June, 1938] The Transportation of New Hampshire Milk 25 



with volume remaining constant, costs of truck operation will vary ac- 

 cordingly. (Studies have estimated the cost of gas, oil, tires, deprecia- 

 tion, repairs, and so forth per mile of truck operation at slightly under 

 five cents. -^ Such a variation in direct costs would be reflected in rates. 

 Eoutes traveling 40 miles to collect 2,000 pounds of milk would have 

 higher average rates than those traveling 20 miles for 2,000 pounds. 



The extent to which difference in total distance traveled affected 

 average rates may be discovered by classifying routes into groups ac- 

 cording to the volume of milk carried and the distance traveled and 

 comparing the charges on those of the same volume but different dis- 

 tances. Table IV on page 21 shows routes so classified. 



All groups do not evidence the same relationship between distance 

 traveled and average charges. While 21 routes carrying 4,001 to 5,000 

 pounds show a decided tendency towards increased charges with in- 

 creased total distance, seven routes carrying 6,001 to 7,000 pounds show 

 no significant relationship. 



In an attempt to measure the degree of variation of average charge 

 with distance traveled, co-efiicients of correlation were calculated for 

 some of the routes. Averages within each volume group were correlated 

 with average rates, and individual route distances were correlated with 

 rates on individual routes. 



In order that the effects of any factors peculiar to only a part of the 

 routes might be eliminated, correlations were made taking only routes 

 operated by independent contractors in the Connecticut valley. Other 

 separate groupings were made of routes in the Merrimack-Coastal re- 

 gion operated by independents, and routes in the Merrimack-Coastal 

 region operated by dealers. By separations such as these it is believed 

 that homogeneity of observations within the group was attained. 



Correlation co-efificients are contained in Appendix Table IV. With a 

 few exceptions, the degree of correlation of average charge with dis- 

 tance traveled was small. The highest co-efficient of correlation found 

 was only .70 and only a few were above .50. 



It may be concluded that while there is a tendency towards increase 

 of average charges with total distance, the relationship is not close and 

 is far below that to be expected under conditions of perfect competition. 



V. Relationship between average charge per route and total volume of 

 milk carried. 



In the preceding section an attempt has been made to measure the 

 relationship of average charge per route and total distance traveled, 

 keeping volume constant. Another comparison of importance is that 

 of routes of similar distances but differing volumes of milk. The same 

 technique as that used in the preceding section has been employed and 

 typical findings are shown in Appendix Table IV. 



Costs of trucking increase only slightly as the load increases. A 

 load of 4,000 pounds does not double costs over one of 2,000 pounds — it 



" Bartlett, R. W., and Caskey, W. F.; Milk Transportation Problems 

 in the St. Louis Milkshed; Illinois Agricultural Experiment Station Bulle- 

 tin 430, p. 448. 



Dow. George F.; Costs and Returns in Operating Milk and Cream Collec- 

 tion Routes in Maine; Maine Agricultural Experiment Station Bulletin 374, 

 p. 154. 



