I do not know, I do not have ESP powers. There a few possibilities: - control of oil reserves (especially that "peaking out" looms on the horizon); - profitable contracts; - throwing a monkey wrench into the EU (US business starts fearing EU, as argued in the last issue of The Nation); - tightening control of Arab states; - aiding Sharon; - re-vamping the military; - military Keynesianism; - undercutting the "internationalist" factions in the US foreign policy and bolstering unilateralists.
These are just a few that come from the top og my head, and which can be appealing to very different constituencies. Of course it is an empirical question whether these constituencies actually supported the war and if so, for what reasons, and what was the impact of that support. It is another empirical question what other, less obvious, reasons might have come into play and who espoused them.
The point I am trying to make is that reasons and rationales for any policy are two very different things. Sometimes they coincide, sometimes they do not. What is more, important decisions at the state level are almost never made unilaterally - they require some sort of consensus among power brokers, and that consensus is usually given for different reasons and with different conditionalities attached to them.
This may not be as elegant and appealing as mono-causal theories though - and also less open to speculations and more dependent on factual knowledge.
Wojtek