130 MISC. PUBLICATION 218, U. 8. DEPT. OF AGRICULTURE 
In Cumberland County, Pa., a study of assessment ratios for 1926 
showed coefficients of 29 for town properties, 29 for village properties, 
and 36 for farm properties. By 1929 a special effort to improve the 
quality of assessments in this county resulted in corresponding 
coefficients of 25 for town properties, 24 for village properties, and 
33 for farm properties ** (56, pp. 29-380). 
In Delaware, assessment ratios, grouped into farm, town, and city, 
had coefficients of dispersion for 1921, 1924, 1926, and 1927 ranging 
from 21 to. 41 °° 6,/9. 27). 
In general, wherever precise studies such as those cited have been 
conducted, they have confirmed the well-known fact of marked in- 
equality among individual assessments. 
INEQUALITY AND THE LEGAL STANDARD 
Undervaluation relative to the legal standard is a generally recog- 
nized characteristic of assessment practice. Evidence on this subject 
is available in reports of assessment ratio studies, such as those specifi- 
cally referred to at other points in this part, and in many State 
tax commission reports. Undervaluation is also indicated by the 
State ratios of assessed value to estimated true value found by the 
United States Bureau of the Census (51, p. 5). 
It is reasonable to suppose that the prevailing tendency to aim at 
an average assessment much below the legal standard leads to greater 
inequalities in assessment than would otherwise occur. If an assessor 
is generally valuing properties at 30 percent of the legal standard, 
discrepancies between individual values are less noticeable and their 
correction is less likely to be insisted upon than if his general level is 
80 or 90 or 100 percent. If this deduction is true, the political units 
with the highest assessment ratios should show the oreatest equality, 
while those with the lowest assessment ratios should show the highest 
degree of inequality. That such is the case is indicated by table 42, 
which is a comparison of the 4 political units represented in tables 
26 to 29 which have the highest assessment ratios, 90 or over, with the 
4 units having the lowest assessment ratios, all of which are under 
50. (In making the selection of these political units, the Oregon 
counties with assessment ratios based on unverified transactions were 
omitted.) The variation in each group is indicated not only by the 
coefficient of dispersion, but also by the percentage of the number of 
individual assessments that differ from the average by 20 percent or 
less (column 4) and by more than 50 percent (column 5). The coun- 
ties with assessment ratios close to 100 percent (group 1) have gener- 
ally coefficients of dispersion that are low compared with those of the 
counties with assessment ratios far from the legal standard (group 2). 
The only exception is Forest County, where the coefficient of dis- 
persion is low because of the weight of several large timber properties 
with assessment ratios not very far from the average. The counties 
with assessment ratios close to the legal standard are also better 
assessed than those with assessment ratios far from the legal standard, 
if the percentage of the number of cases with either low or high de- 
viations from the average is taken as the criterion (columns 4 and 5). 
34 Tt is not clear whether or not the coefficients in this case are weighted by value. If unweighted, they 
are not exactly comparable to the coefficients used in this study. 
35 In this case the coefficients were unweighted by value and therefore not exactly comparable to those 
used in this study. 
