Chapter 6 Further Reading — Risk Management

Nine pointers, ordered roughly from most immediately useful to most theoretical. Where I am unsure of a document's exact designation or current edition, I describe it rather than cite it — you should do the same, and always work from the current version, because these documents change.

1. The standard contract document families — AIA, ConsensusDocs, EJCDC, and FIDIC. These are the primary sources for everything in §6.3. The AIA A201 family of general conditions is the most widely used starting point on American building work; ConsensusDocs was developed by a coalition including contractor associations and tends toward a more collaborative allocation; EJCDC's documents are engineering-oriented and unusually clear on differing site conditions, which matters on civil and utility work; FIDIC's suite is the international reference, and its different books deliberately allocate risk differently depending on who does the design. Read the differing-site-conditions, delay, indemnity, and insurance provisions in at least two of these families side by side. For: anyone who will ever negotiate a contract, which is everyone in this profession. Publishers make current versions available for purchase; many public agencies publish their own modified general conditions online for free, and those modifications are often more instructive than the base document.

2. Your own company's closeout reports and job cost history. I am serious about this being first among equals. Twenty completed jobs of your company's own work is a better predictor of your risk than any published table, because it reflects your market, your subcontractor pool, your staff, and your habits. Ask your project accountant for the final cost reports on the last ten jobs in your sector and read the variance lines. For: everyone. Free, immediate, and almost universally ignored.

3. AACE International — recommended practices on cost estimate classification and on risk analysis and contingency determination. AACE (the Association for the Advancement of Cost Engineering) publishes a library of recommended practices, several of which bear directly on this chapter: how to classify an estimate by the maturity of its design basis, and how to determine contingency by both deterministic and probabilistic methods. Its recommended practice on forensic schedule analysis, RP 29R-03, is the reference work on delay analysis methods and will matter to you in Chapter 33. For: estimators, schedulers, and anyone who has to defend a contingency number to a skeptical audience.

4. The Project Management Institute's body of knowledge — the risk management process area. PMI's framework — identify, perform qualitative analysis, perform quantitative analysis, plan responses, implement responses, monitor — is the vocabulary most owners and program managers use, and knowing it lets you talk to them in their own language. Treat it as a taxonomy rather than a method: it is stronger on process and weaker on the construction-specific content that makes a register useful. For: readers pursuing the PMP or working for an owner's program management organization.

5. The Society of Construction Law Delay and Disruption Protocol. A practitioner-oriented guidance document on delay, float ownership, concurrency, and disruption. It is not law anywhere, and it says so; what it is, is a careful and widely cited attempt to state what reasonable practice looks like on the questions this chapter raises about schedule risk and who owns a buffer. Free to download from the Society's website. For: anyone who wants to understand why "float belongs to the project" is a contested statement rather than an obvious one.

6. OSHA's construction standards, 29 CFR 1926, and the NIOSH research literature. Read these alongside §6.9. The regulatory text tells you what the minimum is; the NIOSH research on how construction injuries actually happen tells you why "the crew was running behind" belongs in an incident investigation. Both are free from the agencies. The Focus Four framing — falls, struck-by, caught-in/between, and electrocution — is the right place to start when you are deciding what a safety row on your register should actually say. For: everyone, and specifically before you write the safety row on your own project's register. See also Appendix F.

7. Your surety and your insurance broker. Not a publication — a conversation, and one that is free and that almost no project engineer ever has. A surety underwriter can explain what actually moves bonding capacity and what a loss job does to it. A broker can walk you through what your builder's risk policy covers during commissioning, what its water-damage sublimit is, and who is a named insured. Both of them will make an hour for you if you ask, and both know things about your company's risk that you do not. For: anyone who wants §6.8 to stop being abstract.

8. Daniel Kahneman, Thinking, Fast and Slow. The general-audience source for the cognitive material in §6.9 — the planning fallacy, the inside versus outside view, anchoring, and why we are systematically bad at exactly the estimating task this chapter asks you to perform. The construction application is not in the book; you supply that. The mechanism is. For: anyone who has ever wondered why smart teams estimate optimistically over and over and never learn.

9. Bent Flyvbjerg's research on megaproject performance, including How Big Things Get Done (with Dan Gardner). The empirical literature on why large projects overrun, and the argument for reference-class forecasting — pricing your project from the distribution of outcomes on similar completed projects rather than from your own plan. It is the outside view of §6.9 turned into a method. Read it critically: the megaproject evidence base does not translate cleanly onto a $6.8 million community center, and the book is more confident than the underlying data in places. It will still change how you look at a schedule. For: readers who want the strongest available argument against trusting their own estimate.

Two cautions on sources

Published contingency percentages and unit costs are starting points, not answers. Any table telling you that a healthcare project "requires" a specific contingency percentage is describing somebody's averaged history, not your job. Use published ranges to check whether your own number is plausible, never to replace it.

Anything legal varies by jurisdiction and changes. Notice deadlines, the enforceability of no-damage-for-delay and pay-if-paid clauses, retention limits, lien procedures, prompt-payment statutes, licensing, and prevailing-wage classification are all state-specific and all move. Nothing in this chapter or this list is legal advice. When a register row's response is a contract clause, the question to bring your attorney is: does this clause do what I think it does, here, this year?