There are no correct answers to the problems listed below, although there may be grounds for applying particular methods, or taking particular options. The aim is to highlight the problems which must be faced by an assessor in determining a fair and equitable tax.
Particular jurisdictions may ignore these problems, or dismiss them as idealistic and over simple. Yet unless the assessor can deal with a very simple case, how can they deal with more complex cases in the real world? Scientific procedure demands that we start with a simple model to make it more readily testable.
Having extensively reviewed the literature on property taxation, I cannot help being struck by the very subjective nature of the assessment process. A property tax based largely on area seems to produce a less fluid and more stable base.
There is often a different degree of transparency in the methods and assumptions feeding into the model used to calculate rental value. Jersey, for example, has no transparency. The greater the transparency, the more scope there is to argue against an unfair assessment.
Consider 3 identical properties, A, B and C.
It is assumed they are in close proximity.
A is rented out at an annual rental of R1
B is rented out at an annual rental of R1
We may therefore deduce that C should be given a putative rental value of R1
Now let us change the parameters slightly.
A is rented out at R1
B is rented out at R2
(perhaps a change in the market for properties of this type)
What is a fair rental value which we may deem for C?
Does it depend upon which of the two (A or B) comes historically later in rental value? In other words, if A is rented at R1, then later in the year B is rented at R2, the current market rental is deemed to be R2?
If so, what if the market has changed, so that if A was re-let now, it would fetch R3, which is again different. In other words,
A is rented at R1, but if re-let would only go for R3
B was rented after R1 at R2, but if re-let would only go for R3
How is R3 determined?
Do we take the average rental for C's putative rental value? (A+B)/2
This assumes the properties are clustered closely, as our original assumption.
Now let us change the parameters slightly.
A is rented out at R1
B is rented out at R2
(the reason is that B is internally in poorer condition that A)
What is a fair rental value which we may deem for C?
Should we assess the internal condition of C? How is this to be done?
Or do we take the average rental for C's putative rental value? (A+B)/2
(which leads us to problems of clustering)
Now let us change the parameters slightly.
A is rented out at R1
B is rented out at R2
(the reason is that tenants of B pay a lesser rent R2, but have to pay overheads O2 which are incorporated in the rent R1, or have some "free facilities" for which they provide a barter/exchange service - e.g. hotel industry)
Is it fair that A and B should be assessed differently, although (say) the rental + overheads for B are greater than the rental of A?
What is a fair rental value which we may deem for C?
Or do we take the average rental for C's putative rental value? (A+B)/2. In this case is C benefiting from B's subterfuge?
Can we overcome B's subterfuge? How?
It may seem fair to consider all the market rentals actually placed on properties and determine putative market rental as the average of these, but problems can still arise.
It is necessary to determine how similar A, B and C are. In our example, we have taken them as identical properties. In a real life situation, they may differ slightly. Perhaps B is slightly larger than A and C is slightly larger than B.
If A1,A2,A3... is in L1 at R1, B1,B2,B3.... is in L2 at R2 and C1,C2... is L3?
An = estate rented out, Bn=estate rented out, Cn=estate home ownership; A,B,C, near identical properties.