FACTORS AFFECTING THE" PRICE OF HOGS 35 
What is really needed in getting a demand schedule is a statement 
of the relation between price and consumption — that is, a statement 
of the quantities which would be consumed during a specified period 
if offered at various prices. Such a statement could be obtained by 
a direct study of the relation of monthly consumption of pork and 
pork products to pork prices and other factors. Lacking such a 
study, the demand schedule must be inferred from the relation of 
the market receipts or slaughter to prices and other factors, as was 
indicated in assumption 5. 
Before stating the means by which this has been done, it is worth 
stating the relative importance of the different factors affecting the 
price, as shown by the multiple correlation just discussed. In the 
system of variables shown, the relative importance of each factor was 
as follows: 
Per cent 
Xi Monthly slaughter 5. 1 
X 2 Slaughter for six months previous — 5. 
X 3 Six-months moving average of slaughter 9. 
X4 Storage stocks 19. 3 
X 5 Business activity — 1. 4 
X 6 Population of United States 27. 2 
X7 Price of steers 1. 6 
X 8 European demand 2. 7 
X 9 Time (changing consumption habits) — 5. 9 
X10 Bureau of Labor all-commodity wholesale price index 35. 1 
Total determination by all factors (R 2 ) 87. 6 
The values preceded by the sign ( — ) are those where the joint 
relation with other variables was such that the net regression with 
price was of opposite sign from the gross correlation. This, however, 
does not mean that they were not of significance as factors influencing 
price; it merely means that their effect on price was, on the average, 
in an opposite direction to the way the price actually moved, rather 
than as a direct contribution. In general, these determination 
coefficients, taken without regard to sign, give a measure of the rela- 
tive importance of each of the different factors, on the basis of the 
specified system of relations. 16 
These coefficients of determination are statements of the propor- 
tion of the total variability of the dependent factor, hog prices, which 
for the period studied can be mathematically related to the variability 
of each of the other factors included in this particular study. It 
must be remembered that factors not included in the regression equa- 
tion may yet influence price ; if they do so either through or concomi- 
tantly with any of the independent factors already included, their 
effect is already included in the coefficient for such variables. It 
must also be remembered that these coefficients are computed while 
simultaneously allowing for the effect of the other variables stated, 
but not taking into account any of the other possible factors affecting 
prices. Dropping any of the significant factors from the group, or 
adding an additional factor to those considered, would cause a shift- 
ing in the computed net relation of the other factors to hog prices 
and hence some change in the values of the coefficients of determina- 
tion. 
16 These coefficients are computed by the following formula: 
Determination of Xn by Xj, in the system of relations described^ (bn, j: 1, 3,4, 5, 6, 7, 8, t, 10) (rn, j) ; deter- 
mination by X3=(bn, 3: 1, t, i, e, e, 7, 8, 9, 10) (rn, 3), etc. For a fuller statement of the meaning of these 
•'coefficients of determination," see Smith {21, p. 4£). 
