Brachtenbach, J.

This is an appeal from a superior court judgment confirming the assessment roll for Local Improvement District 277 (LID) which was assessed pursuant to ordinance enacted by the City of Bellevue (City or Bellevue). We reverse and remand.

In 1987 Bellevue enacted an ordinance ordering improvements to N.E. 4th Street and creating LID 277. The ordinance did not specify a particular method of computing assessments, but rather stated: "In accordance with the provisions of RCW 35.44.047, the City may use any method or combination of methods to compute assessments which may be deemed to fairly reflect the special benefits to the property being assessed." Correspondence, Exhibits, Memoranda and Transcripts (hereinafter Transcript) vol. 1, exhibit 24. The ordinance refers to RCW 35.44.047, but omits a critical word, for the statute authorizes alternative methods "which may be deemed to more fairly reflect the special benefits". (Italics ours.) RCW 35.44.047.

Improvements consisted of widening N.E. 4th Street, an arterial in the central business district (CBD), from four to five lanes for eight blocks and from four to six lanes for four blocks. The project included installation of sidewalks, street lighting and landscaping. Traffic signals were replaced and added. Also, N.E. 4th Street was connected to a newly constructed interchange with Interstate 405. The interchange was not part of the LID improvements. Prior to these improvements, N.E. 8th Street was the CBD's only direct access to the interstate. Eight hundred ninety-four parcels were assessed for 80 percent of the LID costs; the City bore the other 20 percent.

Public hearings were held in 1989 on the proposed assessments. The appellant owners presented testimony from the *401 owners and their experts. The ordinance confirming the final assessment roll was adopted on October 2, 1989. Transcript vol. 1, exhibit 14. Appeal to the superior court followed; the matter was certified to this court from the Court of Appeals.

The issues are:

1. Were the appellants' properties in fact specially benefited?

2. If there were special benefits, did the formula for computing assessments properly determine the amount thereof?

3. Did the City's expert appraiser use proper appraisal principles in reaching his opinions?

4. Were property owners denied due process in the assessment proceedings?

5. Is appellant Tochterman Investment Company entitled to attorney fees?

There are usually two questions in an assessment case: whether the property has been specially benefited; and whether the method of assessing the property is proper. Both questions are at issue here. To analyze these questions in light of the facts of this case, the particular method of assessment used by the City is detailed as follows:

The Trips Generation Assessment Formula

To determine the amount of special benefits, and therefore the assessment for each parcel, Bellevue relied exclusively on the number of vehicle trips (1) generated by existing land uses and (2). which will be generated by future land uses. The formula allocated one-third of the costs to trips generated by existing uses and two-thirds to trips estimated to be generated by future uses of the property. Transcript vol. 1, exhibit 14.

The City offered little justification for the one-third/two-thirds calculation. The only justification given was that City staff originally intended to base the formula entirely on future trips, but after a conference with the Bellevue Downtown Association, they decided to divide the trips. No other validity was suggested, but the City's main witness described *402 the division as "subjective". Transcript vol. 3 (June 12,1989), at 9.

The number of trips generated by a particular use was taken from a publication of the Institute of Transportation Engineers. That document warns that extreme care must be taken in use of the data therein. The publication states that local data should be collected when using the national data, but no local data was taken or presented to the City Council.

Two-thirds of the assessment was based on future trips estimated to be generated from a particular parcel. It is critical to understand that this two-thirds part of the formula is calculated on several pivotal assumptions. First, it is assumed that the property will be redeveloped in the future, regardless of present use. Second, it is assumed that all, we repeat, all 894 parcels, except residential, will be redeveloped for office space. Third, it is assumed that such redevelopment will be to the greatest size and maximum height permitted. In essence, the formula assumes that all of the CBD, except residential sites, will be office buildings built to the maximum size and height. No expert testimony supports these assumptions. Only a staff person, not qualified as an expert, stated that the thrust of new development today is overwhelmingly office, so staff used that assumption. Transcript vol. 1, exhibit 20.

The total LID costs were divided into a trip unit cost. The formula to determine "benefits" from trips generated by existing uses takes the square footage of a building multiplied by the trips generation rate (national, not local) for that use, multiplied times the location or proximity to the improvements, and then multiplied times the trip unit costs. The formula for the two-thirds portion uses the land area and the permitted floor area ratio multiplied by the assumed redeveloped maximum office space, multiplied by the proximity ratio, and then multiplied by the trip unit cost. Adding the two results produced the assessment for each parcel. Transcript vol. 3 (June 12, 1989), at 9-10.

*403 Presumptions and Burden of Proof

Certain presumptions in favor of the assessment arise when a property owner challenges the assessment. They need not be repeated here. Abbenhaus v. Yakima, 89 Wn.2d 855, 860-61, 576 P.2d 888 (1978); Seattle v. Rogers Clothing for Men, Inc., 114 Wn.2d 213, 229, 787 P.2d 39 (1990). However, the principal presumption relevant here is that the improvement is presumed to be a benefit, and "[t]he burden of proof shifts to the City only after the challenging party presents expert appraisal evidence showing that the property would not be benefited by the improvement." (Footnote omitted.) Seattle v. Rogers Clothing for Men, Inc., supra at 231.

The rule is well stated in In re Indian Trail Trunk Sewer Sys., 35 Wn. App. 840, 843, 670 P.2d 675 (1983), review denied, 100 Wn.2d 1037 (1984):

A presumption is not evidence and its efficacy is lost when the other party adduces credible evidence to the contrary. . . . The sole purpose of a presumption is to establish which party has the burden of going forward with evidence on an issue. . . .

(Citations omitted.) The Court of Appeals in Indian Trail then held that the owners' expert testimony there shifted the burden to the City and the City there failed to meet that burden. We agree with the Corut of Appeals' conclusion.

To hold otherwise would make the presumptions in favor of the City conclusive and render the hearing and statutory appeal process on an assessment roll useless. Consequently, the trial court correctly determined the council's decision was arbitrary and capricious and should be annulled.

In re Indian Trail Trunk Sewer Sys., supra at 843.

Here the protesting owners presented oral and written expert testimony before the Bellevue City Council. There were four real estate appraisers and a traffic engineer. There is no challenge to their qualifications nor to their methodology. Those experts were unanimous in their conclusions that (1) the trips generation method of assessment bore no relationship to market value nor to special benefits, and (2) there *404 was no special benefit to the properties involved. Those experts used commonly accepted appraisal methods of location, zoning, access, physical characteristics, and highest and best use. In addition, several analyzed comparable sales.

The burden therefore shifted to the City to prove that the properties were specially benefited. That proof must rest upon competent evidence. It must prove the difference between the fair market value of the property immediately before and after the improvement. Bellevue Assocs. v. Bellevue, 108 Wn.2d 671, 675, 741 P.2d 993 (1987).

Fair market value "means neither a panic price, auction value, speculative value, nor a value fixed by depressed or inflated prices." (Italics ours.) In re Local Imp. 6097, 52 Wn.2d 330, 333, 324 P.2d 1078 (1958) (citing In re Schmitz, 44 Wn.2d 429, 434, 268 P.2d 436 (1954) (quoting Donaldson v. Greenwood, 40 Wn.2d 238, 252, 242 P.2d 1038 (1952))).

With this background, we turn to the first issue, were appellants' properties in fact specially benefited?

Were the Properties Specially Benefited?

A property must be specially benefited by the improvements, as distinguished from a general benefit to the entire district. In re Jones, 52 Wn.2d 143, 146, 324 P.2d 259 (1958). The property owners contend that they receive no special benefits at all. The improvements are presumed to be a benefit. When the owners here produced evidence that there in fact was no benefit, the burden shifted to the City to prove the fact of a benefit, in addition to the amount of the benefit. Seattle v. Rogers Clothing for Men, Inc., 114 Wn.2d 213, 227, 229, 787 P.2d 39 (1990). The City met its burden by showing that in fact there were special benefits to the properties.

None of the properties, except for one parcel of the Bates Oil property, directly abutted the N.E. 4th Street improvements. However, there was testimony that the N.E. 4th Street improvements reduced congestion, provided better access and connected with a new interchange from the state highway. This evidence brings the matter within the *405 holding of In re Aurora Ave., 180 Wash. 523, 529, 41 P.2d 143, 96 A.L.R. 1374 (1935): "That the improvement is a part of an arterial highway from which the public receives some general benefit, does not preclude the idea of special benefit to property particularly served by the improvement." Upon remand, it will be proper to proceed on the basis there in fact was some special benefit to the properties within the LID. It is the amount which must be determined.

The next question is whether the City's expert appraiser used proper appraisal methods in reaching his opinions that appellants' properties were specially benefited. As noted, the burden shifted to the City to establish special benefits.

After the owners' challenges, Bellevue presented the written and oral testimony of its appraiser, Bruce C. Allen. Because the validity of the assessments rests solely upon Allen's testimony, it must be examined in detail. Reference is made to both Allen's written reports and testimony. Transcript vols. 1, 3.

First, Allen's general observations and assumptions are relevant and provide the background to evaluate his conclusions. He states: "It is the position of this report that the entire Bellevue CBD [central business district] benefits from the N.E. 4th Street improvements . . .". Transcript vol. 1 (Sept. 13, 1989), exhibit 6, Executive Summary. He considered the "Traffic Standards Code" and the effect of the LID. That ordinance will be discussed later. Further, "[w]e also consider that, within what we call benefit zones, there will be individual properties that are more benefited than others, based on their potential for redevelopment and assemblage into superblock development." (Italics ours.) Transcript vol. 1 (Sept. 13, 1989), exhibit 6, at 4.

In evaluating each of the benefit zones, Allen heavily stresses assemblage and future development. Regarding zone 2, which includes protester Bellevue Plaza, he states:

Here, the properties will not receive as much special benefit from the NE 4th arterial access as the superblocks adjoining NE 4th, but the improvements will open up substantial devel *406 opment opportunities for the properties on the periphery of the Bellevue CBD.

Transcript vol. 1 (Sept. 13,1989), exhibit 6, at 5. Zone 3, which includes the Tochterman Investment Company property, is assigned a benefit because "further superblock development can occur". Transcript vol. 1, exhibit 6, at 5.

Discussing benefits in general, Allen states:

Within the Benefit Zones, there are areas where this type of assemblage is more or less likely to occur based on the existing uses within the superblock. . . . With the development impetus provided by the NE 4th Street improvements, however, redevelopment of these parcels and some assemblage with other parcels will become more likely. . . . With the potential for assemblage, the marginal value of the individual small properties goes up substantially, particularly where a block is primarily owned by one property owner and there are one or two parcels left to obtain.

Transcript vol. 1 (Sept. 13, 1989), exhibit 6, at 5-6.

Allen attaches a list of a number of land sales within the CBD, but makes no attempt to characterize any one, or all of them, as comparable to any particular property within the LID. Essentially, the reported sales are meaningless as com-parables. In fact, Allen appears to qualify his opinions by stating:

Because the market has long anticipated the improvements to NE 4th Street, precise CBD-wide estimates of before and after property values are elusive based on analysis of comparable sales.

Transcript vol. 1 (Sept. 13, 1989), exhibit 6, Executive Summary.

Several serious flaws in Allen's appraisals are apparent, bearing in mind that the standard is the before and after market value of each parcel. First, quoting Allen, "No attempt was made to appraise an individual parcel, per se." (Italics ours.) Transcript vol. 3 (Sept. 25, 1989), at 51.

Second: "Because of the large size of the CBD, the appraisal methodology develops around land sales only.... Simply land *407 values." Transcript vol. 3 (Sept. 25,1989), at 51. He states that his before and after appraisal

covers an entire zone. But it carries with it the idea of super-blocks will be assembled in the Bellevue Central Business District and average land values will flow out in this manner.

Transcript vol. 3 (Sept. 25, 1989), at 52.

In his oral testimony, Allen confirmed that his appraisal was based on assemblage of parcels beyond present ownership. Referring specifically to Bellevue Plaza, he stated: "Remember that we're dealing with the superblocks, not specifically your property, that can be assembled." (Italics ours.) Transcript vol. 3 (Sept. 25,1989), at 69. He admitted that he had not talked to the owners about assemblage of owned and nonowned property, but "I believe the long term development of that property will occur once that property's put together in one large block." But when asked when that assemblage would occur, he answered: "It could be five years, could be longer." Transcript vol. 3 (Sept. 25, 1989), at 70.

Very telling was an exchange when Allen admitted that he did not know that one of the tenants had a 20-year lease. This was the testimony:

Q: Does that affect your judgment [the 20-year lease] as to length of time that it will take before any assemblage could occur on that property [Bellevue Plaza]?

A: Possibly not.

Transcript vol. 3 (Sept. 25, 1989), at 70-71.

Allen admitted that he had no actual evidence from any seller or purchaser that the price was higher because of the LID improvements. Having testified that the market anticipated the 4th Street improvements and that values had increased 10 to 15 percent, he then admitted that the entire Bellevue area, not just the CBD, had simultaneously increased "let's say seven to eight percent to fifteen percent." It is essential to note, again, that Allen eoccluded all improvements on the properties from his benefit analysis. Transcript vol. 3 (Sept. 25, 1989), at 72-73.

*408 Also, there is the matter of the Bellevue traffic standards ordinance. Allen factored this ordinance into his benefits analysis because the LID improvements would make it easier or less costly for an owner to comply with that ordinance. That ordinance became effective on June 10, 1989; the first LID hearing before the council was June 12, 1989. Allen was not aware that the ordinance in fact was in effect, but admitted that it required further City action to adopt regulations and a means of funding, perhaps a vote of the people. Yet Allen placed great emphasis on the LID improvements as a benefit to owners in their ability to comply with an ordinance he did not know had just been passed. After discussing the traffic standards ordinance, Allen states: "Thus, traffic impacts of new development in the Bellevue CBD is the underpinning of our benefit analysis." Allen was asked to quantify Bellevue Plaza's compliance with the traffic standards ordinance. He answered: "A million dollars. . . . You just [have] to sit back and look." Transcript vol. 3 (Sept. 25, 1989), at 75-76.

Specific Properties — Appraisals

Allen's before and after written appraisals on the three properties for which he gave an opinion occupy, totally, little moré than three pages. We examine them separately.

A. The Tochterman Investment Company property occupies an entire block. Half is zoned 0-2 and half is zoned MU. It must be remembered that the trips generation method assumed future development under the highest zoning with maximum building, as to two-thirds of the assessment, without regard to present use. The property is occupied by a motel, a gas station, several restaurants, several office buildings and parking. Allen considered present uses to be interim uses only; he gave no credence to existing improvements, leases, etc. His analysis is less than one page with a before and after valuation on the land only. Recognizing there are two different zones, he estimated (his word) the before value to be $65 per square foot for the 0-2 land and $55 per square foot for the MU land. He then notes that the City has an option for the "southern" portion of the land to *409 build a convention center. How much is the "southern" portion is undisclosed and no mention is made as to use of the remainder. No terms of the option are disclosed except to note that his values are above the minimum negotiated in the option and this rate is "subject to further negotiation". Despite the speculative nature of this option and the different zoning with different values, Allen concludes:

A blended rate for the superblock is $60.00 per square foot. We further estimate that the value in the after situation is $62.50 per square foot, based largely on the transaction regarding the Convention Center. This indicates a benefit of $2.50 per square foot for the entire superblock, which we believe would be apportioned equally over all the properties in an assemblage.

(Italics ours.) Transcript vol. 1 (Sept. 13, 1989), exhibit 6, at 14-15.

In essence, that is the entire appraisal on a multiuse parcel, appraised by the owner's expert at $22 million.

B. Bellevue Plaza receives little more than a page of Allen's analysis. Allen describes it as a shopping center consisting of a variety of retail shops and restaurants. He then incorporates several parcels not owned by Bellevue Plaza. His appraisal for the assembled superblock, including the properties not owned, is

in the neighborhood of approximately $47.00 per square foot. With the introduction of NE 4th Street and its opening up of access to Interstate 405, we estimate that the value is approximately $49.50 per square foot — for an increase of $2.50 per square foot.

Allen then makes clear the nature of his appraisal:

The effect of NE 4th has been to take a marginally productive property and provide the impetus for redevelopment to the maximum potential under the Zoning Code. Granted, this development will not occur in the next year or two, but the rise in land value captures the speculative nature of the future development potential. The fact that major development is somewhat slowed at this time because of the traffic mitigation ordinance will not affect the development of this property, because we anticipate that the City of Bellevue will be in full reliance at the time this property is suitable for redevelopment.

(Italics ours.) Transcript vol. 1 (Sept. 13, 1989), exhibit 6, at 16. The key phrases, from the appraiser's own words, are *410 that "the rise in land values captures the speculative nature of the future development potential." (Italics ours.) As discussed hereafter, that is impermissible speculation, by its own content.

The appraisal for Bellevue Plaza does not mention its zoning status. Allen stated the before value to be $47 per square foot. Yet in his appraisal for that benefit zone he stated the highest valued zone was MU, with a before value of $45. The maximum after value for MU was $47.25, but he appraised Bellevue Plaza's after value to be $49.50, thus higher for both than the before and after values he stated earlier were applicable to those properties. The inconsistencies are without explanation.

C. The appraisal of the Bates Oil properties is most difficult to comprehend. Allen described it as two parcels located on the Bellevue Plaza superblock. Apparently he appraises those parcels as though they were owned by Bellevue Plaza. He describes another parcel as "located on the block to the north." Part of one of the parcels is occupied by a McDonald's restaurant, under a 20-year lease of which Allen was unaware. Again, Allen's entire premise is that the Bates Oil properties will be assembled with other properties. Once again, the speculative nature of his assumptions is made clear when he states:

Conceivably, this comer property's value could be currently in the neighborhood of $50 to $55 per square foot and could increase more than $2.50 per square foot in an assemblage. At this point, however, we will limit our benefit estimate to the average benefit for the superblock.

(Italics ours.) Transcript vol. 1 (Sept. 13, 1989), exhibit 6, at 16-18.

There are several premises common to the appraisals by the City's appraiser, both in his discussion of benefits to the different "benefit zones" and to the particular properties. First, he ignored entirely existing improvements and existing uses and was unaware of any lease terms. Second, he assumed that parcels owned by different persons would be assembled into superblocks. He was vague about the timing of such assemblage. Third, he did not identify any seller or *411 buyer, or any particular property where the existence of the LID improvements had an effect on the market price. Fourth, he attributed a benefit to the ability to comply with a traffic standards ordinance which had yet to be implemented. His claim that this was worth $1 million to Bellevue Plaza was without any substance. Fifth, he ignored existing uses on the Tochterman properties, assembled two zones into one and then relied largely on an option, the terms of which were undisclosed, to assemble all the existing uses into a convention center on part of the property.

The fundamental starting point for evaluation of the testimony of the City's expert, and its only expert, is clear. "An expert's opinion on the market value of real estate must be based upon those legal principles which define the factors which the expert can or cannot consider in reaching his expert opinion." Doolittle v. Everett, 114 Wn.2d 88, 104, 786 P.2d 253 (1990).

Next, when an appraiser uses a factor "beyond the knowledge of reasonable certainty", it becomes pure speculation. In re Local Imp. 6097, 52 Wn.2d 330, 335-36, 324 P.2d 1078 (1958). The court may disregard the opinion of an expert if he has proceeded on a fundamentally- wrong basis in arriving at that opinion. Doolittle v. Everett, supra at 106; In re Local Imp. 6097, supra at 336.

Determining the amount of the special benefit which may be assessed by reason of LID improvements requires proof of the increase in the fair market value of a particular property caused by the improvements. Fair market value cannot include a speculative value. In re Local Imp. 6097, supra at 333.

As shown above, the City's appraisals were wholly dependent on the assumption of future assemblage of parcels from different owners into superblocks. In the case of the Tochterman property, ownership already existed for a superblock, but the appraisal ignored entirely the existing uses. These assumptions are directly and completely contrary to the extensive holdings of Doolittle v. Everett, supra. While Doolittle was decided after the hearings in this case, there is nothing startlingly new in its rationale. Indeed, one *412 of the basic rules of Doolittle stems from a 1913 Washington case, In re Queen Anne Blvd., 77 Wash. 91, 137 P. 435 (1913). The unanimous Doolittle court stated:

The underlying question is how to determine what is the parcel of land against which an LID assessment may be made when contiguous parcels are in the same ownership, but have been improved and are being used separately.

(Footnote omitted.) Doolittle, at 90. After extensively discussing the determination of the appropriate parcel to be considered in eminent domain proceedings, the court concluded that the same principles apply in special assessment procedures. The court stated: "[MJany courts refuse to consider possible future integrated use in deciding the initial question whether separate parcels constitute a single tract. We agree with this position . . .". Doolittle, at 100.

Further the court noted:

[O]ne leading commentator has nevertheless stated as a general principle that where separate parcels are not actually used as a unit, but instead "simply have such future possibilities," they may not be regarded as one parcel for the purpose of determining benefits. 3 J. Sackman, Nichols on Eminent Domain § 8.6210[1] (3d ed. 1989).

Doolittle, at 103.

These principles of Doolittle are based on cast-iron reasoning. If separate parcels are combined in disregard of present use,

the increase in fair market value will not be attributable solely to the local improvements. Instead, the increase in value will be derived from the local improvements and the combination of the lots, with the owner's actual use disregarded. This has not been the measure of special benefits approved by this court, and it is inconsistent with the principle that the assessment be based on the special benefits resulting from the local improvements.

(Italics ours.) Doolittle, at 103-04.

In this case, there was no effort to determine what was the appropriate parcel for assessment of benefits. We do know that in the case of the Tochterman property, existing zoning and existing uses were ignored. Instead, the appraiser treated at least six separate existing uses as irrelevant, assembled *413 the property into one future use, "largely on the transaction regarding the Convention Center." We do know that in the cases of the other two properties, Bellevue Plaza and Bates Oil, the appraiser ignored existing uses, ignored existing leasehold obligations as long as 20 years, and entirely based his opinion on speculative assemblage into superblocks.

This case far exceeds the facts of Doolittle. There the properties were in one ownership, but the court held as a matter of law that the present use prohibited valuation of the theoretically assembled parcels as one unit. Here the appraiser exceeded that commonsense principle and based his appraisal on the further assemblage of parcels owned by different persons. Despite the common ownership in Doolittle, contrasted to the separate ownerships here, the court held: "The four lots may not be lumped together and the special benefits then determined with respect to possible future use of the whole." Doolittle, at 105.

We affirm the statement in Doolittle that "future use to which property is reasonably adaptable within a reasonably foreseeable time is considered in determining the amount of special assessments." Doolittle, at 104. That principle does not permit the pure speculation of assemblage of separately owned parcels to arrive at present fair market value. We note that the appraiser made no effort to justify such an assumption of assemblage and could reach such speculative potential only by ignoring existing uses and existing leases, and not consulting any owners.

The opinions of the City's only expert were clearly grounded on a fundamentally wrong basis and must be disregarded.

The next question is whether, if special benefits were established (they were not), did the formula for computing the assessments properly determine the amount appellants had to pay?

Assessment by Trips Generation Formula

Bellevue assessed all properties within the LID by the trips generation formula described above. It contends that such methodology is authorized by RCW 35.44.047 and RCW 35.51.030. The historical method of assessment has been *414 zone and termini authorized by RCW 35.44.030-.040. However, RCW 35.44.047 does authorize "any other method or combination of methods to compute assessments which may be deemed to more fairly reflect the special benefits to the properties being assessed." (Italics ours.) See Sterling Realty Co. v. Bellevue, 68 Wn.2d 760, 766, 415 P.2d 627 (1966) (if statutory formula does not fairly reflect the proportionate special benefits, then the authorizing ordinance may specify that the statutory formula will not be followed and an appropriate special benefit formula will be used). RCW 35.51-.030(2) permits the classification of properties according to specified uses and elements, "but in no case may a special assessment exceed the special benefit to a particular property."

Because RCW 35.44.047 requires that an alternative method must more fairly reflect the special benefits, it is incumbent upon the City to make such a finding. The confirmation ordinance makes such a statement, but the record does not support that finding. Bellevue's only appraiser stated specifically that he did not address the issue of assessment methodology. That expert admitted that the trips generation formula was not an appraisal method.

The only witness who attempted to validate the trips method of assessment was the transportation division manager for the Public Works Department. He was not qualified as an expert in valuation or determining before and after values. However, his testimony illustrates the fundamental error in the trips generation method to determine special benefits. He testified:

So, trips become the measure of property's burden on the street system. We take the zoning, the location, translate it into the trips and that becomes the burden on the street system. And so as goes the burden, so must come the benefit. The contribution of traffic to a system must correlate to the benefits assigned to the generator, and that's exactly what we've done.

Further, "[i]n closing, we have a legitimate assessment roll with benefits according to the burden . . .". Transcript vol. 3 (Sept. 25, 1989), at 9, 14.

*415 Other testimony from that city employee illustrates the fallacy of the method. He said:

We took the LID cost here and divided it up into two parts. A third of it goes to the people's existing trips off their property. Two thirds goes to future potential. You sum up all those trips over all the people and divide it into one-third the cost and you get a rate. You take that times an individual's trips and that's their assessment.

Transcript vol. 3 (June 12, 1989), at 2.

Bellevue makes a telling concession in its defense of the trips generation formula. In the trial court, the City's brief contains the following:

Mr. Lema [an owner's appraiser] testified that trip generation is merely a mathematical model that distributes costs.

The City does not deny this. That's what every assessment method does. Square foot, front foot, zone-and-termini, each of them is a purely mindless mechanical model for generating numbers.

Clerk's Papers, at 325. Any formula must ultimately relate to benefits, not merely the distribution of costs. "The critical consideration always is whether the method of distributing cost properly represents benefits to the property assessed." Trautman, Assessments in Washington, 40 Wash. L. Rev. 100, 122 (1965). See Sterling Realty Co. v. Bellevue, 68 Wn.2d 760, 765, 415 P.2d 627 (1966).

The data for trips generated by a particular type of property was taken from a publication of the Institute of Transportation Engineers. The document contains its own limitations. It states:

Because trip generation characteristics for a land use type may vary, extreme care must be taken in the use of the data. . . . Local data should be collected for comparison when considering use of the data in this report.

Transcript vol. 1, exhibit 18, at iii. There was no evidence that local data was used in applying the national averages.

The evidence presented by the City, summarized above, does not demonstrate that the trips generation formula established the special benefit to any particular property as evidenced by a before and after value. It obviously was *416 simply a method of spreading costs, not determining special benefits. This is proved by the City's own witness. After the original hearing, the City made adjustments to 104 parcels. The result was described by the transportation division manager for the Public Works Department:

The philosophical basis was that some of the data we had on the original assessment roll was incorrect, so we got the correct data. And most of the corrections were that — you have my piece of property too big, or you have my zoning too intense. So when you shrink the piece of property, but you're still trying to. collect the same amount of money, the rate goes up.

Transcript vol. 3 (July 17, 1989), at 4.

However, the City's evidence must be measured against the expert testimony presented by the owners. One said: "There is no known relationship between driveway trips and property value." The expert went on to describe the accepted use of the trips generation publication, i.e., to determine mitigation costs for degradation of public streets, such as under a State Environmental Policy Act of 1971 (SEPA) proceeding. He concluded: "The results of such procedures [determination of mitigation measures] have no directly identifiable relationship to increase in land or project value that may result from the off-site street improvements." Transcript vol. 1, exhibit 17.

Another expert testified: "The trip generation method of calculating assessments produces a theoretical number of vehicular trips and distributes the cost among properties in the assessment roll. There is no relationship between the special benefits accruing to the property and the assessment produced by trip generation analysis." Further, "[t]he basic assumption under the trips method is wrong. The market in Bellevue has already demonstrated that this is not true." Transcript vol. 1, exhibit 42.

One appraiser-expert testified there was no appraisal textbook which recognized the trips generation method.

There is another major flaw in the method used. Two-thirds of the trips generation formula was based on future development of all properties within the LID to their highest permitted use under existing zoning. There are six separate *417 land uses within the LID, but it was assumed that all future development would be office buildings, built to maximum size and height, except where land is zoned CBD-R, where residential use was assumed. Transcript vol. 1, exhibit 20. In other words, regardless of present use, it was assumed that every property in what is most of downtown Bellevue would be developed in the future for office buildings, to the maximum size and height permitted.

In electing to use the trips generation method of determining the amount of the assessments, pursuant to RCW 35.44-.047, the City purportedly determined that such method "more fairly reflect[s] the special benefits to the properties being assessed." Assuming, without deciding, that it is presumed that such method in fact does more fairly reflect special benefits, that presumption disappeared when the owners' expert testimony produced evidence that the method had no relationship to special benefits, more fairly or not. It was then the City's burden to prove that there was a special benefit and that the trips generation method more fairly reflected the claimed special benefits. Seattle v. Rogers Clothing for Men, Inc., 114 Wn.2d 213, 231, 787 P.2d 39 (1990). The applicable legal principles are clear.

[Fluture use to which property is reasonably adaptable within a reasonably foreseeable time is considered in determining the amount of special assessments. . . . However, possible future use to which the property is reasonably adapted within a reasonably foreseeable time is to be considered . . . with respect to each of the assessable parcels ....

Further, we express a note of caution to experts who apply the concept of future highest and best use in establishing special benefits in an assessment proceeding. . . .

. . . [A]n owner who is assessed for LID improvements based upon potential highest and best use is forced to pay an assessment on a valuation which may or may not become a reality.

Doolittle v. Everett, 114 Wn.2d 88, 104-05, 786 P.2d 253 (1990).

An expert's prediction of future highest and best use must be reasonable. It cannot be based on speculation. In re Westlake Ave., 40 Wash. 144, 82 P. 279 (1905); Doolittle v. Everett, supra. In fact, no expert testified that this assumed two-thirds future use was reasonable. It is without founda *418 tion. It is sheer speculation to assume that every single nonresidential property in all of downtown Bellevue will be developed into the largest and highest permissible office building. One of the City's documents described the future uses component of the trips formula:

2. Future Uses: Utilizing assumed land redevelopment under maximum FAR [floor area ratio] scenario as per Land Use Code. Only two types of future uses were assumed for private property: (a) Residential Use in CBD-R and (b) Office Use everywhere else.

(Some italics ours.) Transcript yol. 1, exhibit 20. That City document described those assumptions as "subjective".

In summary to this point, we hold that the LID assessments were founded on a fundamentally wrong basis. This is a statutory ground for judicial denial of confirmation of the assessment roll. RCW 35.44.250. There was no evidence presented which showed that the trips generation method "more fairly reflected the special benefits". This is a requirement to use of an alternative method of assessment, at least when challenged by expert testimony. RCW 35.44.047. Apart from that defect, there was no proof which validated the trips generation method. No expert related it to special benefits. Indeed, all the expert testimony found no relationship between trips generation and special benefits. Finally, two-thirds of the assessments were based upon a speculative future highest and best use, totally without evidentiary support.

The City's expert appraisal evidence must be disregarded. The trips generation method of allocating assessments has not been validated as a proper method of showing special benefits to a particular property. The City did not overcome the testimony of experts that the trips generation method bore no relationship to special benefits. Therefore, the City's adoption of the confirmation ordinance was without factual or legal foundation and therefore arbitrary and capricious. It must be and is nullified.

The next questions are whether due process was denied and whether appellant Tochterman is entitled to attorney fees.

*419 Attorney Fees

Tochterman requests an award of attorney fees pursuant to 42 U.S.C. § 1988, contending there was a taking because there is no special benefit, and even if there were a special benefit, it was imposed without due process. Given our holdings, there is no entitlement to § 1988 fees. Tochterman cites Brower v. Wells, 103 Wn.2d 96, 690 P.2d 1144 (1984) as authority, but that case is readily distinguishable because it involved a taking under constitutionally inadequate foreclosure proceedings. It is unnecessary to reach the due process argument.

The assessments on appellants' properties are nullified. There must be reassessment proceedings pursuant to applicable statutes.

Reversed and remanded to the trial court with directions to remand to the Bellevue City Council for assessment proceedings consistent with this opinion and applicable statutes.

Andersen, C.J., and Utter, Durham, Smith, Guy, and Johnson, JJ., concur.