The description of INCDISAB says that it includes income from “worker’s compensation”, but the description of INCWKCOM doesn’t specify whether it is different or whether it is included in INCDISAB. The questionnaire text doesn’t appear to be available.
Update: After comparing the actual values of INCDISAB and INCWKCOM in the ASEC from 2000 to 2014 (the sample I’m interested in), it does not appear that INCWKCOM is included in INCDISAB. If this were true, then INCDISAB >= INCWKCOM should hold for all observations in the appropriate universe. However, a significant number of observations do not meet this assertion - there are many observations where INCWKCOM > 0 and INCDISAB = 0. This is inconsistent with the reply of the Staff member above and, I think, warrants investigation.
You are right, this is an inconsistency. The documentation provided by the Census Bureau on the FTP site is a bit ambiguous on this issue. However, I am able to confirm that INCWKCOM and INCDISAB are, in fact, mutually exclusive. Therefore, INCWKCOM is not included in INCDISAB. The easiest way to verify this is by looking at the Comparability Tab for INCTOT. For years 1988 and onward both INCWKCOM and INCISAB are among the list of variables that sum together to produce INCTOT. Therefore, they must be mutually exclusive.
It seems the reason for this is the wording of the questions for INCWKCOM and INCDISAB, respectively. INCWKCOM asks about income as a result of “job related injury or illness” and INCDISAB asks about income as a result of “disability or health problems”. Although the source, worker’s compensation, is the same, the reason is different.