Richard Boehm hasn't changed a thing in his N Highview Avenue house since it was built in 1992 except move some furniture.
Yet his assessment rose from $111,100 in 1993 to $133,700 last year, a 20 percent increase. His tax bill leaped from about $1,638 to $2,090.
"I just want to be treated like everyone else," said Boehm, 70, a retired executive for a smelting firm. "Why pick on me?"
With Boehm's permission, the Times asked Property Appraiser Ron Schultz to lay out how he assessed Boehm's house. (See chart.)
The exercise provided an inside tour of a complicated method Schultz introduced to Citrus County last year to estimate what every home in the county would sell for. That prediction forms the basis for a home's assessment and tax bill.
The formula used a mazelike computer program that crunches the sales prices of similar homes in the same neighborhood that were sold recently.
Until last year, Schultz used a more traditional method based largely on a house's replacement cost.
His use of the computer program helped fuel the large increases in assessments some homeowners saw last year.
Newer homes like Boehm's got a big hit, for example, because they had been assessed too low by his former technique, Schultz said.
But overall, the computer estimates are considered more accurate than those produced by the replacement-cost approach, according to several appraisal experts. Computer programs are widely used by county appraisers, especially in urban areas.
"The methodology is sound," said Susan Wachter, associate director of the Wharton Real Estate Institute at the University of Pennsylvania.
Over time, Schultz and the experts agreed the new approach should make the tax roll fairer by eliminating some of the high and low assessments in the tax roll.
Some people, like Boehm, distrust computers, especially when it comes to valuing their house.
"I don't think a computer can ever replace a human," he said. "The computer goes ping, ping, and out comes a number."
However, Schultz's office would quickly run out of money if he tried to do the kind of in-depth appraisal that, for example, banks require for mortgages. A commercial fee appraiser typically charges $300 a house for that kind of precise estimate.
With a budget of $1.4-million to value 141,000 lots, Schultz has only $10 a parcel to work with.
So he and other county appraisers try to estimate all the homes' values using a sample of the recent sales.
Still, Boehm wonders how a machine could account for differences among houses. He asked, "How can you take another house on another street in my neighborhood and tell me it's exactly like mine?"
Schultz and the experts agree comparisons among houses aren't perfect, but the computer program does a reasonably good job at a reasonable cost within a limited amount of time.
Here, generally, is how the computer values your home:
Schultz has selected 10 characteristics of homes that in his judgment are reliable predictors of a home's selling price. These include things common to all homes: square feet of living space, number of bathroom fixtures, age and style of the house and whether it has desirable frontage, say facing a lake or golf course.
The computer assigns a starting value to each characteristic, then tallies up those values. Let's say the number it reaches is $10,000.
It then compares its result with the price of each home that sold recently in the neighborhood. Let's say that number is $50,000.
The computer's tally doesn't match the price, so it adjusts the values it has assigned to the 10 characteristics. It makes those adjustments by analyzing the houses in your neighborhood that sold recently.
For example, it arrives at an average square-footage figure and assigns a value to that. If a house is above or below that average, the value of that characteristic for that house will rise or fall.
The program keeps making adjustments until it has a set of values it can use to predict the market value of any home in that neighborhood. That figure then is used to determine your assessment.
Schultz had the computer run those numbers last year for 77 neighborhoods. He has identified more neighborhoods but was unable to run the computer calculations there last year because they lacked enough recent sales to provide a statistically reliable sample for analysis.
But by next year, he expects he'll have enough data to do all the neighborhoods. Schultz is required by state law to update property values annually.
have their limits
Schultz and the experts agreed, however, that any computer-based model has limits. It will never be perfectly accurate.
The more similar the homes in a county, the easier it is for the computer to compare them, and the better its predictions.
But Citrus has a varied mix of new and old housing, some in new golf-course communities, some in older waterfront areas.
"We have too much variation within the community" to be perfectly accurate, said George Donatello, Schultz's deputy appraiser.
Moreover, appraisers make a subjective choice when they decide which variables to study with the computer. That subjectivity can be reduced, but an appraiser needs several years of practice to do that.
Some experts were surprised Schultz didn't replace the old assessments with the new computer-driven ones all at once.
Instead, as the accompanying illustration shows, Schultz mixed the computer's figures with the replacement-cost method in a way skewed toward the cost method.
Generally speaking, the computer-generated market values "should stand on their own. . . . If they're the market value, why wouldn't you use them?" said Wachter of the Wharton Real Estate Institute.
However, Schultz's caution was warranted, according to Bruce W. Sauter, director of valuation services for the New York State Board of Equalization and Assessment. "The first time you use it (the computer program), there's no way to avoid volatility," he said.
Schultz said he will give the computer-based numbers more weight as he gets more practice using the program and working out its kinks.
"This is more than a one-year process," he said. "If you study us three years from now, I think we will have improved."
Another reason Schultz set up the formula the way he did was because he expected the computer-based figures would boost assessments on some Citrus homes. He mixed the two methods to soften the increases.
That wasn't much comfort to Boehm. "Don't shoot me in the leg and tell me, "I could have shot you in the head,' " he said.
After complaining last year, Boehm got Schultz's office to revise the value of one of the variables, which knocked his assessment down to $124,200.
That's still a 12 percent increase over 1993, but it brought his tax bill down to $1,902.
As much as Boehm might grumble about it, the computer program generated an estimated market value for his house that seems to be in the ballpark.
Boehm said he bought his lot for $15,000 in the late 1980s and spent $155,000 to build the house in 1992. The computer estimated his house would fetch $179,942 _ the market value.
"I'm sure it would never sell for that," Boehm said.
Boehm still thought his assessment was a little high, but he didn't formally appeal. "I figured I wouldn't do any better."
THE APPRAISER'S DICTIONARY
The property appraiser uses a step-by-step method of figuring your property's assessments. Here are some of the terms, and their definitions, used in that process:
Market value: Also called appraised value, this is the estimate of what your home would probably sell for as of Jan. 1 of each year. The property appraiser starts with this figure in preparing your home's assessment.
Assessed value, assessment: In Florida, assessed value is defined as market value minus the cost of sale.
Cost of sale: The state allows a deduction from the sales price, ideally 15 percent, to cover the cost of sale. This figure includes a range of expenses involved in selling a house, such as broker's fees.
Taxable value is assessed value minus exemptions, such as the $25,000 Homestead Exemption. Your tax bill is figured on this number.
TRIM notice: The Truth in Millage notice is mailed out in August and lists assessed and taxable values from the year before, and proposed 1994 taxes.
Tax bill: Mailed in November, your tax bill is figured by applying the millage rates set by the County Commission, School Board and state water district to the taxable value.
Jeffrey Brainard, 32, has been a reporter for the Citrus Times since August 1993. He lives in Inverness and covers Citrus County government. Previously, he was a reporter for Newsday on Long Island and newspapers in Middletown, N.Y., Albany, N.Y., and Northampton, Mass. A native of New York City, he received a bachelor's degree in philosophy from Williams College.
T. Christian Miller, 24, started working for the Citrus Times the day five inmates broke out of the Citrus County jail in February. Miller lives in Floral City and covers business and general news in Citrus. He has worked as a reporter for the San Francisco Chronicle and the Los Angeles Times. After growing up in Charleston, S.C., Miller attended the University of California at Berkeley, where he received his bachelor's degree in English.
How one home's assessment was figured
If you don't understand how Property Apraiser Ron Schultz figured your assessment, you are not alone. He used a statistical technique that makes balancing your checkbook look like child's play. However, for those curious enough, and not numerically faint-hearted, here is how Schultz's formula works, step by step. The formula is a mixture of the results of two different appraisal techniques, the cost-based and market-based methods. We offer as an example the home of Richard Boehm of Citrus Hills (shown at right), who volunteered for the exercise. His three-bedroom, two-bath house was built in 1992.
The cost-based method
Compared with the other parts of Schultz's formula, this one is relatively a breeze. The appraisal is made up of two parts: the building's value and the land's value.
Under the cost approach, the building's value is calculated using standard formulas derived from the construction industry. Those formulas show, in current dollars, what it would cost to build the house based on its dimensions and design.
Replacement cost, Boehm's home $104,950
Minus depreciation $2,950
Total, building value $102,000
Land value is estimated by a county appraiser based on recent sales in the neighborhood.
Estimated value, Boehm's land $15,200
X 89% to deduct cost of sale X .89
Total, land value $13,528
Total, building and land values $115,528
X 128% (1) X 1.28
Total, cost-based method $147,875
Mixing the methods
All the calculations for the market-based figure are only slightly reflected in the final assessment figure. Schultz deliberately weighted the market figure less to avoid larger increases in assessments.
Boehm's assessment = 80%
(cost method) + 20% (market method) X .805 (5)
.80 ($147,875) + .20 ($179,942) X .805 = $124,200 GRAND TOTAL
The market-based method
Using a computer, Schultz did a complex set of calculations on most homes in Citrus County to estimate what they would sell for today. Below are a sample of the calculations for Boehm's house.
Schultz broke down the home into component parts. Then he assigned a general value to each part based on how homes in the neighborhood were selling. Last, he massaged the values with adjustment factors to reflect the specific size and traits of Boehm's house.
Here's what the headings in this chart mean:
Variables _ the component parts Schultz used, such as square feet
Boehm's house _ the actual values for Boehm's house
Boehm's neighborhood _ the values for Citrus Hills, where Boehm's house is located.
Adjustment factors _ used to fine-tune the estimate
Boehm's Boehm's Adjustment
Variables house neighborhood factors
Living space 2,648 X $29.65 / X 97.9% = $76,864
sq. ft. sq. ft.
Other space 1,394 X $28.84 / X 100% = $40,202
(garage, etc.) sq. ft. sq. ft.
Miscellaneous features $12,710 X 156.6% = $19,904
(pool, shed, etc.) (cost)
Bathroom fixtures Six, at $630 X 99% = $3,742
(sinks, baths, toilets) each
(3) x 122%
Land $15,200 (2) X 92.3% = $14,030
X 96.9% (4)
Total market-based method $179,942
1. Schultz multiplied by 128% to correct for differences between the cost-based appraisal and actual market conditions. That correction is based on a sales ratio, meaning the ratio of all the cost-based values of homes sold in Citrus County in 1993 to all the sales prices of those homes. The cost-based values tend to be lower than the sales prices.
2. Land value is estimated by a county appraiser based on recent sales in the neighborhood.
3. An adjustment factor that reflects the size, age, type and location of Boehm's house and the quality of construction materials.
4. A correction for the difference between the market-based method and actual market conditions. The correction is based on the ratio of all the market-based appraisals on homes sold in Boehm's neighborhood in 1993 versus all the sales prices of those homes.
5. Schultz used this adjustment factor countywide to deduct the estimated cost of sale, which is included in a property's market value but is not taxed. The state Department of Revenue prefers an adjustment factor of 85%, but allows a factor as low as 76.5%
Source: Citrus County Property Appraiser Ron Schultz