In the Israeli general election of June 1992, non-Jewish voters compri
sed 12.3 percent of the electorate. Theoretically, this would be suffi
cient to elect 15 non-Jewish members to the Knesset if there was a uni
ted effort to elect only Arab representatives. However, three Arab-sup
ported parties took only 4.88 percent of the valid vote and won five (
4.17 percent) of the 120 seats in the Knesset. This understatement of
potential electoral strength is due almost entirely to low voter turno
ut, unfocused voting patterns among non-Jewish voters who, in addition
to voting for the three predominantly non-Jewish parties, also cast t
heir votes for the full range of Jewish (Zionist) parties, and the ina
bility of the Arab parties to agree on distribution of their surplus v
otes among themselves. It would not be improper to say that the distri
bution of the Arab votes in Israel defies statistical explanation. Sta
tistical analysis of the voting shows that it is extremely difficult t
o explain or predict the electoral behaviour of the Arab population in
Israel, as neither geographical nor socio-economic variables reveal a
ny clear pattern. This study indicates the need for detailed investiga
tions relating directly to local and neighbourhood effects in the Arab
vote in Israel.