The news on privately maintained (and possibly built) roads is not really news. The "innovative approaches" look very much like the old approach of PFI. Private capital provides the investment (so no need for public debt), in return for a guaranteed profit, while the government underwrites the risk. It's also worth noting that much the same navvies will be out in their high-vis jackets doing their Tarmac magic. The Highways Agency routinely subcontracts to private companies.
David Cameron asked a bizarre question as part of his announcement: "Why is it that other infrastructure – for example water – is funded by private-sector capital through privately owned, independently regulated utilities, but roads in Britain call on the public finances for funding?" The answer is: because the water companies get to send each household a bill, generating revenue which covers the cost of capital. The water analogy only makes sense if the government anticipate tolls as the rule, rather than the exception, which I doubt they do.
Some of the early debate has focused on whether private business can deliver more effectively and efficiently than public provision (the opportunity argument), while there have also been concerns about the risks of a privatised monopoly (the risk argument). What has occasioned much less debate is whether roads are a sensible place for anyone to put their capital.
We know from the history of railways in this country that a fully freed market creates over-capacity. This ultimately led to the famous Beeching Axe of the early 1960s, the rail network having been nationalised in 1948 due to it inefficiency and under-investment. The earliest line closures had occurred only a little after the railway boom in the middle of the 19th century. They gathered pace in the 1920s and 30s as goods transport started to move to the roads with the coming of the lorry, eroding the railways' profit. It was this shift that convinced many of the need for an integrated transport strategy, which was another driver for nationalisation.
Infrastructure is deemed to be a money-spinner because there is so much of it, and everyone needs it. I remember the over-pricing of Cisco (data network manufacturers) in the late 1990s was largely based on the meme that they were the modern equivalent of the firms that sold picks and shovels to the California gold-rush miners. Their shares reached just shy of $80 in early 2000 and are at $20 today. This over-valuation was not a failure to judge the merits of Cisco, as they remain a major player in an expanding sector, but an over-estimation of the ROI from infrastructure as a class, amplified by the prevailing dot com madness.
The problem with infrastructure projects is that they are capital-intensive. This means that constant capital (plant, machinery, materials etc) makes up a much larger share of the total cost, relative to variable capital (labour), than the norm. Constant capital components produce little profit themselves. This is because adding a fat mark-up makes the work uncompetitive relative to other bidders who add a smaller mark-up. Assuming you have a free market, the margin on constant capital will be minimised. You can gain an edge by having better quality constant capital, e.g. more efficient plant, but road builders acquire their plant from third parties - they don't invent and develop it themselves. The same applies to railway companies, water companies, cable companies (anything to do with Richard Branson, in fact).
To add value, you need labour to transform the constant capital (e.g. a more talented worker, putting the same components together, can produce a better quality product, which justifies a higher price because it is more valuable to the consumer). Larger profits are produced by projects that entail significant variable capital (i.e. labour) and thus have the potential for extracting greater surplus value (i.e. the difference between the value added by the labour and the cost paid for that labour).
This is true not just in terms of building new stuff, but can also be seen in respect of projects aimed at improving existing processes. Genuine productivity gains, e.g. doubling the output with the same input, are rare, if only because the necessary circumstance, where an increased supply to the market corresponds with a natural increase in demand, thus maintaining prices, is unusual. The same output with fewer inputs is much more common, as this doesn't depend on a volatile market nor does it disrupt a stable market. However, this improvement tends to peter out as the competition take similar steps and then underbid to gain market share. This drives down the profit margin across the sector to what it was before.
The most profitable process improvement projects are those that focus on quality/innovation. These leave the inputs and outputs unchanged but allow for an increase in product price (because of unique features), or the maintenance of a price premium in a commoditised market where unit prices are otherwise falling. As the cost of constant and variable capital does not change, an increase in price effectively increases the surplus value of the labour.
Apple is not the richest corporation in the world because it has invested large amounts of capital. It's recent decision to free up some of its massive cash pile as a dividend indicates it has not been working its available capital all that efficiently (the cash pile is roughly 20% of its market capitalisation). Much of its manufacturing R&D was outsourced after the mid-90s, after its soup-to-nuts style of operation in the 80s proved increasingly unprofitable, with the result that its corporate R&D (mainly architecture and OS) is now only a little more than 2% of revenue. Keeping design in-house does not require large constant capital.
Apple's real dynamic as a profit-generator is the combination of premium pricing and outsourced labour. By offshoring the bulk of its manufacturing to China, the profit on labour can be increased. Expensive goods with small capital costs can only generate large profits if labour does not soak up the majority of the non-capital price. As a former manager at their outsourced supplier Foxconn has said: "Apple never cared about anything other than increasing product quality and decreasing production cost".
The problem for investors is that the conditions that allow for Apple's profitability cannot be replicated in the case of UK roads, for the simple reason that we cannot physically offshore their building and maintenance. While it is feasible to imagine the wages of local road workers being squeezed further, there is no possibility of pushing them down far enough to generate very large profits. So why would anyone want to invest in our roads? The most credible answer is that the dearth of investment opportunities (of which Apple's cash pile is a symptom) means that even projects with a low yield are attractive if the alternative is cash sitting in the bank earning negligible interest.
The problem for us is that if road projects cannot deliver a significant profit margin, they will be more vulnerable to adverse shocks, such as a rise in the price of oil or other materials. This leaves the private supplier with two alternatives: either reduce costs to maintain the profit margin, or renegotiate the contract with government upwards. The first of these can have catastrophic consequences, as the sorry litany of rail disasters in the late 90s and early 00s shows. As a result, the second is the more likely.