292 THE FISCAL PROBLEM IN MISSOURI 
ratio as computed is in a sense a weighted average, that is, it 
indicates the relationship between total taxes and total 
profits of all corporations. It does not follow, however, that 
the average is typical. Each year a number of corporations 
report net losses rather than net profits, and the larger the 
number and size of such concerns, the less representative 
the ratio becomes. For example, since certain Missouri 
corporations reported net losses, it follows that a large 
number of corporations must have had profits much greater 
than $1.00 for every $0.38 paid for taxes. Even though each 
corporation reporting from Missouri had a net profit, the 
ratios for individual corporations would doubtless vary 
widely. It might be found that there was no concentration 
at or very near to the average computed from the combined 
data. Another possibility is that by means of an analysis 
of individual corporation returns it could be shown that 
there was a definite modal group comprising corporations 
with ratios of not less than 259, nor more than 309, and 
that the average of 389, was accounted for by a compara- 
tively few corporations with ratios much higher than the 
average. If the latter should be the case, the mode would be 
much more significant than the average. 
A third reason for the limited conclusions that can be 
drawn from such data is the fact that variations within a 
state are no doubt usually much larger, even in the same class 
of industry, than the variations among the averages for the 
several states. In many states property tax assessments are 
decidedly variable, and, since property taxes ordinarily 
account for the largest part of state and local taxes paid by 
corporations, the reasons for the above statement should 
immediately be apparent. In view of the importance of the 
property tax item and the variation in assessment ratios, 
it would seem to be unwise to attach any significance to 
interstate comparisons that are based upon data in the form 
of averages without first ascertaining the extent of the 
variation within a particular state. It would, however, be 
permissible to make comparisons between cities in several 
states, if comparable plants could be considered and all of 
the necessary information obtained. 
As yet no wholly satisfactory approach to the problem of