For all the attention paid to the back-and-forth over the House and Senate versions of the economic stimulus bill, the version that REALLY mattered is the one that appeared last: the conference committee report. The early debate and early votes certainly contributed, since they helped set the parameters of what was possible for the legislation (i.e., what the critical players in either body would tolerate), but the votes to send the bill through each body were really prologue, since the conference committee would write the actual final draft that would eventually go to the president.
The lobby team at my last nonpofit (enviro) job drove home this observation repeatedly, putting the procedural victories and losses along the legislative way in perspective — sure, the Senate might strip a critical provision out or add in something nasty while voting on floor amendments, but the matter really wasn’t settled until conference. Conferees are of course still bound by the need to get the final bill passed in both bodies, but working behind the scenes they’re more able to make the swaps and changes to shape a bill (however ugly) that’ll get enough votes to survive.
Naturally, perhaps the critical choice in a contentious bill’s history is who goes on that conference committee, since selecting one member over another will ultimately favor one set of provisions or one philosophical approach over another. In the case of the stimulus bill, the process had a surprise ending: a version actually smaller than what passed either the House or the Senate. Whether it works or not? Future cloudy; ask again later.