The goal was to make it harder for such frauds to take place--but in practice, as with most such laws, small companies were generally harder hit by the regulatory complexity and costs than large companies. While it did not by itself cause the screeching decline in startups after 2000 (it was not passed until 2002), it was certainly a contributing factor to the relatively small number of such business attempts.
There was a big decline in startups anyway because so many of the dot-com companies were built around tulip bulb crowd madness and other signs that too much money was chasing too few legitimate ideas. My favorite example of this was an email I received from a South of Market Street (in San Francisco) dot-com that was trying to recruit me with such valuable benefits as a "clothing-optional workspace."
Still, Sarbanes Oxley is a contributing factor. This article from the December 2, 2011 Idaho Statesman mentions that one of Idaho's U.S. Senators is a sponsor of a bill to provide a temporary easing of these rules for small companies:
The legislation would "temporarily [scale] back certain regulatory requirements imposed by the Sarbanes Oxley Act and the Dodd Frank Act," Crapo said.
He added that "the average cost for a company to go public is $2.5 million, and the annual cost to stay public is $1.5 million."Believe me, I would love to see startups happening again on a big scale. Who knows? Perhaps it could revive the economy, and create a few software engineering jobs for people with wrinkles.