Howdy

Here is an update on the distribution of box and auto cars on the UP between Laramie and Rawlins in 1938, with tables that show the cars classified by the ICC region of the owners. This analysis and my previous comparison of data for specific railroads (# 89886) lead me to conclude that the G-N hypothesis does not work very well for the UP in 1938.

The first table lists the names, abbreviations, and number of box/auto/ventilated cars of the ICC regions. The number of cars is for cars in interchange service on ICC Class 1 railroads and their lessees only and is based on the January 1938 ORER.

ICC Region: Abbrev; ORER N of cars
Central Eastern: CE; 130,863
Central Western: CW; 144,296
Great Lakes: GL; 156,429
New England: NE; 23,947
Northwestern: NW; 126,201
Pocahontas: POC; 18,694
Southern: S; 100,611
Southwestern: SW; 56,970
TOTAL: 758,011

The next table gives the number of cars listed in the three Freight Conductors' Train Books by Fitz, Ferguson, and Fraley, classified by region.

Abbrev: N of train book cars
CE: 145
CW: 1,408
GL: 269
NE: 3
NW: 218
POC: 16
S: 73
SW: 111
TOTAL: 2,243

In addition to the cars shown in the table above, the train books list 5 cars from the CN and 19 others whose reporting marks were illegible or could not be found in the ORER. There are thus a total of 2,243 + 5 + 19 = 2,267 box/auto cars reported in the train books.

Data from these tables may be combined to provide a test of the G-N hypothesis of freight car distribution. My understanding of this hypothesis is that it predicts that the proportions of cars observed in a sample of real trains should approximate the national proportions. However, the home road cars should be removed before comparing the proportions.

In this case, the home road is the UP, part of the CW region. There are 27,624 UP box/auto cars listed in the 1938 ORER and 935 in the conductors' train books, so these should be removed from the appropriate tables. This leaves 116,672 and 473 cars in the CW region for the ORER and train books respectively, with totals of 730,387 and 1,308 cars. Once this is done, the percentages of cars in the ORER and train books are as follows:

Region: ORER, Train books
CE: 17.9%, 11.1%
CW: 16.0%, 36.2%
GL: 21.4%, 20.6%
NE: 3.3%, 0.2%
NW: 17.3%, 16.7%
POC: 2.6%, 1.2%
S: 13.8%, 5.6%
SW: 7.8%, 8.5%
Total: 100.0%, 100.0%

Several of these percentages (GL, NW, and SW) appear very comparable. However, the CE, NE, and S regions are greatly underestimated in the train books while the CW has over twice the representation that it should have.

Another way of comparing the tables is to allocate the 1,308 box/auto cars (2,243-935 UP cars) from the train books according to the ORER distribution and compare the actual and predicted numbers of cars:

Region: Actual; Predicted
CE: 145; 234
CW: 473; 209
GL: 269; 280
NE: 3; 43
NW: 218; 226
POC: 16; 33
S: 73; 180
SW: 111; 102
TOTAL: 1,308; 1,307 (Predicted does not sum to 1,308 because of rounding errors.)

The CW region contains the SP as well as the UP and we have had many discussions over many years on this list about the well-known over-representation of its cars in trains across Wyoming. 201 of the train book cars belong to the SP, whereas it "should" have only 44. This leaves an "excess" of 201-44=157 cars, which arguably should be removed from the 473 train book cars in the CW region. If they were removed, then the CW region would be left with 473-157=316 cars which is still half again as many as the 209 it should have.

So does this data support the G-N hypothesis? In my opinion it doesn't, but I stress that this is just an opinion and nothing more – others may decide it provides fine support, or the data have so many flaws it should be ignored. One flaw, a small sample size, will be ameliorated in the months and years to come: I still have over 120 train books to transcribe for the Laramie-Rawlins run in the late 1930s. With an average of 2,300 cars per book, 450 of them foreign box/autos, eventually the sample size should exceed 50,000 foreign box/autos.

I think I recall Tim Gilbert writing that the G-N hypothesis fits the data well during economic prosperity but does less well during recessions and depressions. The year 1938 was during the Great Depression, of course. The high proportion of home cars (41%) is another indication that companies may be keeping their cars close by. It might be that modelers wishing to have a realistic mix of cars on their trains should pick an era first – or perhaps even a specific year and season, and then check what was happening in the national economy at that time. The choice of the G-N vs. a regional (or any other) model for an accurate freight car composition may well depend on such ephemera.

Best wishes,
Larry Ostresh
Laramie, Wyoming