Why are there system missing (SYSMIS ) values in the table below?

You should have written the following command sequence (or calling up the Transform > Compute > If dialog separately for each of the logical conditions) and then have produced the frequency table seen to right. :

COMPUTE PM=0.
IF (q11a=3 AND q11b=1) pm=1.
IF (q11a=1 AND q11b=3) pm=1.

IF (q11a=1 AND q11b=2) PM=2. 
IF (q11a=3 AND q11b=2) PM=2.
IF (q11a=1 AND q11b=4) PM=2. 
IF (q11a=3 AND q11b=4) PM=2.

IF (q11a=2 AND q11b=1) pm=3.
IF (q11a=4 AND q11b=1) pm=3.
IF (q11a=2 AND q11b=3) pm=3.
IF (q11a=4 AND q11b=3) pm=3.

IF (q11a=2 AND q11b=4) pm=4.
IF (q11a=4 AND q11b=2) pm=4.

VARIABLE LABEL PM "Postmaterialism Scale".
VALUE LABELS PM 
   1 "Materialist" 
   2 "Rather materialist"
   3 "Rather postmaterialist" 
   4 "Postmaterialist".

The command sequence needs to cover all possible combinations between all values of the two priority variables. When you had a look (if you didn't, do it now) at the frequency tables of each variable, you certainly noticed that in addition to the four items, both variables (q11a and q11b have a certain number of missing values coded "5" (DK=Don't know). Our sequence of commands does not take care of respondents who replied "Don't know" on one or both questions, therefore they have a value of SYSMIS.

There is a second possibility of getting a SYSMIS value. Remember that people were asked to select a first, then a second priority. In our transformation logic, we assumed that it was not possible to select, for instance "Maintain order" as first, and then again as the second priority, which makes no sense. But are we sure, that the interviewer let that pass and noted twice the same item? As a general rule, we should never be too optimistic in these matters. This was one of the reasons you were asked you to produce a crosstab between the two variables, where can check whether there are observations on the diagonal; there should be none when the row and column item are identical).

A third reason for getting SYSMIS value could be that you forgot or mistyped one of the conditions. For instance if your wrote IF (q11a=2 AND q11b=.4) pm=4. instead of IF (q11a=2 AND q11b=4) pm=4. the expression would never be true... and particular condition would not be covered by one of the conditions.

Technical note: Before processing the above command sequence starts, PM does not exist; when it appears on the first command line, SPSS creates it and sets all values of the newly created variable to SYSMIS, and then modifies it in sequence according to the conditions specified on each IF command. At the end, all observations that have not been modified explicitly keep their initial SYSMIS value.

If you have more SYSMIS values than the number shown in the table, you must have made an error when specifying the condition. Try to find the (logical) error, but also check the Viewer) to see whether there has been a syntax error with the commands..