« A defense of employer-linked health insurance | Main | Why Most Published Research Findings are False (2) »
Broken Buildings, Busted Budgets
That's by Barry B. LePatner and the subtitle is How to Fix America's Trillion-Dollar Construction Industry.
The key problem is that building or new construction owners become completely dependent on information provided by their contractors. The contractors experience cost overruns and the commissioning owners have to suffer delays, cost increases, and the general feeling of having been screwed over. Opportunism and recontracting are rampant. According to the author, no institution successfully helps commissioning owners distinguish between good and bad contractors. Ex post the construction industry is not competitive, even though it has a large number of firms ex ante.
According to LePatner, the answer is to move to truly fixed-rate contracts. This means no more passing cost onto owners, real penalties for delays, and much more risk on the construction company itself. That in turn requires larger and more integrated construction companies instead of today's small proprietorships.
If this is such a good idea, why doesn't the market provide such a solution already? Surely there is enough capital to support large construction companies. I don't know, and for that reason I cannot evaluate the author's central proposal in an informed manner. Nonetheless this book is argued on a very serious level and I found it definitely a worthwhile and stimulating read.
A must for anyone interested in the economics of construction, and a possible read for others as well.
Posted by Tyler Cowen on September 20, 2007 at 06:00 AM in Economics | Permalink
Comments
Interesting.
When all details do eventually become public knowledge, the fantastic contract structures and management in constructing Terminal 5 for Heathrow in London would illustrate these principles nicely.
Some prelim information can be found here: http://www.baa.com/portal/site/default/menuitem.1b52c81dd9043acc0fb42410c02865a0/
No vested interest, but as a London resident and LHR user, I am just keen that BA's lost baggage stats will fall, that the airport will be less crowded and nicer, and LHR will stop getting the awful press that it does!
Posted by: Shefaly at Sep 20, 2007 6:56:21 AM
There is a pretty close analogy with computer systems. The problem with computer systems is that the client typically changes what he wants as the system is being built.
With construction projects, the same thing can happen. The client will make changes in the project along the way. Those changes affect the cost. You have to allow for that.
A fixed-number contract would tend to preclude clients from making changes after the contract is written. I'm guessing that this would be an inferior solution.
Posted by: Arnold Kling at Sep 20, 2007 7:14:11 AM
Agree with Mr Kling above. The projects tending to have the most overruns are the ones where clients demand a lot of flexibilty. Suburban homes when built by a developer don't have as many cost over-run issues.
The smartest developers therefore build the basic flat/house structure and leave the interior decoration to the future occupants.
Another point, why do clients want to make changes as they go along. But if you look around yourself then almost all manufacturing and service delivery systems have some felxibilty built into them. I think therefore that therefore the market therefore sorts itself into two groups- one mass markt builders which can crank out cookie-cutter houses at a super fast pace and small partnerships which specialize in aesthestic desgin and then sub-contract the actual work out to their long-time builders.
Posted by: sa at Sep 20, 2007 7:32:57 AM
Arnold Kling and SA are right on.
Two other important issues are sunk costs and bankruptcy. A fixed-cost contractor can go bankrupt in the middle of a project if it is more expensive than expected. It will then be optimal for clients to pay the entire contract value over again to a new contractor to get the building finished, since the value of the finished building is unchanged and the money spent so far is just a sunk cost. It is more efficient to just pay the original contractor more. A similar situation applies with overbudget Hollywood movies.
And yes, a better capitalized construction company may reduce the bankruptcy risk, but you need to pay for that capital, and you still need really good lawyers to make sure the contractors' lawyers haven't found a way to hide the capital from disgruntled clients.
Posted by: DK at Sep 20, 2007 8:15:03 AM
According to LePatner, the answer is to move to truly fixed-rate contracts. This means no more passing cost onto owners, real penalties for delays, and much more risk on the construction company itself.
I don't think this necessarily holds. There are real information problems in construction. Builders have lots of ways to cut corners that are invisible to all but the most obsessive, knowledgeable, and eagle-eyed, customers. So the cost is passed on in the form of shoddy workmanship, or possibly in the need to hire someone to closely inspect the work.
Arnold Kling is correct, also, to suggest that not allowing changes is a poor idea. As a project proceds, whether software or building, new possibilities emerge. Feature creep can be endless, but there are still many worthwhile things that can be done that were not obvious in the planning stage.
Posted by: Bernard Yomtov at Sep 20, 2007 8:59:24 AM
Some of us have good experience with contractors. It involves:
1) screening the contractor closely
2) not using the lowest bid as an absolute (asking for trouble)criteria
3) reviewing the building plans in complete detail, down to the last light switch
4) asking lots of questions and anticipating everything ("will that tile floor be easy to maintain?")
5) using a competent attorney to button down the contract
6) have a very detailed funds draw procedure and a very detailed change-order authorization process
7) include penalties and incentives when advisable
8) have the contractor bonded
I am always amazed when people and companies do one of the biggest transactions they will ever do and do it like buying a used car. Duh.
Posted by: save_the_rustbelt at Sep 20, 2007 9:04:59 AM
Some of us have good experience with contractors. It involves:
1) screening the contractor closely
2) not using the lowest bid as an absolute (asking for trouble)criteria
3) reviewing the building plans in complete detail, down to the last light switch
4) asking lots of questions and anticipating everything ("will that tile floor be easy to maintain?")
5) using a competent attorney to button down the contract
6) have a very detailed funds draw procedure and a very detailed change-order authorization process
7) include penalties and incentives when advisable
8) have the contractor bonded
I am always amazed when people and companies do one of the biggest transactions they will ever do and do it like buying a used car. Duh.
Posted by: save_the_rustbelt at Sep 20, 2007 9:06:06 AM
I suspect that this problem will go away, as long as we don't try to hard to commit to a solution: the flexibility sought by many construction customers will be replaced by improved computer modeling (of function as well as form). As it gets easier to develop a complete description of what you want, contracts based on such will become more common.
Maybe.
Posted by: Tom Myers at Sep 20, 2007 9:41:31 AM
This is similar to the argument in favor of fixed-price R&D contracts in defense procurement. Problem is--as several respondents have already pointed out--it keeps you from making changes while the work is in progress (since the contractor will always find a way to "get well" on them changes), it tends to wind you up with contract-compliant junk, and of course, there's always the risk that even the biggest firms will go sneakers up in the middle of the project: not on your average four-bedroom suburban house, to be sure, but on a major project such as an office complex or a mall.
Posted by: David Hecht at Sep 20, 2007 9:53:04 AM
I work for a developer, the concept of a fixed price contract is very much a part of virtually every major construction contract. It's called a GMP (Guaranteed Maximum Price) the problem is owners make changes and or there are unforeseen conditions that cause delays/cost overruns.
On a side note, what's wrong with the construction industry?
If your worried about cost overruns and delays I would argue most of that can be mitigated by smart developers and air tight contracts.
If your worried about cost in general, I have a lot of ideas starting with eliminating the power of unions, city inspectors, community groups and everyone else who likes to put their hand in the cookie jar.
Posted by: Fever at Sep 20, 2007 9:54:16 AM
The problem with construction is more a matter that anyone can pick up a hammer and a pick-em-up and go into business with very little licensing/regulation. States where there is a "defective home litigation problem" have a "defective home" problem because unqualified people are building houses. States with low incidences of such lawsuits have licensing requirements for all the site workers. Unfortunately for homeowners, the "cure" being pushed by the administration is to limit the recourse of homeowners to sue and recover damages from incompetant construction companies. And to limit damages, so that if the contractor builds the roof wrong, so that it leaks and damages your sheetrock and carpeting, they're only liable for the roof and for no other damages incurred by their incompetence. Contractors have squeezed the subcontractors so much that speaking Spanish is now a requirement for most subs, and hiring legal, documented workers will bust the sub's budgets.
Posted by: Peter at Sep 20, 2007 10:15:47 AM
Bernard makes an important point - the problem is information asymetries between the contractor and customer. The contractor can do one of two things mid-stream: increase price or decrease quality. Economically, the two are equivalent.
If you have a fixed price contract, then quality will become unfixed, and the contractor will likely cut corners (they may anyway). So not sure fixed price is really a solution.
Posted by: ah at Sep 20, 2007 10:17:43 AM
I think there has been an increase in the number of "design-build" projects and "construction manager at risk" projects, at least in the southwest. Usually, on those contracts, there are still mechanisms in place for the owner to negotiate changes after the project is underway.
There are lots of agency problems, also, since you have subcontractors, a contractor, an architect, and an owner in most projects. That many players is bound to lead to problems, and I don't really know how you'd fix it. Add to that the fact that a lot of medium-sized projects are municipal (schools, water treatment plants, etc.), and each of those players is affected by the fact that the principals for the project are only acting as proxies for the public, so now instead of using intuitive processes to make the project successful, they have to follow sets of rules that prevent fraud.
Posted by: kebko at Sep 20, 2007 10:44:53 AM
See the housing bubble? Know anything about zoning? other people's money? Construction is intimately intertwined with politics. Anyone who doesn't understand the politics will never understand the economics.
Posted by: Joel at Sep 20, 2007 10:50:48 AM
Excellent points by Bernard and ah above regarding the price/quality tradeoffs and one becoming unpegged if you hold the other constant. You could hold both of them constant but that would mean substantial monitroing costs and specialized knowledge - essentially like a credit rating agency for builders. There could be a market in this if the building quality rater amortizes it's cost of acquiring information about builders over many customers.
So why don't we see this happen? My guess is that unlike the debt markets where a small no. of institutional players buy a lot of debt repeatedly, most clients(I'm talking big projects here) engage in very sporadic orders to build projects. The role of quality raters has been instead replaced by advertising like in many other industries.
Posted by: sa at Sep 20, 2007 10:50:49 AM
"...truly fixed-rate contracts."
As another mentioned, the construction process is filled with many variations of such contracts so I wonder what the author means.
Could he be suggesting contracts with no change orders? Sounds very odd and of no benefit to anyone, contractor nor owner. And practically impossible really when it comes to geotechnical conditions as a contractor would have to carry an absurd contingency for that work.
Posted by: David Sucher at Sep 20, 2007 11:39:58 AM
Btw, Tyler, I don't believe that your central premise is correct. You write: "The key problem is that building or new construction owners become completely dependent on information provided by their contractors."
I don't think that is so. Sophisticated or even just diligent owners develop their own source of information to triangulate and verify costs. One can hire an estimator or even one's own "shadow contractor" to assess line-items for plausibility. As well, there is competitive bidding among sub-contractors.
Yes, it can be difficult to tease-out these costs; and the change-order game offers tremendous possibilities for abuse of the owner. But I don't quite get how by itself a "fixed-price" contract can solve the many problems inherent in any construction process, unless the number of projects going with "estimates" is larger than I think. I guess I will have to try to read this book.
Posted by: David Sucher at Sep 20, 2007 11:52:50 AM
What does the size of the company have to do with it? Defense contractors are truly gargantuan (Boeing,
United Technologies, GM), and they almost all run on a "cost plus" basis.
Posted by: Person at Sep 20, 2007 12:00:18 PM
One economic reason preventing the emergence of large integrated building contractors is bonding capacity. Many public works projects in California are joint ventures not because the prime contractors bidding can't manage the project on their own, but because they cannot obtain the bonding required on their own. So they joint-venture, where one partner is primarily providing bonding capacity in exchange for a percentage of the fees, and sometimes a small piece of the work.
Posted by: Anthony at Sep 20, 2007 12:20:44 PM
So far, nobody on this blog has mentioned the nightmarish process of getting a building permit. The problem with building is that it's not about building. It's about how much red tape you can stand. As long as a licensed, bonded engineer has signed-off on a project and assumed liability for it, the building department should allow it to proceed without their endless plan checking.
Posted by: Mike Sproul at Sep 20, 2007 12:56:09 PM
In general, construction is a highly regulated, unionized, and corrupt industry. The original post seems to ignore these anti-competitive factors. I sense that Tyler would benefit from learning more about the construction business at the micro- rather than macro-level. I wonder if the book deals with these issues?
Posted by: Ken Hirsh at Sep 20, 2007 2:47:44 PM
I thought the query was to the point that market forces" have not resulted in "adequate" consolidation efficiencies in construction contracting; not about basic operational defects and their causes.
To the original point, I would respond that there is another "market force," -- open entry, or at least relatively open entry.
You will find this the effective limitation because construction is not integrated, it is structured on the principles of General Contractor and sub-contractors, the latter having essentially "open or free" entry.
Posted by: R. Richard Schweitzer at Sep 20, 2007 2:58:14 PM
Lots of good comments already.
Another reason that true-fixed-price-and-conditions contracts are not common in the smaller markets (houses, small offices) is that they are expensive. The difference in price between a fixed-price and a cost-plus contract will be that on average, buying insurance costs something--the fixed-price contract insures the owner,a nd the contractor has to charge for that.
Also, it's worth re-noting that the randomness of government demands makes good prediction much more difficult.
Posted by: SamChevre at Sep 20, 2007 3:50:28 PM
Ken:
Only some segments of the construction industry are unioned, home building
is almost entirely non-union (and heavily into undocumented workers in some regions).
In some areas the trade unions still have great influence, and in some urban areas the
whole business is rife with corruption. In other areas neither applies.
Construction tends to be very segmented (home, commercial, highway, utilities) and each
segment has its own characteristics.
It is very difficult to generalize.
Posted by: save_the_rustbelt at Sep 20, 2007 3:52:48 PM
I would argue that where the construction business is highly inefficient, you will find some combination of excessive regulation, unionization, and/or corruption. Certainly, Boston's "Big Dig", the example given in the Amazon summary for the book, suffered massively from all of these problems. Home builders might suffer from fewer regulations, unions, and corruption, but it would be odd to accuse them of having productivity problems!
Posted by: Ken Hirsh at Sep 20, 2007 4:18:25 PM
