112 MISC. PUBLICATION 218, U. 8S. DEPT. OF AGRICULTURE 
In Richmond they worked entirely independently in order to economize time, 
and the results in this town must be regarded as subject to the differences in 
judgment which practically always occur between two or more qualified appraisers, 
especially where there is no background for their work in local-valuation data 
systematically collected and compiled. 
In the town of Loudon the appraised values of the land, exclusive of buildings, 
were computed by means of a statistical formula. This formula took into 
account the principal factors affecting value and was based on values of 134 
individual properties, as determined by one of the same expert appraisers who was 
responsible for the individual appraisals in Fremont and Richmond. Also these 
appraised values were tested by the available information in regard to sales and 
were found to be in accord with selling prices. The larger number of basic values 
made the use of the standard statistical method more desirable than the inspec- 
tion method of correlation which was used in Minnesota and Wisconsin. The 
building values were approximated by a formula which took into consideration 
the cost of production and depreciation, as well as the obsolescence owing to over- 
building and other causes. 
The elements of real-estate value in Loudon were improvements, merchantable 
timber, and bare land. The improvements consisted largely of farm buildings, 
and the merchantable timber was principally second-growth white pine, 30 to 50 
years of age. The unmerchantable timber (young growth) was included with 
pare land. In all of these appraisals land and improvements were valued sepa- 
rately. 
After experimenting with the various possible factors affecting timber and land 
values, the following were combined and treated by multiple correlation: 
X o=value per acre of total area. 
DF — vouumne of merchantable timber per acre of total area in thousand board 
eet. 
x — xX 1°. 
X3;=ratio of area in crop land to total area plus 0.45 times ratio of area in pasture 
to total area. ; 
X,=ratio of area in forest to total area. 
n=age of the unmerchantable timber in years. 
b;-b4 =the unknown unit values, to determine which the correlations were carried 
through. 
The equation became: 
Xo=6,X;+0.X.+ b3.X3+ b4(1.04) aX,. 
The introduction of (1.04)* implies that the unmerchantable timber increases in 
value by 4 percent of the total value of land and timber each year, an assumption 
which was checked by inspection of the appraisals. It was found, however, that a 
slight alteration of timber values would improve the accuracy of the formula. 
When the residuals from the unaltered formula were plotted against timber 
volume per acre, the freehand curve drawn indicated that the low and high 
volumes per acre were overappraised and that those in the middle range were 
underappraised. Adjustments were made for X,; and X2 by substituting readings 
from a second freehand curve drawn in terms of unit value per thousand and volume 
per acre for the values computed in the original formula. 
SUMMARY OF ASSESSMENT RATIO DATA 
The data obtained in original studies of assessment ratios are sum- 
marized in tables 23 to 29. These tables show the assessment ratios 
of the principal groups of properties with which these studies are 
concerned, together with the aggregate assessed values and aggregate 
estimates of true value on which these ratios are based. They also 
show the areas of land in each group for which this information is 
available. The number of properties and coefficients of dispersien 
are given for the studies in which the individual properties were 
valued by considerations realized in sales, or by fairly intensive 
appraisals. The more important of these figures will be repeated in 
smaller tables in connection with the following discussion of assess- 
ment results. 
