Private infrastructure finance and secular stagnation

by John Q on November 19, 2015

For most of my academic career, I’ve been working on (more precisely, trying to demolish) the idea of private investment in public infrastructure, exemplified by the Private Finance Initiative in the UK and the Public Private Partnerships program in Australia. Here’s my first published article on the subject, from 1996. I conclude that

The current enthusiasm for private infrastructure, like the enthusiasm for public ownership which it replaced, has been based more on ideological beliefs in the virtues of one sector and the vices of the other than on any systematic economic analysis …Analysis of the relative performance of the private and public sector in different phases of infrastructure provision suggests that, in most cases, the private sector will be most efficient in the construction phase but the public sector will be best equipped to handle the risks associated with ownership.

Twenty years later, this analysis seems finally to have been validated. The UK Auditor-General recently reported that

Analysis of the 2012-13 Whole of Government Accounts (WGA) implies that the effective interest rate of all private finance deals (7%–8%) is double that of all government borrowing (3%–4%)

As a result of the excess costs, and some spectacular failures, bipartisan enthusiasm for the PFI has finally turned to disillusionment. Here’s the Telegraph, correctly putting much of the blame on New Labour. And, for balance, here’s the Guardian. There hasn’t been a similar admission of failure in Australia, but the flow of PPP projects has greatly diminished, and most new ones rely on a substantial component of public capital.

Unfortunately, the failure of private finance hasn’t led governments to resume the high levels of public investment that prevailed in the Keynesian era of the 1950s and 1960s. So, even with central bank lending rates at zero, there has been no real recovery in infrastructure investment. Apart from the direct effect of lower investment, there’s a strong case that infrastructure investment increases the returns from private investment in general and therefore stimulates growth.

I think this analysis applies more broadly. The financialization of the global economy has produced a hugely costly financial sector, extracting returns that must, in the end, be taken out of the returns to investment of all kinds. The costs were hidden during the pre-crisis bubble era, but are now evident to everyone, including potential investors. So, even massively expansionary monetary policy doesn’t produce much in the way of new private investment.

This raises the question of how the hypertrophied financial sector we observe has succeeded in displacing less costly alternatives. I plan another post on this but my short answer rests on regulatory and tax arbitrage. These, and not the efficient allocation of capital, are the key functions of Wall Street, the City of London and their various counterparts around the world.



Marshall 11.19.15 at 4:43 pm

You forgot sucking money out of companies and into the hands of shareholders and executives.


Metatone 11.19.15 at 5:44 pm

Part of the problem is that one reason for the popularity of PFI (far from the only one, but an important one) is that is allows politicians to keep some things “off the books.”

Hence, the reason there hasn’t been an upswing in public investment has to be traced back to the “on the books” phobia – well illustrated by the current UK proponent, George Osborne.

(He’s so keen to keep things off the books, he did a frankly bizarre deal with China over a nuclear plant…)


cassander 11.19.15 at 6:04 pm

It’s definitely true that public/private partnerships tend to combine the worst aspects of both. But the decline in infrastructure spending has far more plausible causes than the rise of such partnerships, the rise of finance, or some decline of Keynesianism.

First, you have simple limits on what is useful. If spend 20 years building dams on any river than looks at you funny, you can’t spend the next 20 going around damming them again. Second, you have the rise of environmentalism, and the conversion of the left from pro-infrastructure to anti-infrastructure. The US has seen a similar decline in such spending to the UK, and the fight over the keystone pipeline is a perfect demonstration of at least part of the reason why, the left went from wanting to build dams to wanting to tear them down. Last, you have the massive rise in entitlement spending crowding out everything else in government budgets.


SamChevre 11.19.15 at 6:20 pm

I’m curious about other countries, but in the US infrastructure has gotten VASTLY more expensive to build since the 1950s; at some point, it costs too much to be worthwhile.

For example, the Interstate Highway system was estimated in 1956 to cost about $1 million per mile; it came in fairly close. That would be $8.74 million a mile today–round up and say $10 million a mile.

A county where I have friends (Chesterfield County VA) recently adopted a plan to build bike paths; the bike paths are estimated to cost $1 million a mile.

If you used to get a divided four-lane interstate for your money, and now for the same money you can get ten bike paths–it’s easy to see why this doesn’t look like such a good deal.

Another factor is the huge opposition on the left to building any new infrastructure–name a pipeline, port, highway, or bridge project that hasn’t had huge opposition from the left, because I can’t.


MPAVictoria 11.19.15 at 6:28 pm

“Another factor is the huge opposition on the left to building any new infrastructure”

The people on the left are the only ones who support infrastructure spending.


SamChevre 11.19.15 at 6:39 pm

MPAVictoria @ 5

The people on the left are the only ones who support infrastructure spending.

The left is not monolithic–but I’m curious; what highway, port, pipeline, freight rail, or bridge project had substantial support from the left?


Will G-R 11.19.15 at 7:00 pm

I’m tempted to interpret public-sector privatization by way of analogy with the appropriation of Church lands during the Protestant Reformation. Obviously the Catholic Church was an imposing force and kings or princes often fought it at their peril, but from a certain perspective the achievement of proximate ideological goals (whether clergy could sell indulgences, whether Henry VIII could divorce, etc.) is beside the point: the key effect of Reformation was simply to obliterate the largest feudal lord and redistribute its holdings to capitalist landlords, giving the economic transition from feudalism to capitalism a shot in the arm.

In a similar sense, even if any given privatization project isn’t cost-effective for capitalists or their politicians and leads to red ink or lost elections, the point is to herald and accelerate a far more important long-term economic trend: the neoliberal erosion of secure, unionized, salaried employees in favor of an insecure, fragmented, outsourced “precariat”. Public-sector workers are the single largest constituency providing ideological inertia for the old Keynesian/Fordist paradigm, and thus must be combated through outsourcing and privatization if that paradigm is to be thoroughly eradicated. From a certain perspective, any immediate-term costs incurred in this struggle are just a bit of necessary overhead.


John Quiggin 11.19.15 at 7:02 pm

@4 That’s interesting, but I’m pretty sure your 1956 example is unadjusted for inflation, which makes it pretty much useless. It got me interested enough to Google and I found a Politifact Check on a claim by the Mayor of Portland that the city’s entire cycle network (319 miles according to this source) cost less than a mile of freeway

Politifact ranks it Mostly True.

And of course, this is an example of infrastructure spending that has plenty of support from the left. Similarly renewable energy. So, the response to your question is that the left is more supportive of aggregate infrastructure spending, but has different specific preferences. To put it simply, the left wants modern infrastructure, rather than being sentimentally attached to 20th century technology, like the Republican Party.


SamChevre 11.19.15 at 7:12 pm

John Quiggin @ 9

The change from “a million a mile” to “8.74 million a mile” is the CPI adjustment for inflation, from the BLS, here.

That’s why I said 10 miles of bike lane, not one mile.

I believe the mayor’s claim (freeway costs 300 times what bike lane does per mile) and mine (cost of any given investment has gone up hugely more than inflation since the heyday of infrastructure building, to the point that a bike lane today costs as much as a two-lane road did then) are entirely compatible.


Bruce Wilder 11.19.15 at 7:40 pm

The Right in the U.S. seems to be attached to the American Dream, 1950s auto edition, and gets hostile whenever any other vision is proposed.

In the annals of partisan disputes over infrastructure spending, the 2010 election in the U.S. resulted in the shelving or revision of several rail projects that had broad (not particularly left) support. Chris Christie — since famed for his administration’s brilliant investments in empty casinos and pointless mega-malls as well as closing lanes of the George Washington Bridge — cancelled a plan for new cross-Hudson rail tunnels, an ill-considered decision that may have profound effects on the region, as existing tunnels will have to close for repair and refurbishment. Wisconsin cancelled a plan involving rail connecting Madison, Milwaukee and Chicago, at a considerable cancellation cost as well as damage to the viability of a Milwaukee railcar manufacturer. Florida cancelled its plan for a Tampa-Orlando high-speed line; a semi-private plan to bring back the Florida East Coast railway has since gone forward, but with much more modest goals as far as corridor speeds are concerned. Ohio lost Federal funds to support that state’s efforts to build a rail network connecting its major urban centers.


A H 11.19.15 at 7:52 pm

“If only it could be all ordered from China, shipped in containers and sold at 500% markup… Or is there some other business model?”

See the Bay Bridge


John Quiggin 11.19.15 at 7:58 pm

@10 Actually, Politifact gives the cost of a *four-lane* rural freeway today as $12.4 million a mile, so it looks as if costs might have fallen (I don’t what the cost ratio is between four lanes and two). Certainly, relative to mean income, and implied value of time, construction cost of freeways looks to have fallen.

The Mostly True ranking for $60 million is based on the assumption that the comparator is urban freeway which is hugely more expensive, (I suspect in part because the costs of demolishing people’s homes can’t be externalised as easily as it was in the 1950s)

The Portland numbers suggest you can get 5 miles of average bikeway for $1 million, so the ratio is 60:1 which looks pretty good.


John Quiggin 11.19.15 at 8:09 pm

@11 If they dislike rail, they absolutely hate wind and solar PV. In Australia, most of the right has signed on to an anti-vax style theory about health effects generated by the almost inaudible sound generated by wind turbines, exactly the kind of thing they once mocked in the loopier elements of the left.

Of course, climate science denial paved the way for thi. They’ve got used to thinking that everything is a matter of opinion, and that tribal identity is the best basis for forming an opinion.


Bruce Wilder 11.19.15 at 8:14 pm

Unfortunately, construction costs are not something that just comes off the menu. On Big Projects, there’s a tendency for the political process to produce not the Portland outcome, but the Seattle Alaska Way viaduct replacement answer or the Tappan Zee Bridge replacement outcome. Or as A H @ 12 points out, the San Francisco Bay Bridge.

In Seattle, they cast around for years for a plan, and somehow they ended up with the most fantastically expensive option, which, of course, is also not technically feasible, but is now consuming some unbelievable fraction of the whole state’s transportation spending budget. All for a giant digging machine that’s stuck in the mud.

All of these projects also have in common that they attempt to reproduce an infrastructure architecture for petroleum-fueled automobiles and trucks, which, in all probability, will be obsolete within a generation.


mjfgates 11.19.15 at 8:35 pm

The Alaskan Way viaduct replacement has some genuinely hard problems; build a major thoroughfare, in a space that Does Not Have Room, that has to be earthquake-safe, on mud. Then you add in the issues that crop up on every big public project. It was guaranteed to be a mess from day one. There never was any argument about whether we were going to do it, though.

In fact, the Democrat-controlled governments in Washington State have been quietly funding and building new infrastructure for decades. Replacements for both floating bridges across Lake Washington, a new span over the Tacoma Narrows, the bus tunnel under downtown Seattle, the light rail line that is slowly growing to connect everything between Tacoma and Everett. There’s been no particular opposition to any of it.

The one project here that did get cancelled was the coal-export terminal in Bellingham. Apparently leftists are okay with things that actually help us, but not with enriching coal-company executives at the expense of fucking over the world. Who knew?


PlutoniumKun 11.19.15 at 10:24 pm

I’m very glad for all this work, even though the notion of public private partnerships being more efficient and cheaper than the public sector alone is a zombie that refuses to die – I was reading some months ago about a major conference for development donors specifically aimed at encouraging PPP as a development model for developing countries.

For a long time I’ve puzzled over the popularity of PPP as its always been very easy to demolish any argument that they are cheaper – all you need to do is point to the far cheaper financing costs for the public sector, combined with the lack of any evidence that the public sector was any less efficient at running contracts than the public sector made it obvious. I think one of the more insidious ways they have been made popular in the Eurozone in particular is that borrowing rules imposed on governments have made ‘off book’ borrowings more popular. Certainly here in Ireland, the government is very keen on PPP for precisely this reason.

On the subject of the cost of highways and cycleways – there is a huge variation in costs, but I’d offer one reason in particular why highways and other infrastructure costs have risen so much – and thats related to rules related to public sector tendering. Back in the 1950’s and 1960’s when public bodies had far larger budgets (proportionally) and had more discretion on spending, they were able to build infrastructure using long term rolling contracts. In comparison, its now common for each individual item to have to go to tender separately. There are enormous costs involved in this – every major project requires a long period of gearing up, and winding down – its impossible for contractors to invest long term in either tooling, or in staff knowhow. To give one example, the French have been able to build high speed railways far cheaper than elsewhere, simply because they have permanent staff in place in the national railways designing and managing contracts, and they use small numbers of experienced contractors on long term rolling contracts (a comparison cost between the Channel Tunnel Rail Link and almost any equivalent stretch of French TGV will show the difference). The use of long term rolling contracts invariably proves vastly cheaper than the alternative. The Madrid underground system is a another well known model for how to do it right. But this type of long term commitment is becoming increasingly difficult under procurement rules.

Another factor – and this probably applies as much to bike lanes as it does to highways – is that modern construction standards are actually a lot higher – there does seem to be a stochastic principle involved with engineering standards – as errors are corrected, the quality and complexity goes up. Just take a look at any 1950’s motorway and look at the modern equivalent – there is very little comparison.


Bruce Wilder 11.19.15 at 11:21 pm

mjfgates: There never was any argument about whether we were going to do it, though.

That’s misleading, since there were a number of alternative plans developed, before someone pulled a deep-bore tunnel (of dubious feasibility) out of a hat. (Did I mention that construction is in the hands of an ad hoc private consortium practically designed to take the money and run?)

The political problem, of course, is getting a process with a lot of stakeholders with hold-up power, to “yes”, (and particularly where the political support of private business corporations in construction business is needed). I’m just saying that solving that problem in the U.S. today often involves increasing the scale of the project and impairing the rationality of project design in ways that inflate costs relative to function.


Watson Ladd 11.20.15 at 1:29 am

Any analysis of public investment in infrastructure needs to address diminishing returns and the invisibility of need. Before you have a major bridge, the need for one is obvious, and the return immediate. After you have one, needing to set aside money for a replacement or repairs is a much less visible expense, and the return not as directly felt. So looking entirely at expenditures doesn’t really make sense: you would have to ask whether the need for increased investment is still there.

I do have to push back against the idea that the finance sector raises the cost of capital. Sure, every banker would love to be a necessary intermediary, but there are plenty of investment banks for would-be issuers to pick from, and for governments in particular, lots of ways to bypass them. I know this was an offhand remark at the end of a post, but there needs to be a mechanism for would-be intermediaries to intermediate and extract more value then one would think they could.

As far as reduced rates, they have contributed to a flood of money into venture capital with deals structured more like bonds then equity. There are also articles about increased investment in emerging economies, but the returns haven’t been as great as expected. So it seems that the usual mechanisms are working, at least anecdotally. I eagerly await being proven wrong.


david 11.20.15 at 2:11 am

@6 seizes on the historical insight – before PFI there was the enormously successful Homes before Roads, a movement ensuring that there would be neither roads nor homes.

A doubled interest rate sounds like a lot, but the (correct) New Labour insight is that public provision implies a level of public consultation and multiplicity of veto points that then ensure that the project is never carried out at all.


BBA 11.20.15 at 2:46 am

This kind of privatization is relatively rare in the US. I think it comes down to the US never nationalized much compared to other countries – freight rail, electricity, telephones were always private – so there hasn’t been a push to privatize what we see to be governmental activities – roads, airports, waterworks – or to make new construction private.

The aforementioned barriers to public construction are, if anything, higher in the US than in Britain or Australia, but since private construction is a nonentity, the result is that nothing gets built at all, which is fine by a large chunk of the electorate. Taxation is theft and a penny spent is a penny wasted, after all.


Peter T 11.20.15 at 8:33 am

6 and 20

Yet governments (at least here in Australia) seems to have no difficulty in over-riding local opposition or “streamlining” approval processes when a developer wants to build an office tower or replace public housing with an up-market apartment block. Wonder how that is?


Colin 11.20.15 at 8:46 am

Attributing PFI schemes and similar apparent white elephants to ‘ideological beliefs’ may be overly charitable to the moral character of politicians, when they have a rational interest (as private individuals with business contacts, and also as recipients of donations to their political party) in obfuscating the business of public expenditure and bypassing the usual rules for awarding government contracts. 4% per annum extra interest is a small price to pay for keeping voters in the dark. From the cynical politicians’ perspective, the ‘failure’ isn’t in the economic inefficiency, but that voters have started to see through the charade.


reason 11.20.15 at 9:14 am

My view is that infrastructure investment is needed to combat centralising forces and high land prices, which are a major contributor to inequality. People from Australia really to look around say the Randstad in Holland or the Rhein-Main-Gebiet in Germany to get an idea of what multi-city development looks like. You need more good well serviced places to live, not sprawl.


Tim Worstall 11.20.15 at 10:51 am

Not entirely sure that your two parts to the argument fit together.

“Analysis of the relative performance of the private and public sector in different phases of infrastructure provision suggests that, in most cases, the private sector will be most efficient in the construction phase but the public sector will be best equipped to handle the risks associated with ownership.”

Part of the PFI argument is that the private sector should do that building, the bit they’re most efficient at.


Ginger Yellow 11.20.15 at 1:04 pm

Part of the PFI argument is that the private sector should do that building, the bit they’re most efficient at.

The private sector has always done the building. PFI’s innovation was that it let the private sector do the financing* and the operation. At least until government accounting was cleaned up, allowed investment to be done off the books as capital was transformed into expenses.

* Though in practice, a lot of PFI/PPP projects have explicit or implicit government guarantees.


Blackbeard 11.20.15 at 2:45 pm

I’m a civil engineer with 40 years plus experience doing major infrastructure projects in the U.S. Quite of few of those projects were “privatized” but that term is so vague as to mean nothing.

If by privatization we mean private ownership then I quite agree it almost never makes sense. The cost of money for any remotely creditworthy public entity is much less that for a private entity even before you add in the private entity’s need to make a profit. I always told that to my public sector clients and, in my experience, the ones who wouldn’t listen were elected officials who wanted to do something big and move on to higher office before the long range costs began to bite. Naturally there were always plenty of contractors and investment bankers happy to help if that was the goal.

But if we broaden the definition to include construction (design/build and design/build/operate) and operations (contract ops) then I think privatization can offer real benefits. Essentially what privatization can do is bring competition to areas, for example, garbage collection, that formerly appeared to be natural monopolies. I have seen dramatic increases in efficiency in the provision of public services once they are opened up to competition. Of course, if you are going to bring private contractors in you need strong contract administration skills on the public side as these guys are hardly in it for the fun of it.


Peter K. 11.20.15 at 2:47 pm

I agree with most of what you say here except for the bit about monetary policy.

“The financialization of the global economy has produced a hugely costly financial sector, extracting returns that must, in the end, be taken out of the returns to investment of all kinds. The costs were hidden during the pre-crisis bubble era, but are now evident to everyone, including potential investors. So, even massively expansionary monetary policy doesn’t produce much in the way of new private investment.”

I agree that fiscal policy and infrastructure work *better* but that’s not to say that “massive” monetary policy doesn’t work.

Compared to “normal times” yes monetary policy has been “massive” but it hasn’t been compared to what’s needed by macro policy. Monetary policy has been compensating for unprecedented fiscal austerity. That’s a lot of work.

It may be more accurate to say that the financialization of the economy is creating the need for more macro (monetary, fiscal, trade) policy by the government in order to support aggregate demand. This has mostly fallen to the central banks.

I think it is mistake for the left to say that monetary policy is ineffective. This just reinforces the rightwing arguments that government macro policy doesn’t work:

“Fiscal policy is like sugar and monetary policy is like morphine.”

The Austrians argue monetary policy and low-rates are causing bubbles and “malinvestment.” I would counter we need more regulations and macroprudential policies to prevent bubbles and more macro policy to keep aggregate demand up. If monetary policy (or macro policy) is too loose, we’ll see inflation. Otherwise monetary/macro policy is helping support aggregate demand, job creation and wages and not simply creating bubbles.


Peter K. 11.20.15 at 2:49 pm

“extracting returns that must, in the end, be taken out of the returns to investment of all kinds”

Can’t they also be taken out of wages causing lower labor costs, or can’t they cause higher prices for consumers especially in monopolies which don’t have competitive pricing?


John Quiggin 11.20.15 at 4:03 pm

@Peter K Good points, which I will try to address in my next post


reason 11.20.15 at 4:21 pm

Peter K @28
I don’t see much reason for treating “Austrian” arguments seriously. It is not obvious that low interest rates cause bubbles, and I don’t think it is empirically supported. Low interest rates make marginal real investment profitable that otherwise wouldn’t be and bubbles are the result of more capital being available than can be profitably turned to real investment. And as far as “malinvestment” is concerned – if we are worried about “malinvestment” why isn’t everything centrally planned (properly this time). The fact is “malinvestment” is a fact of life in a decentralised society (with investment decisions made independently). And it is also hard to see how “malinvestment” is more of waste than involuntary unemployment. (As a test for how useful a guide it is – ask “Austrians” to pick which investments are malinvestments in advance. If they refuse, then ask them why “malinvestment” isn’t just a part of normal price discovery – people will get it wrong sometimes.)


Peter T 11.21.15 at 12:55 am

The contractor state is hardly new. It was the norm in, for instance, Britain 1750-1850. One firm maxim at all times was that, if you are contracting out, you need to reserve some part under your direct control (so have Royal Dockyards alongside private yards, or in-house IT alongside outsourced IT). Otherwise you will have no benchmark, no insight into the technical issues, no check on the inevitable gaming of the system (because you won’t know where you are being gamed). An old and obvious truth, consistently disregarded.


Sebastian H 11.21.15 at 5:56 am

I’m a bit surprised by the pushback on the idea that infrastructure projects appear to be very expensive in the US. See for example here <a href="; cost of developed country subways.

The thumbnail sketch is that US subway/tram construction is massively more expensive than similar projects anywhere else in the world. The only one even close is the Crossrail project in London, which is also massively more complex than any of the other projects.

And these aren’t cheap cities either
NYC 4 billion per/km, 1.7 billion per/km, and 1.3 billion per/km.
London Crossrail 1 billlion per/km
London Jubilee Line Extension $400 million per/km
Amsterdam $410 million per/km
Tokyo $280 million per/km
Berlin $250 million per/km
Paris $230 million per/km
Copenhagen $170 million per/km
Seoul $110 million per/km

So something appears seriously out of whack with US subway construction costs. Does anyone have comparable numbers on stadiums or other big city projects?


Sebastian H 11.21.15 at 6:04 am

Sorry about the link. Hopefully this works


Peter K. 11.21.15 at 2:32 pm

@30 @ 31

Thank you for your replies. Interesting ideas.


underdoglet 11.21.15 at 5:01 pm

Blackbeard, we have privatized garbage collection here. Competition opened up and now we have 6 or 7 different garbage collector trucks going down every block once a week wearing out our neighborhood streets. There are downsides to private garbage collectors in our urban areas.


Blackbeard 11.21.15 at 8:30 pm

Underdoglet: What you describe is not what I would call privatization, although, as I said, the term has become so vague as to be nearly useless. What your community has done is to franchise several collectors all of whom are then free to compete for business. As you say it is an inefficient and wasteful system. This is something that smaller communities, who cannot afford their own collection system and don’t have the management capabilities to run a privatized system, sometimes do. In a real privatized system the community would conduct some sort of bidding process and select a single contractor to provide collection services for the whole community for a defined period, usually around five years. Theoretically the bidding process keeps the contractors honest and having one company do the whole town improves the efficiency.


Bruce Wilder 11.21.15 at 9:12 pm

Theoretically the bidding process keeps the contractors honest and having one company do the whole town improves the efficiency.

I wonder why that is the theory. It seems like an idiotic theory to me.

It seems to me that any scheme, public or private, open-bid or municipal mandate, can and will be subject to all the ills and pitfalls, which ought to be familiar and accumulate in any polity. We have to reform. And, depending on the status quo, that might mean progress is in the direction of privatization or the direction of public provision.

There is no obvious single, ideal stasis. That does not mean every proposal has merit, judged from a public good perspective. What underdoglet described seems almost certain to be wasteful. In a polity where “competition” is an ideological shibboleth and opening the public to profitable predation a perennial project, it isn’t hard to see how such a silly scheme could get off the ground.

But, we can also imagine an incumbent monopolist corrupting politicians in the pursuit of an easy life of indifference to performance, or a municipal department of underpaid civil servants nickel-and-dimed into impaired effort. There are lots of ways for things to go wrong. Changing the rules of game in some radical way may be the way to reset behavior and expectations.

Comments on this entry are closed.