One of the things that even fairly small government sorts think is a reasonable government job is building roads, railways and other similar infrastructure. Yes I know the real minarchist libertarian sorts will point to the 19th/early 20th century and note that fairly large infrastructure projects, such almost all the major railways in the world, were built with private funds but a smart slightly larger government person will point out that in most (all?) cases the builders had a lot of help from government in terms of acts of parliament or similar that permitted compulsory purchasing of the land and things like that.

Anyway, ignoring the ideological purists, suffice it to say that there is general agreement that since “government is just the word for things we do together” and building roads etc. is a particularly cooperative enterprise, governments are, these days, expected to take the lead on the building process. Die hard socialists and Californians (but I repeat myself) seem to think that governments should actually directly employ the thousands of workers, buy the construction equipment and supplies, and do the whole thing in house. Not many other people do these days because it is pretty blindingly obvious that unless the government is planning on sequentially building stuff across the entire land under its control, this is a vast misuse of resources. Thus almost all government infrastructure projects involve the government contracting out most or all of the actual construction to construction companies who do that for a living – particularly large bits of brand new infrastructure (smaller projects: extensions, widenings etc, may or may not be contracted out). [Note that there’s nothing wrong with the government directly employing maintenance people and having serious repair equipment, just as long as there’s enough infrastructure that needs maintenance and repair that the workforce can be employed continuously, but the skills required to repair and maintain are not necessarily the same as the ones required to build new stuff.]

This is a good thing as long as the contracts for the construction companies are written well and the companies that win the contracts are well run and have a clue. Sadly we see that these days this is not always the case.

Firstly, over in the UK, we’ve seen the bankruptcy of the major government contractor – Carillion. Part of the problem here is that although it started as a construction company, Carillion became a conglomerate formed by the merger/take over of a number of companies. Some were construction companies but others did all sorts of other things and about the only thing they had in common was that their customers were government at one level or another. In other words, Carillion’s expertise morphed from actual construction to extracting money from government (and in the process becoming a dedicated implementer of “diversity” and other politically correct buzzword bingo). The reason it died is that it turned out to be rather poor at pricing its bids correctly and had to resort to various forms of financial engineering (a.k.a. fraud) in order to stay alive. Eventually all the financial engineering was unable to keep the house of cards stable and it collapsed. Now of course all its competitors are busily licking their lips and planning to bid for the contracts the lost to Carillion at far higher rates and the subcontractors of Carillion – who are almost certainly not going to get paid for the work they have already done  – are going to get screwed

Secondly, in California, we see the amazing ever increasing cost of building the bullet train to nowhere:

Only two years ago, the California rail authority unveiled an ambitious plan to begin operating a segment of bullet train service between San Jose and the Central Valley by 2025. It would take nearly every penny in its checkbook, but the rail authority assured the public it would work.

But that plan has been crushed by the acknowledgment Tuesday that the cost of building just 119 miles of rail between the farm towns of Madera and Wasco has soared from about $6 billion to $10.6 billion, siphoning off money that the authority had planned to allocate to the ultimate goal of connecting Los Angeles and San Francisco.

First you have to wonder why it originally was assumed to cost $6 billion and why the projected cost rose about 75% to $10.6B. The sums are outrageous: 119 miles at a cost of $10.6 billion works out at about $90 million per mile (at $6B it was a mere $50m/mile). That’s not a bad cost for mountainous or urban construction, in fact comparing it to, say, the NY Metro expansions at over $2billion per mile it looks like a bargain, but the 119 miles are pretty much flat farmland. You can’t blame land cost for this; since this is not metro New York (or even the SF Bay area) land prices are probably well under $10,000/acre for vast majority of the route – I’d guess a lot is under $1000/acre to be honest. If we assume a 100 yard wide strip (which works out at about 36 acres/mile)  and a generous total average land cost of $1,000,000 per mile that works out at over $27,000 per acre cost and that notional $1m is barely more than 1% of the $90m projected cost per mile of the undertaking.

At $90 million per mile in terrain that has little or not tricky topography or geology you have to wonder what exactly the money would be spent on. Even if the whole this is elevated, that seems excessive in terms of material costs – $10 million can buy a lot of concrete and rebar. Moreover you have to wonder just how expensive the tricky bits (over the mountains into San Jose/Gilroy and eventually down south into LA) would be. Plus of course the cost of actually valuable SF Bay area land….

In other words it looks like the money will be spent on something other than concrete and the workers/equipment to place it. Oddly enough one of the contracts went to a joint bid that included a company (Tutor Perini) with a major shareholder who turns out to be the husband of California Senator Diane Feinstein, and that joint bid has already seen a healthy 18% increase in its payments. I haven’t checked where the rest of the money is going but it looks like Tutor Perini – unlike Carillion – are successfully managing to shake the government down for more.

Neither way – the UK way where the contractors get carried away with diversity, over extend and fail or the California way where they successfully milk the tax payers for more – is good for the efficient provision of infrastructure. And we can be fairly sure that the politicians, bureaucrats and CEOs in question are all going to skate away from the messes they make leaving future tax payers on the hook to clean them up.

I’m not sure what the fix is but I’m sure something needs to be done to stop this sort of waste. In the California case I think the answer is to simply make it harder for governments to fund this kind of idiocy themselves by insisting that such projects be financed entirely by private money. If they can’t fund it but, as in the 19th century, are only able to pass eminent domain rules to support it then that might help. Solving the UK one is harder, but on the other hand it may be self-correcting because now politicians and bureaucrats know to not trust large conglomerates, even if said conglomerates are run by their friends and conform to their preferred virtue signaling