In situations where density is unknown, the CIS allows the user to enter shipment class, which is then converted into a density. That density is the guideline used by the National Motor Freight Classification (NMFC) committee when assigning the class to a specific commodity. But how accurate is the use of such a default? The NMFTA does not release actual measured density information on commodities, meaning that little data is available to determine how well freight of various classes actually conforms to the guideline. The one widely-used density study that is available to the industry is the Density Study, published by Southern Motor Carrier Rate Conference (SMCRC), which is an accumulation of actual measurements for commodities done by that Bureau. The publication, which can be purchased directly from the SMCRC, groups commodities by NMFC item number. It contains over 12,000 actual shipment observations.
Recognizing that the study is not a random sample of all freight, we still thought it would be interesting to see how the SMCRC commodity measurements compared to the NMFC guideline. Thus, the TCG staff looked up and recorded the classes for all of the commodities in the study. The commodities were then sorted and averaged by class. Having no data by which to weigh the observations, two different averages were developed: One average assumes all observations had the same weight and the other assumes all observations had the same cube. Neither, of course, is the true average of the freight, but, again, no other information was available.
The results are shown in the table below:
| Class | Low Density | High Density | Avg. Density - Weight Basis | Avg. Density - Cube Basis | NMFC Guidline |
| 50 | 6.1 | 244.6 | 59.7 | 30.8 | 50.0 |
| 55 | 3.1 | 117.8 | 33.5 | 22.7 | 35.0 |
| 60 | 5.4 | 247.7 | 38.5 | 24.1 | 30.0 |
| 65 | 0.2 | 210.9 | 37.5 | 8.2 | 22.5 |
| 70 | 1.4 | 219.7 | 25.3 | 14.0 | 15.0 |
| 77.5 | 3.5 | 155.8 | 26.3 | 13.0 | 13.5 |
| 85 | 0.5 | 101.2 | 17.0 | 9.4 | 12.0 |
| 92.5 | 3.7 | 190.6 | 14.3 | 8.9 | 10.5 |
| 100 | 1.6 | 79.3 | 13.2 | 8.2 | 9.0 |
| 125 | 1.2 | 75.0 | 7.7 | 5.5 | 7.0 |
| 150 | 0.1 | 79.0 | 6.6 | 3.5 | 6.0 |
| 175 | 1.2 | 14.8 | 4.5 | 3.3 | 5.0 |
| 200 | 0.2 | 109.8 | 6.0 | 3.2 | 4.0 |
| 250 | 1.4 | 52.8 | 5.5 | 3.3 | 3.0 |
| 300 | 0.4 | 32.6 | 3.9 | 2.0 | 2.0 |
| 400 | 0.3 | 9.4 | 2.7 | 1.3 | 1.0 |
| 500 | 0.9 | 0.9 | 0.9 | 0.9 | 0.8 |
When a study is used for a purpose other than that for which it was designed, as is the case here, the results should always be viewed with caution. But the data, as compiled here, does show that, as we have stressed in the past, the use of class to determine density may not be bad on the average, but it should be a last resort, when no other data is available. The large swings between highest and lowest density for each group is particularly noteworthy: Always be aware of it when using class to obtain density for a particular customer's freight.