There are utilities which are automatically cartels. Public transport is like a utility in that they are necessities which everyone uses and which the public as a consequence demands supervision over. The public will not permit private interests to, say, quadruple the train fare or the water bill because as a practical matter the entry cost of the industry is so high that competition is practically impossible. Can private investors lay new sewer lines ? Isn't the cost of massive infrastructure investment so high (the Brooklyn Bridge, subway systems, commuter rail, sewer systems, etc) that the taxpayer is the final guarantor ? So since the taxpayer was the underwriter, the taxpayer will be the determiner of what a "fair" profit is.
There's a myth that the top players in industries have a live and let live understanding, but it's a myth. Competition is fierce in almost every industry, if a business isn't expanding then it's probably shrinking, the business cycle is not condusive to a company sitting on a plateau, the only way for a business to expand is to take business away from somebody else.
Even in war there are hot sectors where combat is intense and vast quiet sectors of little or no activity. You can't spread out your armor and aircraft carriers everywhere. They are concentrated on key objectives. Competition for competition's sake is wasteful. Plateau ? Technology plateaus. When it does competition drops as now. The breakneck technological change of the 80's and 90's in personal computers saw feverish competition that did in Wang, Digital Equipment Corp, Ashton-Tate, Osborne, Timex, CP/M, Wordperfect, Dbase, Clipper, etc. Now it has stabilized because the marketplace has settled on a standard and X86 technology has plateaued. You see, the tendency is inevitably towards cartels because cartels enforce technological standards and the marketplace wants technological standards, be it in word processors, DVD recording formats, VCR tape formats, or operating systems. The marketplace wanted first IBM and then Microsoft to determine the technological standards. People want to invest in developing software that they know will still be supported 8 years from now. If I came up with a fantastic new X86 operating system the marketplace would not want it because they are not going to invest in something that major players will not support. Nobody wants orphan production software.
It is almost like the progression of history where states are created by warlords on horseback and consolidated by bureaucrats. Businesses are created in the "heroic age" of the industry by entrepeneurs on horseback and managed a generation later by bureaucrats who are risk minimizers.