VII. THE MONTH OF NOVEMBER IS NOT A FAIR AVERAGE MONTH IN ANY RAILWAY YEAR OR ONE THAT IS TYPICAL OF A YEAR'S BUSINESS AND ITS USE AS THE SOLE BASIS OF THE POSTMASTER-GENERAL'S CALCULATIONS WAS SO UNFAVORABLE TO THE RAILWAYS AS TO DEPRIVE THE RESULTS OF ANY VALUE EVEN IF IN ALL OTHER RESPECTS HIS METHODS WERE BEYOND CRITICISM.

All the Postmaster-General's calculations, reported in Document No. 105, and by him relied upon therein, and elsewhere, to substantiate his attack upon existing railway mail pay, depend solely upon data for the single month of November, in the year 1909. It is obvious, therefore, that the validity of his conclusions, if all the rest of his processes were accurate and his deductions otherwise sound, would depend upon whether November is sufficiently typical of the railway year to be safely used as the sole basis for conclusions applicable to a whole year. The truth is, however, that November is not a typical or average month and that all of its deviations from the averages of the year are such as greatly to favor the result which the Postmaster-General was seeking.

It may well be doubted whether the railway year contains any month that can properly be regarded as typical of the whole period but if it does, the month of November, with four Sundays, two holidays and only twenty-four working days, is certainly not such a month. The Interstate Commerce Commission publishes the monthly aggregates of railway receipts and these official data conclusively prove that November, 1909, was the one month for which the data were most strongly favorable to finding, by the Postmaster-General's method, an abnormally low apparent cost for the passenger train services and, consequently, for the mail service.

It is a month in which substantially Winter conditions prevail in a large part of the country and, on this account, one during which much of the ordinary work of maintenance of way and structures must be suspended. Such work occasions a large fraction of the yearly expenses of all railways and these expenses pertain in a relatively large proportion to the passenger services because the higher speed of passenger trains results in greater relative wear and tear upon road-bed and structures than that caused by the slower trains of the freight service and the requirements of safety to passengers carried at high speed impose more costly standards of maintenance than would otherwise be necessary. Consequently a month in which these maintenance expenses are necessarily below the yearly average cannot typify the full annual cost of the passenger train services. Figures showing the facts are contained in Appendix B.

It is, of course, understood that the respective expenses of the passenger and freight services must move upward and downward with the fluctuations in the volume of each sort of traffic. No month can furnish a reliable basis for estimating the proportion of the total expenses that is caused by the passenger service unless during that month the volume of passenger traffic bears a normal relation to the volume of freight traffic. But in November, 1909, as will appear from official figures for each month in the year contained in Appendix C, passenger traffic, as measured by receipts therefrom, was much below the average month of the year while freight traffic was far above the average. The November receipts from passengers amounted to only 21.5 per cent. of total receipts, the lowest relation shown for any month in the year. Of course, under these conditions passenger expenses were curtailed and freight expenses relatively enhanced. Certainly the use of data resulting from these abnormal relations could not possibly produce results fairly typical of a normal period, that is of a whole year. The results so obtained must have diminished the apparent cost of the passenger train services below the true cost, by just as much as the figures for November were below the average figures of the year.

These considerations fully establish the truth that, if every other feature of Document No. 105 were absolutely beyond criticism, the fact that it rests wholly upon estimates based upon data for the single month of November would render its conclusions illusory, misleading and seriously prejudicial to the railways.