IT projects are legendary for high failure rates, cost overruns, and schedule overruns. Hence the motivations for questions like this.
Is there any actual empirical research that compares the failure rates of IT projects to those in other domains, in particular to engineering domains?
For example, has anybody compared cost and schedule overruns in IT projects to construction projects of comparable scope? There should be publicly available data here from government projects, but I'm not aware of anybody who's gone and done comparisons.