Institutional Issues: Transparency
Continuing with my series on institutional factors relevant to construction costs, I’d like to turn to a culture of transparency, or lack thereof. It’s unfortunate that the exact breakdown of costs by items and factors that we’ve seen in Italy and Turkey and are seeing in Sweden does not exist in the English-speaking world. It’s further unfortunate that there is an adversarial relationship in the Anglosphere between the civil service and academic researchers like us or the broad public.
It’s a delicate subject, because the cultures of opacity we’ve encountered, the American and the British, certainly correlate with high costs, but we cannot be perfectly certain that they cause them. The peripheral Anglosphere learns many things from the US and UK, so it could just be part of the general correlation between Anglosphere membership and high costs.
That said, we do have reasons to believe this matters. The opacity we’ve encountered in the US and UK is so severe that it ensures there is no proper oversight. A system that punishes junior workers for reporting problems will just not know they exist. It’s best viewed as the Xi Jinping school of governance: demand that people follow the line and not air out problems, until subordinates lie to you just as much as you lie to the public, and local party officials arrest doctors who report to the public about corona while Taiwan is already warning the world about it.
The organization of information
Many episodes of Yes, Minister prominently feature the red boxes of papers for the minister to review. The civil service prepares documents every night for review, and the minister, who thinks he is a reformer, demands to know everything – so the permanent secretary’s office gives him interminable work to look at, down to and including stationery requisition. Needless to say, the minister does not come out of this experience informed about the department’s workings.
One of the obstacles we’ve encountered to a clean itemized comparison of construction costs is that in the US (and apparently also the UK), the information either does not exist or is not made public. We know how much Second Avenue Subway cost; we know how much individual stations cost and even how the stations break down between the civil work and the finishes, but each of these is still $200-500 million in unitemized costs, given as a lump sum contract. There are independent itemized estimates used as a benchmark, but they’re confidential, since the MTA uses them to rate contractor bids.
Any further breakdown we’ve seen is at the level of the minister’s red boxes, stating individual salaries and contracts for concrete and widgets; it’s not even complete information, since most of the work is subcontracted, and what’s subcontracted is opaque even to the independent cost estimators. To the extent we have estimates at a level that’s at all useful, that is high single-digit millions to low tens of millions, they’re cobbled together from many different examples, to the great frustration of people who were hoping for a perfect recipe for them to solve the cost problem.
I must stress that this is in a relatively cooperative environment. I don’t think Janno Lieber is sitting on a detailed breakdown of contracts to tranches of about $30 million and is just making sure we don’t know it. I doubt that this information exists in an organized fashion at all – it lives in the lore of numerous private-sector middle managers each of whom knows a few items.
An example from Italy
Italy has a well-known problem with tax evasion. Pellegrino-Zingales consider tax evasion among small businesses to be one of the root causes behind Italy’s economy stagnation in the last generation, arguing that it encourages firms to hire and promote by loyalty and nepotism (alongside patronage-based credit networks) rather than by merit, and that this has been an especial drag in the age of IT.
More recently, D’Agostino-de Benedetto-Sobbrio consider this question from the other end: what makes people choose to evade taxes? They look at the impact of government spending, proposing two opposed theories: the government as a grabbing hand, which taxpayers perceive as out to get them, and the government as a helping hand, whose spending helps ordinary taxpayers. The grabbing hand model predicts that bigger government leads to more tax evasion, the helping hand model predicts the opposite. While Chinese tax evasion follows the grabbing hand model, Italian tax evasion follows the helping hand model: Italian government spending induces the taxpayers to perceive the Italian state as on their side, reducing tax cheating.
All of this should be treated as background to the fact that, in Italy, the public data on construction costs and their breakdown is of very high quality. Marco Chitti has obtained breakdowns at the level that is useful for us for the upcoming Italian case study, and having read a draft of his report, I can speak with relative confidence (less than he can, of course) about wages, staffing levels, techniques, relative costs, and what the problem sections are.
Transparency and openness
That Sweden has very high-quality public data is probably not surprising to readers who know even a little bit about Nordic institutions. Here, for example, is the published breakdown of one set of contracts for Nya Tunnelbanan. Nordic transparency is a general feature, seeping to so many places, to the point that academic hiring committees in Sweden produce public-facing spreadsheets of all applicants and brief comments on them, and if the comments are positive, along the lines of “almost makes our shortlist but we have too many good candidates this year,” then applicants use that in their next application.
But when it comes to infrastructure megaprojects, we’ve found high transparency wherever we’ve looked in Continental Europe. Italy has the same information quality – because the Italian state works for the people of Italy rather than lording over them in secret.
This transparency extends to analysis of problems. The cost overruns on Grand Paris Express have led to a report by the Cour des Comptes about what happened, with detailed analysis and cost breakdowns (albeit not at granular enough level for our case studies). The report is earnest because the French state, elitist as it is, still works for the public, and acknowledges errors. Likewise, academic work in Italy on cost problems, such as Paolo Beria’s paper on the high costs of Italian high-speed rail, is in wide circulation. In Sweden, there is not only academic student work about the cost overruns of Nya Tunnelbanan but also a brief report a civil servant involved in the project sent us explaining the issue of mid-project changes in regulations.
Then there are the UK and US, where the situation is different. The UK, and countries it influences like Australia, barely even informs the public of the expected cost of a project until it is time to approve it; David Levinson told me that in Australia all communication about costs comes from leaks and trial balloons, unlike in the more open US. Even learning the highest-level breakdown of Crossrail costs required using a freedom of information request, and project-level questions about the cost of individual stations are often redacted (Crossrail 2 made available information about relative costs of stations, not a specific number per station). Reports about critical technological change like driverless trains, increasingly adopted in Paris, are not available to the public in London except through leaks. American governance is somewhat more transparent in the early stages, but key information about choices is hidden in confidential documents; the freedom of information process takes forever and officials freely redact documents or reject handing them over. The American and British freedom of information process screams, the government doesn’t work for you – its relationship with you is adversarial.
Transparency and language
In the last year or so, observing ever more central circles of political activism in the United States, I’ve realized something important: federal policymakers, and state policymakers who interface with them, speak an inscrutable language of bureaucrats who nudge but do not do. This is best illustrated with examples:
- The Bipartisan Infrastructure Framework has a line item about federal funding for previously funded but canceled projects, inserted by Maryland’s senators in order to fund the Baltimore Red Line, which Governor Hogan canceled for racist reasons in 2017. Instead of openly including it as an earmark, or else letting the federal process play it out, the language uses a circumlocution.
- At a meeting with activists, another advocate asked Beth Osborne of Transportation for America (T4America) about bus shelter. Instead of dealing with the direct issue, she gave a soporific answer about the need for federal standards, which may be Washingtonian for what in English would be rendered as “yes, I’ll do what I can to make sure the feds fund bus shelters,” or it may be Washingtonian for endless process and yet another round of red tape; not speaking the language, I could not and still cannot tell.
The contrast is with the concrete, plain language I see elsewhere from civil servants, to the point that it’s easier for me to go through the Cour des Comptes report, in a language I speak imperfectly, than to try to translate from Washingtonian to English. All of this matters – the use of a form of language designed to speak only to Beltway insiders is itself a form of opacity and American civil servants need to train themselves to on the one hand be more technical when necessary but on the other hand be very clear about what they’re doing.
Transparency as a goal
The path forward must involve treating transparency not as an imposition to be fulfilled through checklists, which produce red boxes, but as a positive goal. It involves ensuring agencies are helpful to regulators, academics, and the broad public, rather than hiding decisions behind walls, often because the reasoning behind such decisions is weak. An academic is expected to make data available to peers and the public, and so must agencies and regulators.
Trust in the civil servants is crucial for public infrastructure to succeed. Results can speak for themselves even in a low-trust system – streets really do come before trust – but the US and UK have poor results. The adversarial relationship with the public produces bad outcomes, and people whose expertise is in stonewalling and making excuses must be replaced with people whose expertise is in building things and accurately reporting on what they’re doing so that others can replicate their success.
Pedestrian Observations 
I was going to say this reminded me of the BBC’s “Yes Minister” when I got to part that references that great TV series from the 1980s. As for the opaque language of bureaucrats, once again Sir Humphrey said nothing while saying a lot, to the befuddlement of his minister/prime minister. I just bought all three books which compile the TV episodes as diaries of the “Right Hon. James Hacker” and they are just as good a read as they were watching. The lack of transparency in America is very frustrating, making it hard to identify the how and why of infrastructure projects, a good example being legislative members of the NYS Senate and Assembly still not knowing how the new Tappen Zee Bridge was funded by the State of New York.
However bad you think it is in the US, it is far worse in Australia. US journalists on assignment to Oz are constantly amazed at the level of government secrecy on all matters.
“demand that people follow the line and not air out problems, until subordinates lie to you just as much as you lie to the public”
Thermocline of truth
How do you rate the famous Willie Brown quote? He is completely transparent and unapologetic about how the government purposefully lies to public. The subtext is government officials are saying to the taxpayers “You can’t handle the truth”. [insert Jack Nicholson yelling from the witness stand in A FEW GOOD MEN].
Willie Brown, former SF mayor, San Francisco Chronicle – July 28, 2013
“News that the Transbay Terminal is something like $300 million over budget should
not come as a shock to anyone. We always knew the initial estimate was way under the
real cost. Just like we never had a real cost for the [San Francisco] Central Subway or
the [San Francisco-Oakland] Bay Bridge or any other massive construction project. So
get off it. In the world of civic projects, the first budget is really just a down
payment. If people knew the real cost from the start, nothing would ever be approved.
The idea is to get going. Start digging a hole and make it so big,
there’s no alternative to coming up with the money to fill it in.”
The fact that this admission did not lead to the collapse of the entire party machine under the weight of state and federal investigations is a blot on American government.
Not that willie brown is in any way trustworthy, but I do wonder if there is something to this. Namely in the US (and much of the rest of the anglosphere), a strong libertarian element in the public limits politicians’ ability to make “big” expenditures (or significantly raise taxes to finance them). E.g. The price of gas in the US is still lower than in much of Europe but is seen as a political liability for biden.
The rest of the Anglosphere did not have abnormally high costs until the 1990s-2000s.
It took that long for Reaganism/Thatherism private sector worship to work it’s way out to the provinces?
I bet Adirondacker is right. A central element of Reaganism is inserting extra middlemen, who are profiteering and extracting wealth, and it is in their interest to conceal this wealth-extraction by creating opacity.
Despite the lower prices per gallon, Americans spend more on gas per capita than Europeans, because they have to drive so much longer distances.
They did have real costs, but if they had used those the projects would never have got funding…….so you convince yourself you can remove contingency and lower the costs and get started…….
I think something is getting lost in translation here. Are you sure you aren’t just comparing Design Build (DB) in the US to Design-Bid-Build in Europe (DBB aka traditional procurement?) I.e. are you conflating delivery method with location?
For large infrastructure projects in the US, DB is lump sum, whereas DBB is typically itemized. DB may have a “schedule of values” but that’s really in essence a set of milestones used for progress payments. DBs also generally have some (should be a small percentage) of force account (FA) work which is done on a time-and-materials basis, and as you said can be a nasty accounting mess of labor hours, equipment hours, and supplier and sub invoices. FA is necessary evil for some kinds of work, so if you’re not seeing it in European contracts you’re probably not looking hard enough (sometimes it’s inappropriately used and becomes an unnecessary evil, but that’s a different point).
Is it possible to do itemized DB? Is that really being done in Europe, or are you confusing with DBB? I’m genuinely curious about how that could work. What happens if there’s an overrun? Do you pay the design-builder for a bust in his design? Seems like a very perverse incentive. And if you don’t pay on quantity times unit price, how is that really an itemized contract?
I agree that DBs have a transparency issue, but I don’t think that’s the case for US public agency DBB – itemized contractor invoices for the latter can be found online, at least for some agencies.
Maybe your real argument here is that the US does too much DB for large infrastructure projects, and we should reverse the trend and go back to more DBB? Or maybe
Second Avenue Subway Phase 1 was DBB and lump-sum, and the itemized bits that exist are buried under layers of trade secrets for both the private and public sectors. The DB bug in the United States is very recent, and has emerged more or less simultaneously there and in the rest of the Anglosphere, which came to believe it is superior in the 2000s.
Is that the project that you mentioned had an independent cost estimator? Sounds like a CMAR, although I’m used to seeing the bid tabulations for the ICE vs. CMAR published as public info.
An evolutionary view of policy would suggest that transparency thrives in places where it is rewarded, while opacity becomes the norm in places where being forthcoming is mainly a risk.
I suspect continental transparency might stem from an emphasis in administrative law on demonstrating proper procedure, while opacity becomes the norm, if anything you say can be used against you in a court of law, as the standard American disclaimer goes.
The US loves proper procedure, so it keeps developing these thick books of procedure and requires hundreds of pages of reports (to provide surface area for citizen lawsuits) and checklists.
That might be the case, but what I mean is that broadly speaking continental law assumes an agency has certain responsibilities towards citizens and once it has demonstrated that it has fulfilled those responsibilities, it’s clear to proceed. There are rules, you know what the rules are and you can start digging a hole if you follow the rules. Thus it makes sense to demonstrate how well you’re following the rules.
Common law-systems, on the other hand, seem to be more centered around rights, which are often quite nebulous in their definitions. There’s a lot of compliance, but ultimately it doesn’t protect you from a negative ruling. It’s like playing Calvinball, you never quite know what the rules are, so you keep mum so as to not incriminate yourself by accident.
This philosophical difference can be seen in other domains of construction as well. On the old continent, building permitting is generally by-right, so you are virtually guaranteed a permit if your plans match the zoning ordnances. In the UK, on the other hand, getting a permit to build something is discretionary (although there are a lot of national guidelines and bulletins on the topic). The US is a jumble of different approaches, with Californian jurisdictions being the most discretionary.
https://www.economist.com/special-report/2020/01/16/politicians-are-finally-doing-something-about-housing-shortages
Not sure where you are getting your information but in the US detailed cost estimating is undertaken at each stage of design development for Design Bid Build projects. As the design develops the amount of contingency is decreased in an attempt to develop a genuine Engineers estimate for the project bidding costs. The one factor no owner or Engineer can really take into account of though is the market forces that the Contractor prices in. On many non transit related projects that I am aware of the Engineers estimate is often within 5 to 10% of the accepted bid. The ASCE has different levels of estimate that are well understood and used throughout the infrastructure industry. When developing these estimates we use the same estimating software that Contractor’s would use and develop the estimates form the bottom up. We make use of the various cost databases that exist that list out construction commodity costs, labor costs can be had from prevailing wage or collective bargaining agreements. So detailed estimates can and are developed from the detailed design drawings. Certainly for the MTA projects I have worked on very detailed cost estimates were developed for budgeting purposes.
But here is where things get funky.
The MTA for example will require the Contractor to fill out the Schedule E which is a list of “Bid Items” that when totalled up get you to the lump sum price for Design Build Projects. But no quantities are given by the MTA. So the Contractor has to scramble and develop quantities and price the job in the 3 months typically that is given to turn the bid around. Compare and contrast that to for projects I have worked on in Hong Kong or Turkey where detailed Bills of Quantities are developed that cover every element of the project and for which Methods of measurements are provided so that what is included in the quantity is well understood. Hong Kong for example has a Standard Method of Measurement. One of the biggest items of dispute in the US will be on rebar quantities. In other places I have worked the Engineer/Designer will prepare detailed rebar quantities and include them in the drawing set which makes it very easy for the Contractor to price as this will also include the laps and splices etc. Only handling steel will be omitted by the designer as we may not know the means and methods of placement that the Contractor will use, but the Contractor will and will have historical records of this allowing him to develop a unit rate for the design steel and the handling steel as the unit rate has to cover that and wastage. And then there is a fundamental difference in project administration. The MTA for example don’t want the hassle (in their mind) of having to remeasure everything during construction, hence the Lump Sum and everything is now on the Contractor unless there is a change approach, whereas other places prefer to remeasure each month and pay for what is installed with a reconciliation of quantities at the end of the project. Hence why the MTA for example does not provide detailed quantities, as they also don’t want to be on the hook for quantities and neither does the Engineer due to liability issues, and hence why the detailed costs are tricky to tease out of MTA/US projects. Having said all that a Detailed Cost Breakdown is required to be submitted once the project is awarded that list the value of each line item in the schedule. But again its a value not a cost and includes whatever risk the contractor wants to spread to the particular line item of the DCB.
So there are fundamentally different drivers between different parts of the world and between different agencies, in the US not all agencies do what MTA does but its commonplace. Meaning deriving the real costs for activities is always gong to be difficult. Also I can tell you that
Design Build though in the US is always going to be Lump Sum with no transparency. Why would a Contractor advertise what his potential profit is? Heck even our design fee is provided to the Contractor as a Lump Sum with limited breakdown for the same reason as we have to cover risk and profit somehow, we have multimillion dollar design contracts based on a single sheet cost proposal. Obviously we have an internal detailed cost build up, but if we have to take the risk of a lump sum covering redesign for contractors preferences then there is no way he gets to see how much contingency money I have built into my cost. If you want transparency then it has to be Time and Materials and arguing over every hour spent on the project and what someone’s rate is……the accounting will cost almost as much as the contingency in a Lump Sum. If you use the NEC Target Cost style contracts in the UK then you may have a better chance of understanding what the costs are. But your never going to have a Bill of Quantities remeasured Design Build because to do that you need the design finished so you can take off the quantities and then your back to design bid build. I can just imagine the chaos that would cause on the systems elements of a DB contract……the biggest problem with DB in the US is that there are practically no transit agency clients with the skillset to implement it properly, they just seek it as a risk and liability shedding exercise from them to the Design Builder, but still want to meddle and insist on their preferences allowing the Owners Engineer (who really should not exist) to meddle and interfere with the design progression. The water industry in the US is using DB and progressive DB quite effectively as are some highway agencies for straightforward projects, but when it comes to mass transit…hopeless.
There are detailed independent cost estimates, but they are trade secrets; the agency does not publish the results, but instead keeps them private as a way of rating the contractors’ bids. Then a lot of items are hidden behind subcontracting, hence our reliance on consultant and contractor estimates for the labor:materials:equipment split rather than exact numbers.
Actually many agencies publish the Engineers estimate alongside the contractors bids on DBB projects. At least the headline number. So that everyone can see where the numbers lie. Increasingly when we bid for a design project we are asked about our estimating capability and have to provide examples of our estimates versus contractor bids. Because of this we keep a database of costs from projects, production rates etc that helps inform our estimates. But yes owners rarely publish the detailed estimates with individual line item costs. Why should they? Costs are considered confidential during a bid as why would contractors want their overheads, profits etc broadcast across the industry. So as much as these are trade secrets then no they won’t get published. The detailed cost estimates are used to review the contractors bids but if it’s a low bid project it’s not that relevant, although you may look to see where rates are loaded or unloaded which may give you a hint as to where a contractor sees the possibility for a change. But ultimately Lo price wins.On a negotiated contract they are more heavily used to look for risk items and decided whether you can adjust the contract to better share the risk for example, usually be creating an allowance item and reducing the risk money built into the contract.
I’m really not sure what your argument is about subs. Subcontracting is used the world over. But ultimately the Prime is responsible for the price. The bids I have reviewed in multiple countries never show sub prices, it shows the price for the job. Let’s face it a contractor will have two or three subs providing pricing for parts of the work and after award goes back to them and negotiates them down. The project price though is owned by the Prime as for as the owner is concerned. So the price you see is the price of the work including managing the subs. If that’s what you mean by hiding behind layers of subs then maybe, but no project is happening without subs especially in the US.
The difference is more local than you might think. I’ve always found NICTD (South Short Line) process to be remarkably transparent. (And I don’t think Indiana is any good at transparency in general, so I really do think it is per-institution.)