court and this court may well differ on this point with an administrative agency. This is not a violation of the separateness of the branches of the government but the exercise of a check and balance. Likewise, in Silverberg v. Industrial Comm. (1964), 24 Wis. 2d 144, 128 N. W. 2d 674, this court held there was no substantial evidence to support a finding denying a license to an employment agency. This finding related to a statutory standard, whether the number of licensed employment agencies in a given area was sufficient to supply the needs of employees and employers.

Such a determination is not by nature legislative whereas a finding of what constitutes “public interest” for various purposes and circumstances and without guidelines has uniformly been held to be a legislative function. Gateway City Transfer Co. v. Public Service Comm., supra; In re City of Beloit (1968), 37 Wis. 2d 637, 155 N. W. 2d 633. Likewise, the terms “public convenience and necessity” in relation to transportation are comprehensive and broad terms. One of the qualifications of the statute requires the commission in making such a determination to take into consideration existing transportation facilities in the territory proposed to be served. In Gateway we held this language did not require or promote either a monopolistic or competitive view of transportation. We affirmed this view in Priebe *660v. Public Service Comm. (1968), 38 Wis. 2d 635, 157 N. W. 2d 600.

The volume of business between two terminals maybe such that two common carriers are needed to serve the public convenience and necessity. Likewise, the volume may only justify one line. When two lines are allowed, the cry is made of favoring competition; when one line is allowed, the charge of monopoly is made. This case is no exception.

The evidence on behalf of Robertson consisted of the testimony of some 22 witnesses who were shippers and expressed dissatisfaction with the service of Gateway, especially its slowness, damage to shipments and delays in claim settlements. The service in respect to time of Robertson was no better because of the joint-line restriction.

In rebuttal, Gateway used the testimony of their Janes-ville terminal manager who offered exhibits covering four weeks of four different months showing concrete examples of actual shipping time between Milwaukee and Monroe and between Madison and Monroe. This evidence contradicted the oral testimony offered by Robertson in the sense it would support an inference in respect to Gateway’s service contrary to the inference which could be drawn from Robertson’s evidence. A reasonable person acting reasonably and considering the testimony of Robertson’s witnesses could conclude such testimony represented the exception and that Gateway was furnishing reasonably adequate service. If we were to weigh the evidence in terms of preponderance, we might agree the evidence in favor of Robertson is of the greater weight. However, the evidence in favor of Gateway is substantial. It is for the commission to evaluate the evidence in the first instance and to draw the reasonable inference that Gateway’s service was of such quality as to satisfy the public convenience and necessity and conclude that the public interest on this *661record did not require any additional single-line service between Monroe-Madison and Milwaukee.

Basic in the commission’s thinking was the view that the volume of business between Milwaukee and Monroe would not support two single-line systems. However, it is evident that Robertson could improve the quality of its service to the public with a single-line operation. This view would seem to result in a great economic waste by failing to allow a common carrier to improve the quality of its service and better serve the convenience of the public, but this policy determination was entrusted to the PSC. A majority of this court concludes, as did the trial court, that there is substantial evidence in view of the entire record to sustain the findings; a minority would not.

Robertson claims the findings are arbitrary and capricious, mostly on the proposition the commission acts inconsistently in various cases. Consistency, of course, is a virtue both in administrative and in judicial determinations but inconsistencies in determinations arising by comparison are not proof of arbitrariness or capriciousness. Dairy Employees Independent Union v. Wisconsin Employment Relations Board (1952), 262 Wis. 280, 55 N. W. 2d 3; Nick v. State Highway Comm. (1963), 21 Wis. 2d 489, 124 N. W. 2d 574; see 2 Davis, Administrative Law Treatise, p. 528, sec. 17.07. We have said that an agency does not act in an arbitrary or capricious manner if it acts on a rational basis. Chicago, M., St. P. & P. R. R. v. Public Service Comm. (1954), 267 Wis. 402, 423, 66 N. W. 2d 351. Arbitrary action is the result of an unconsidered, wilful or irrational choice, and not the result of the “sifting and winnowing” process. Olson v. Rothwell (1965), 28 Wis. 2d 233, 239, 137 N. W. 2d 86; Iron River Grade School Dist. v. Bayfield County School Comm. (1966), 31 Wis. 2d 7, 13, 142 N. W. 2d 227.

By the Court. — Judgment affirmed.

(dissenting). Monroe, Wisconsin is served by two common carriers in truck shipments to and from Madison and Milwaukee. One, Gateway Transportation, provides single-line service operating through its Janesville terminal. The other, Robertson Transportation, provides joint-line service, meaning that it can truck between Monroe and Madison or Monroe and Milwaukee by transferring cargoes for a portion of the trip to trucks of cooperating carriers. The applicant-appellant, Robertson Transportation, seeks by its application to provide single-line service instead of the joint-line service presently authorized. This straightening out of the kinks in the authorized service of Robertson Transportation is described by the Public Service Commission in its decision as a “streamlining” and a “coordination” of routes. That seems to be an accurate description, and it is all that is here involved.

The majority affirms, upholding the commission’s decision that such streamlining and coordination of an existing service would not serve the public interest. I would reverse, relying upon the fact-finding of the commission that “Applicant could provide a more efficient and economical service to the public if its certificate is streamlined and service coordinated between routes,” and finding no substantial evidence in the entire record that such improvement in an existing service is not in the public interest.

The legislature has clearly determined that this type of application is to be granted or denied “. . . as the public interest may require, upon a finding of public convenience and necessity, . . .” taking “. . . into consideration existing transportation facilities in the territory proposed to be served . . . .” (Sec. 194.23, Stats.) The legislature has as clearly provided that the judicial review of a commission decision “. . . shall be confined to the record, . . . The court may affirm the decision of the agency, or may reverse or modify it if the sub*663stantial rights of the appellant have been prejudiced as a result of the administrative findings, inferences, conclusions or decisions being . . . unsupported by substantial evidence in view of the entire record as submitted; . . .” (Sec. 227.20 (1) (d), Stats.) As the majority opinion points out, the term “substantial evidence” requires only that “. . . upon an examination of the entire record, the evidence including the inferences therefrom, . . .” it is clear that “. . . a reasonable man, acting reasonably, might have reached the decision; but, on the other hand, if a reasonable man, acting reasonably, could not have reached the decision from the evidence and its inferences then the decision is not supported by substantial evidence and it should be set aside.” 1 (Emphasis supplied.)

Our position is that the facts here established do not sustain what the statute calls the “. . . inferences, conclusions or decisions . . .” made by the commission. The “reasonable mind” test applies to the drawing of inferences, the reaching of conclusions and the issuing of a decision as well as to the fact-finding function.

Most of the space in the briefs and most of the time on oral argument dealt, pro and con, with the following paragraph in the public service commission’s decision: “Gateway serves single-line from Madison, Milwaukee, La Crosse, and Eau Claire to Monroe and Janesville. It has generally provided an overnight service, although there were some allegations from public witnesses that such was not the case.” Appellant’s counsel spiritedly contends that the finding of “overnight service” made by the commission is unsupported by the evidence and cites complaints by shippers made at the hearing. In rebuttal, Gateway presented one witness, its terminal manager who produced records of shipments made during four *664selected weeks to corroborate the claim of “overnight service.” As to the disputed issue on whether Gateway service was in fact “overnight service,” the writer would agree that the commission’s finding that it was resolves the issue. However, twenty-one shipper-witnesses who testified complained, not only that the Gateway service often took more than twenty-four hours, but also that deliveries were made at inconvenient times (late in the day), that goods were mishandled, lost and delivered in a wet condition on occasion, that service on complaints and tracing lost shipments was poor. Such complaints were not answered so there is no foundation for any finding that service was “generally satisfactory.” Reasons for dissatisfaction with both the single-line Gateway service and joint-line Robertson service were given by the witnesses who testified at the hearing. From the testimony presented, no reasonable inference can be drawn either that the present service is satisfactory, or that existing services could not be improved. As to the question of satisfactoriness of existing services, the trial court, reviewing this record, stated:

“In conclusion, we feel constrained to add that if we were only dealing with the issue of quality of service as applied to the public interest, we would feel compelled to reverse the commission. Our own feeling is that the quality of service offered to shippers at Monroe would be improved should the requested amendment to Robertson’s authority be granted.”

Actually, either as to the single-line service of Gateway or joint-line service of Robertson, the uncontroverted testimony clearly establishes that the presently existing services are not satisfactory in all regards. While on oral argument, the commission attorney stated that there are always cranks to complain about any service, there is nothing in the record to identify the twenty-one shippers on the Monroe-Madison-Milwaukee run who appeared in support of the application as either cranks *665or unworthy of belief. For what it adds up to, their testimony is the only testimony in the entire record as to quantity and quality of service, except for the sample four weeks of documentary evidence related only to the issue as to “overnight service.” Such documentary evidence standing alone is insufficient to support an inference or the conclusion that the existing services are “generally satisfactory.” Certainly, it was not satisfactory to the users of the service who testified. The documentary evidence of next-day shipments simply did not deal with nor relate to all the areas of shipper complaint.

It seems obvious, however, that, in reaching its decision to deny the applicant’s request to improve its service, the commission was not basing its rejection on enthusiasm for the adequacy of existing services. Its action rests on its consideration of the interests of one of the two common carriers authorized by it to serve the Monroe area. In the commission brief, it is stated: “In this case we are dealing with transportation between points where the traffic is ‘thin.’ No responsible person would advocate that there be further dilution of traffic that is already described as ‘thin,’ absent some unusual circumstances, which is not present here.”

The “thinness” of the shipping traffic using either Gateway or the Eobertson service is overwhelmingly established by the record. As the commission brief points out: “The Wisconsin intrastate traffic to and from Monroe is very light . . . The Gateway witness testified on the eastbound movement that ‘we are very lucky to fill a trailer a week ....’” At the hearing before the commission, it was undisputed that Monroe shipments were very light, the traffic out of Monroe being about one trailer per week. The commission concluded that so miniscule a pie was too small to further divide. This knife cuts two ways. In its decision, the commission correctly states: “There is merit in removing restrictions wherever pos*666sible without damaging the competitive situation in the common carrier field . . . However, restrictions should be lifted as to each individual route only in those cases where doing so would not materially affect major operations of objecting carriers . . . .” Gateway is a giant in the common-carrier field in the United States. The record in this case includes the company’s annual report for the year preceding the calendar year in which this application was filed. With gross operating revenues for the year of over forty-one million dollars, and a net operating revenue of over three and one-half million dollars, Gateway is no dwarf in the transportation industry. Can the Monroe-bound traffic here involved of less than a trailer a day and the Monroe-originating shipments of less than one trailer per week be said to have a material effect upon major operations of Gateway? Can the marginal and peripheral interest of a nationwide shipper in a city from which it transferred its terminal building be reason for withholding approval of an application that would otherwise be in the public interest?

Consideration of existing facilities is related directly to the impact, favorable or unfavorable, upon the totality of service to the area, present and future, of granting or denying an application. Here the commission itself has found that an existing transportation facility “could provide a more efficient and economical service to the public” if “its certificate is streamlined and service coordinated between routes.”

We deal here with no question of establishing a competing common carrier service ab initio. The classic confrontation of monopoly versus competition in the transportation field is not present here. By commission action and policies, there are two common carriers, not only authorized but actually transporting freight between Monroe and Madison, between Monroe and Milwaukee. The question of whether there should be one or two common carriers serving Monroe has been answered. Not only is Robertson, as well as *667Gateway, franchised to serve the Monroe area, hut, on the strength of such commission authorization, Robertson has established a terminal in Monroe. The relevant statute includes all existing facilities when it authorizes the commission to take into consideration “. . . existing transportation facilities in the territory proposed to be served . . .” Both Robertson and Gateway are such existing transportation facilities, already serving the territory. Monroe is not a territory proposed to be served by a new carrier. It is a territory proposed to be better served by an existing facility that the commission has found could provide “. . . a more efficient and economical service to the public . . .” if its application were to be granted. In this situation, with the fact of improvement of service found to be a consequence of granting the application, there appears to be a compelling reason for putting first and foremost the predictable overall improvement of the existing services. To do otherwise is to dilute the test of serving the “public interest, upon a finding of public convenience and necessity” to a pale, pale shadow.

We would not quarrel with this statement in the commission’s brief: “It is also well known that if the traffic on any tranportation operation is too ‘thin’ and hence of marginal profitability at best, the service tends to drop off, which drop-off is accentuated with a further drop in traffic.” This is what has been happening to Monroe shipments to and from Madison and Milwaukee. This could be called “The Case of the Descending Spiral.” The present shipping setup is, it is conceded, improvable. The present volume of business is, everybody admits, not very great. Regardless of whether the chicken or the egg came first, the two factors will continue to interact. Several shippers testified that, dissatisfied with the existing common carrier service, they are using their own trucks and even buses for shipments between Monroe and Madison-Milwaukee. The service “tends to drop off” as the traffic declines. The traffic declines as the service “tends to *668drop off.” The status quo predictably will deteriorate. If the streamlining and coordinating of applicant’s service could, as the commission found, “. . . provide a more efficient and economical service to the public . . .” is that not in the public interest considering both the public convenience and necessity and the slender impact of so “thin” an operation of the common carriers involved? I think so and would hold that the Public Service Commission has in fact issued a certificate of public inconvenience, drawing inferences and reaching a conclusion not warranted by the facts established at its hearing. The foundation of facts does not support the superstructure of the conclusions reached.

The legislative mandate to the Public Service Commission is to promote the public interest, considering all existing facilities in determining what will best serve the public convenience and necessity. Where the commission has found that the service to the public provided by an existing facility would be improved by granting an application to streamline operations and coordinate routes, the application should have been granted.