Virtually, every taxing jurisdiction relies on self-reporting of property additions and deletions. Annual reports are submitted. Municipalities then keep records by taxpayer showing on a cumulative basis, by year of acquisition, what they consider to be the current value of the taxpayer’s overall assets.


To arrive at an assessment value, which is supposed to be at FMV, most tax  jurisdictions apply “COST INDEXES’ to prior year acquisitions, in an effort to show what the ‘current cost’ is of the assets still on hand.

Implicit is the assumption that the current cost, based on indexed original cost is the fair market value (FMV) of the assets.

Inasmuch as retirements have to be reported by year of original acquisition, the municipalities’ records purport to show the total current value of all assets owned by the taxpayer as of the reporting date. A moment’s reflection reveals that the accuracy of the tax assessors’ values is going to be a function of the accuracy of the cost indexes and the accuracy of the asset detail submitted by the taxpayer. Errors in either, or both for that matter, will cause companies to pay more tax than they need to.

Cost Indexes represent a shortcut to estimate current FMV. The basic assumption of a cost index is that it purports to reflect changes in selling prices by vendors, purchase prices that would be paid today, were a company to acquire all prior assets at today’s costs.

The construction of a cost index is both simple and complex, as shown by the federal government’s attempts to develop the consumer price index that impacts such things as Social Security payments.

It is simple because it takes a standard complement of assets each year and obtains current price quotes for each item in the index. Weights are usually assigned to individual components based on relative purchases, or use, of the specific item.


Let’s assume a simple fruit and vegetable price index. Every year at the same time (to avoid seasonal influences) the person constructing the index would obtain prices from the local supermarket of carrots, potatoes, and lettuce. The relative weighting of carrots in the index would reflect the relative purchases of carrots by consumers in relation to all other fruits and vegetables a typical family buys.

So the first year of the index construction the statistician would determine typical purchase patterns (or sale patterns if this was to be a store index not a consumer index) of vegetables. If this were a national index then samples would have to be obtained from areas throughout the country since prices of oranges in Florida might differ from oranges in Alaska.

Whatever the final complement of the index, the maintenance of the index on a yearly basis would require getting new price quotes each year, assuming that relative consumption would not change. In other words, even if the price of oranges went up, and the price of lettuce went down, the index assumes constant relative consumption and that people would not substitute grapefruit for oranges or tomatoes for lettuce.

To the extent that relative purchases are not directly influenced by price changes such an index can faithfully reflect price changes in fruits and vegetables consumed by the average family. In the real world, however, prices do change behavior and this makes construction of price indexes quite difficult. In fact, statisticians devote entire lifetimes to refining indexes.

Ones, at this point, may ask themselves if the price of lettuce went up—because of a drought in Calif—but tomatoes produced locally had gone down, would we not use more of one and less of another?

The answer, of course, is that prices do drive consumer behavior. But if the statistician rebalanced the index every year, it would be hard to have a price index; rather you would have an index of how much consumers spend on fruits and vegetables—which is not the same thing as a price index.

Now, to carry the analysis one step further, it is easy to assume that fruits and vegetables hardly change in content or format. A bunch of carrots from ten years ago is going to look identical to a bunch of carrots bought yesterday. A change in price from carrots ten years ago to today undoubtedly reflects only a change in price.

Well, those are fruits and vegetables. In fact, not everyone live in the same business. Keeping consumer expenditures in mind, what if we are trying to develop a price index of computer desktop sets?

Okay. Five years ago very few flat-screen display were sold, and the few flat-screen sets available were very expensive. Over the five-year period, the price of computer desktop sets has gone down—everyone will have either a plasma or LCD set, with size depending on one’s budget. So, how do you develop a price index when the item itself is changing because of technology and customer demand? To put it in a nutshell, developing price indexes of technology items requires a great deal of professional judgment—which is a rare knowledge for the rest of us.

The federal government has the same problem in evaluating the consumer price of new cars. How much of a price increase is attributable to a new GPS system now made standard?

Ten years ago no cars had GPS, now perhaps 50% have them. Undoubtedly, some of the total price increase in autos over that time period reflects new features, and the remainder of the price index reflects true price increases on a ‘‘like for like’’ basis.

In this post I am not trying to make you, the reader, into a statistician, only to point out in easily understood terms the real difficulties in developing, and then using price indexes.

With respect to property taxes, municipalities do apply readily available indexes to all the types of assets owned by taxpayers. But can a price index capture price changes in machine tools where computer controls increase the price but also increase the productivity?

The answer is that price indexes do a good job where changes in technology are absent, or even quite limited; indexes do a poor job where technology and productivity changes are great. This digression was to put the problem of indexes into place, because of the tremendous reliance that municipalities place on the use of indexed original cost. Indexed costs can approximate FMV in certain cases, but in others are more than likely to overstate assessed value relative to true FMV.

Municipalities do make one further adjustment. They apply a depreciation factor to the indexed historical cost to allow for the diminution in utility as an asset gets older and suffers physical wear and tear. Again, as in our discussion earlier about companies using tax depreciation lives for financial reporting, ‘‘rules of thumb’’ or arbitrary lives used by tax authorities simply do not (and cannot) measure actual reductions in value due to physical wear and tear and functional obsolescence.

The end result of these adjustments made by property tax authorities is their own estimate of FMV, not what the company itself might consider its true value. You should keep in mind that in virtually all taxing jurisdiction property taxes are supposed to be collected based on an ad valorem1 approach.

As both statutes state, and court decisions have held: Fair Market Value is defined as the price for which property would exchange between a willing buyer and a willing seller, each having reasonable knowledge of all relevant facts, neither under compulsion to buy or sell, and with equity to both.

The question then has to be asked: if you start with the original cost of an asset (purchased years ago)—and have adjusted it upward by a cost index and downward by a depreciation factor, how close is the resulting amount likely to be to today’s FMV?