Construction documents - are they worse than ever?

Two women operating ENIAC; Wikipedia Commons

One of the presentations at the 2017 convention in Providence was a panel discussion titled Hot Topics and Emerging Trends, which included comments about the decline in the quality of construction documents. I found this to be an interesting subject, as I had seen many attacks on document quality over the years. Not only that, but I had made presentations on the subject.

In 1997, Michael Chambers and I presented “Document Coordination” for the Minnesota chapter of AIA. We discussed the roles of drawings and specifications, document quality, coordination techniques, short-form specifications, and MasterFormat 1995. Our handout included reprints of several articles about document quality; some, with scary titles, tried to prove that construction documents were atrocious and getting worse, while others how quality depended on coordination of construction documents.*

The frequency of problems in construction documents makes it easy to accept claims that they are getting worse. In 1997 I believed those claims, but I now believe the opposite. I would argue that overall, construction documents are better than ever before.

Since the presentation Michael and I made in 1997, I have continued to collect articles about the quality of construction documents. Most of the articles address current document quality, but a few discuss a change in quality. The main difference is, while the first group of articles describe specific problems, the articles that talk about changes of quality lack specificity. Rather than explain how documents have changed, they rely vague expressions of individual perception.

For example, the Construction Management Association of America (CMAA) has published several annual reports, often in conjunction with the Facility Management Institute (FMI). These reports frequently refer to a decline in the quality of documents, with conclusions based on comments obtained by surveying facility owners, but they do not include supporting information. I have seen thirteen of these reports, going back to 2000.

The reports consistently claim that quality of construction is a major concern, and sometimes say there has been a decline in the quality of documents. The 2003 survey report was the first to assert that “there is a general decline in document quality,” along with declining skill levels. There is no support for the claim, but the report does include an interesting exploration of reasons for that decline.

The 2004 survey asked, “Have you experienced a decline in the quality of design documents?” More than 70% of responders said yes. Even so, it’s worth noting that about 30% said documents at the beginning of construction were adequate or excellent.

From then until the 2010 survey, survey reports mentioned document quality only tangentially, noting that quality is always a concern, but making no specific reference to a change in quality.

From FMI/CMAA Eleventh Annual Survey of Owners

The 2010 report states about 30% of owners report that the quality of design documents worsened in the previous two years. That sounds bad, but the graph provides additional information.

Even though about 30% of owners said document quality had declined, more than 35% said there had been no change in quality, and 25% said they were better!

While we should know of problems with construction documents, cherry-picking statistics is unnecessary and unjustified.

The most recent CMAA report, published in 2015, states, “as major challenges, the poor quality of documents tops the list.” It goes on to say, “This finding is consistent with … the 2010 study, i.e., 34 percent said the quality of design documents had declined … and 33 percent made the same claim about construction documents. … as long ago as [2005] more than 70 percent of respondents had cited a decline in the quality of design documents.” Again, the report uses only some of the information; it uses its own reports as sources but adds nothing new. The only other reference to document quality appears in a graph that shows poor document quality is an urgent challenge for owners.

One of the articles Michael Chambers and I used as a handout, “Contractor Survey Finds That Specs Don’t Measure Up,” was based on a survey conducted by Engineering News Record (ENR) and the School of Building Construction at the University of Florida.

ENR sent surveys to 500 contractors and received responses from 120 of them. Asked about the quality of specifications, 37% were rated good, 35% were rated fair, and 17% were rated poor. Compared to drawings, 85% of respondents said specifications were “sometimes or even more often” of lower quality. They reported that more than 84% of specifications “sometimes, often or generally have major omissions.” Contractors complained that specifications are boilerplate and contained irrelevant information. As was the case with the CMAA reports, the ENR survey summary expressed only subjective opinions.

How can this be? 

In 1997, I accepted both claims about construction documents - that they had many errors and that they were getting worse. I had seen enough of them to know that defects were common, and because all I had heard about the change in quality was negative, I believed what I had read. In the time since then, I have noticed that every few years, the decline in construction document quality again becomes a popular topic. But, if document quality was declining twenty years ago, and has continued to decline since then, how is it that we can build facilities today that are more complex than they were in the ‘90s?

In a sense, this is the opposite of what we often seen in advertising. Every time a product is changed - and, I suspect, sometimes when it hasn’t changed - it is promoted as “New! Improved!” If laundry detergent, for example, has been improved many times since it was introduced, it should be perfect by now, but it’s not. And chances are, within the next year or two we’ll see more “improved” versions of many common products.

I contend that the quality of construction documents not only is not declining, but is, in fact, improving. Some of the improvement can be attributed to our tools. As software evolves, it makes it easier to avoid many types of mistakes. Both graphic and text processing programs now incorporate features that eliminate some problems, reduce the frequency of others, and help the user make correct choices. Also, the basic data used by computers has improved by becoming more standardized, and by being continually revised to incorporate real-world information. Many design firms have libraries of proven details and specifications that can be used as-is in many cases, and that can be easily modified to meet project-specific requirements. Building models now can incorporate complete, actual dimensions of structural elements, mechanical systems, and many products, allowing generation of more accurate dimensions, and software can analyze models to find conflicts.

I’m not saying documents are perfect. I continue to see mistakes in both drawings and specifications, and it’s likely they will never be eliminated. There will always be new employees who need to learn the correct way of creating drawings and specifications, there will always be new contractors and subcontractors who must learn how to use construction documents, and there will always be new products and processes that will challenge designer and contractor alike.

I see the problem as one of perception. Assume a typical project has 10,000 items. If 100 of them present problems, it’s likely that the 9,900 - or 99% - that were not a problem will be forgotten, and the one percent that didn’t work will be the ones that are remembered.

A word about boilerplate

As noted above, contractors and suppliers frequently complain about text that is repeated many times with little or no change. What they don't seem to understand is that some requirements do not change much from one project to another. Specifications aren't prose; they're documents that define products and processes used in construction. If a given window is used in two projects the specifications may well be identical because that particular window is required in both projects. Similarly, the general conditions may be identical in multiple projects, and even the supplementary conditions may vary only slightly from one to another.

Boilerplate isn't bad; it's necessary. However, the amount of boilerplate can be minimized by proper use of Division 01 and industry standards, and by elimination of redundancies and nonessential text.

What have you seen? Are contract documents getting worse? If you think so, please post a comment (below) to explain why, and provide evidence!

* Partial list of articles reprinted for 1997 AIA presentation: "Contractor Survey Finds That Specs Don't Measure Up,” “Contractors seek more detailed drawings, greater coordination,” “Field Interpretation and Enforcement of Specifications,” “Avoiding Liability in the Preparation of Specifications,” “Sum of the Parts: Complementary Documents,” “The Standard of Care,” “When Drawings and Specifications Conflict,” “Study pinpoints reasons for building problems.