"The World's First Railway System: Enterprise Competition and Regulation on the Railway Network in Victorian Britain" by Mark Casson, OUP, 2009 looks at how efficient the investment in railways in the the 19th century were. There are lessons in this for Australia's planning for the National Broadband Network.
The conclusion drawn by Casson is that there was duplication of railway infrastructure and some inefficient placement of capacity. He concludes that this was due to a failure of the political process, with MPs not having the courage to make decisions for the good of the national as a whole, unable to choose between competing local interests and so making suboptimal decisions. Casson argues that private enterprise with some government planning could have resulted in a more efficient railway system and that this was done in India. This work provides some insights for Australia, with Casson also commenting that Australia failed to learn from the UK's problems and introduced three different railway gauges, a lack of standardisation which is a problem 100 years later.
The National Broadband Network is similar to the model Casson suggests, with central government planning and private investment. I suggest that the Internet protocols are the equivalent of the standard gauge which was missing from Australian railway planning. With it we will be able to have a meshing of multiple private and public networks, seamlessly carrying data round Australia and around the world. Without it we will have a series of little branch lines, with packets of data having to be loaded and unloaded between different data standards, just as goods still have to be transferred between different gauge railway lines in Australia today.