We used Map Calculator to perform the function "aGrid.Combine (aGridList)" to take data fields from each of the two grids being compared and place them together in a new grid. The function's format is [gridA].combine({[gridB], [gridC], ... [gridN]}). For example,
([G1976].combine({[G1980]})). The result was a file with four fields:
Value
Count
Gyyyy
Gyyy(y+4)
where Gyyyy was the PartyValue field from the grid (Republican=-1, Democrat=1).
The two PartyValue fields could have 9 possible value pairs, which represent several conditions (see Table 5 for the possible pairs).
Gyyyy | Gyyyy(+4) | Possible results |
1 | 1 | Democrat both elections |
1 | 0 | Democrat to Tie or No Data |
1 | -1 | Democrat to Republican |
0 | 1 | Tie or No Data to Democrat |
0 | 0 | Tie or No Data to Tie or No Data |
0 | -1 | Tie or No Data to Republican |
-1 | 1 | Republican to Democrat |
-1 | 0 | Republican to Tie or No Data |
-1 | -1 | Republican both elections |
switchCode | Gyyyy | Gyyy(y+4) | Label |
1 | 1 | 1 | Dem |
2 | 1 | -1 | Dem to Rep |
3 | -1 | -1 | Rep |
4 | -1 | 1 | Rep to Dem |
5 | One or two zeroes | No Data |
When all six grids were created and coded, Map Calculator was used to add together [( [1976_1980 . switchedParty] + [1980_1984 . switchedParty] + [1984_1988 . switchedParty] + [1988_1992 . switchedParty] + [1996_2000 . switchedParty]) + [1992_1996 . switchedParty]] the ŒswitchedParty' fields to yield a new grid showing the number of times each area of the grid changed party allegiance over the past 7 elections (see Figure 7).
The possible values ranged from 0 (voted for the same party in every election from 1976 to 2000) to 6 (voted for a different party in every election from 1976 to 2000).