In Defense of Stick Framing, Part 2: Speed Demons

Someday robot carpenters will frame homes in minutes with materials 3D-printed on site or teleported from the lumberyard. This four-part series examines some of the issues we should probably deal with while we're waiting.

Offsite construction systems—or more accurately, offsite done the European way—have been the hottest topic in construction in recent years.

For analysts and academics, the rise of offsite is the culmination of close to 90 years of appeals to the industry to modernize what they call "virtually the sole surviving large-scale hand industry in a machine age" (Final Reports of the President's Conference on Home Building and Home Ownership, 1931).

To say they're excited is a serious understatement. In an Oct 2018 post at the blog for its Industrialized Wood-Based Construction Conference, Forest Economic Advisors called offsite "the panacea to the housing crisis" and declared that housing is at long last "getting its own 'Model T' moment."

FEA is convinced that moving home building from jobsites to factories "vastly reduces build times, costs, pollution, and waste, and makes housing affordable."

The purpose of this four-part series is to examine those claims. Part 1 looks at construction waste; you can find it here.

Time is Money

Claim No. 2: Offsite construction vastly reduces the time it takes to build a home.

Cycle time is the average cost per day for each day a home is under construction. One common way to measure it is from the day the lot is purchased (which triggers loan interest) to the day of the closing (when the loan is paid off).

Framing the dried-in shell—i.e., building the structure, then covering it with rough sheathing—is the single most time-consuming task in that process. When you stick-frame, it happens entirely on the jobsite. When you prefabricate, time is consumed both on the jobsite setting components and in the factory designing and building them.

Prefab components are always faster than stick framing if you only consider the time you spend on the jobsite, and there are good numbers to prove it.

In 1995 and again in 2015, the Madison, Wis.-based Structural Building Components Association (SBCA) conducted a demonstration project it calls Framing the American Dream. Two four-man framing crews built two identical single-family homes. One home was completely stick-framed; the other was built with roof trusses and wall panels, with only the floors stick-framed on site. SBCA tracked the time it took each crew to build each dried-in shell.

The stick framing crew took 375.5 man-hours to complete the project, or 9.4 ten-hour days for a four-man crew. The component crew got the job done in 152.1 hours on the jobsite, which works out to 3.8 ten-hour days. That's a 60% savings vs. stick framing.

European-style offsite construction is even faster by virtue of the fact that the floors are panelized. Builder magazine reports that Ripon, Calif.-based Entekra erected a 3,200 sq. ft. home in three days in 2017, a 70% savings vs. stick framing.

The thing is, you don't need a time study to tell you components are faster than stick framing if you only count time spent on the jobsite. Common sense will tell you that. The question is whether it's a fair comparison.

If we were talking about construction waste, it's pretty clearly not fair to compare waste generated only on the jobsite. 100% of the waste produced by the project is on the jobsite when you stick frame. If you use components, nearly all the waste is in the factory.

A true apples-to-apples comparison covers all the waste produced during shell construction: 2% to 3% of total material consumed when you build offsite or 5% if you stick frame. So shouldn't the same principle apply to cycle time reduction?

Depends on how you look at it. Just like waste, nearly 100% of the time consumed is on the jobsite when you stick frame (minus a few hours to build and deliver material packages). When you panelize, more than half of the time invested is spent in the factory designing, engineering, and building components.

Ultimately the builder pays for that time, too. Following this logic, a true apples-to-apples comparison would count both the time spent on the jobsite and the time spent in the factory.

But the counterargument also makes sense. Some would say the time spent in the factory designing and building components shouldn't count because it happens concurrently with other activities. Whether the builder pays for that time is irrelevant—the only cost we're addressing here is cycle time cost and time spent in the factory doesn't add to that.

It's a valid point. Every construction project consists of a long list of tasks: 20 or 50 or a few hundred, depending on how granular you get. Some have to be completed in sequence (you can't start framing until the foundation is cured). Others can be completed concurrently (rough-in plumbing and electrical can be done at the same time).

Construction management is all about minimizing the gaps between tasks. If you can turn a sequential task into a concurrent task, that's always a win. The appropriate question here is whether it's a big win or a small one.

Under the Hood

As defined by the NAHB Research Center (now Home Innovation Research Labs, or HIRL), cycle time consists of four components: 1) the cost of money (loan or invested capital), 2) management costs (overhead), 3) new sales opportunities (the faster I build, the more homes I can sell), and 4) reduced labor costs.

Those four are combined to calculate an average cost per day that accrues from the day the project begins to incur cost (typically when you purchase the lot and trigger interest on your construction loan) to the day the finished home closes and everyone gets paid.

When NAHB's analysis was first published in 1998, the examples the author used worked out to a total cost of $291 per day per home. But he also noted that "other builders calculate a day to be worth between $50 and $500." In other words, the value of cycle time can vary dramatically depending on your situation.

Adjusted for inflation, $291 in 1998 equals $454 now. Builders' calculations vary widely, but generally speaking, large production builders have always valued cycle time cost higher than local builders. These days you'll see numbers as low as $100 or as high as $1,100 per day or sometimes even more. Here's what goes into them:

The cost of money plus overhead represents just under 23% of the total. That's not huge, but it isn't insignificant, either. The catch is that it's not easy to put a serious dent in loan interest or fixed overhead.

New sales opportunities make up slightly less than 26%, but this is one area where you can have a significant impact. If you can sell 24 homes in the time it used to take to sell 20, you just gave yourself a 20% pay raise. But only if you sell all 24 homes. That's easy when the market is booming and you're selling homes as fast as you can build them. But the value of new sales opportunities can evaporate quickly when the market goes soft.

Reduced labor costs are by far the biggest opportunity to reduce cycle time cost. HIRL explains how: “Streamlined production processes allow contractors to complete jobs in less time and produce more homes with the same crews. While material costs may not change significantly, there are substantial savings in contractor labor costs. Builders can share in those benefits.”

Sometimes they can, but not always. Builders pay framing subs by the square foot, not by the hour or day. If they want to use components to reduce labor costs, they need to go to their subs and say, “I can build 60% faster if I use wall panels, but I need you to give 60% of your paycheck to the component manufacturer.”

Go ahead, try it. See what they say.

Resistance from framers is the No. 1 obstacle to panelized construction and it's hard to blame them. From their perspective, panels simply take money out of their pocket and put it in the pocket of the panelizer.

That resistance does appear to be diminishing. Some believe framers have finally seen the light and begun to embrace efficiency. Maybe. Or maybe there's just so much work available that they're not worried about where the next job will come from. Right now, that is.

But the important thing about cycle time is that it's a little bit theoretical. The value is all over the board and the opportunities for improvement depend as much on your situation as they do on the effort you make.

That is not to say it isn't worth pursuing. But the notion that offsite "vastly reduces build times" is...well, let's just call it an oversimplification.

The Finish Line

The Census Bureau has been tracking the average start-to-completion time for new single- and multifamily homes since 1971. In 2017, Census estimated the average time to build a single-family home at 6.5 months. That's 197.7 days.

The average size of a new single-family home in 2017 was 2,631 sq. ft. That's smack in the middle of Entekra's sweet spot, so let's assume you'd get the maximum benefit from offsite: A dried-in shell that would have taken ten days instead takes three.

If your average build cycle is 198 days, switching to offsite construction reduces it by 3.5%. That's not chump change—seven days at $500 per day is still $3,500. But Entekra claims that its FIOSS? method (Fully-Integrated Offsite Solution) "reduces the overall build cycle by up to 33% compared to stick framing."

Not by itself it doesn't. What FIOSS does is enable a builder to reduce build time—maybe by less than 33% or maybe by 43%. Whatever the final number, it's a lot more than 3.5%.

It doesn't really take 6.5 months to build a typical single-family home. If it takes that long, you spent three to four months building the house plus two to three months fighting delays.

Some delays can't be helped—maybe you got hung up by an inspection, for example. But more often, one task takes longer than it should have so the next sub can't start on time. That throws the next sub off schedule and the next one after that, and before long you've got the kind of cascading failure that causes builders to drink brown liquor and beat the dog.

The upfront design and engineering work that an offsite supplier like Entekra does simplifies and clarifies details that would otherwise need to be worked out in the field. It identifies bad specs so they can be fixed (e.g., oversized headers), and resolves conflicts between systems (e.g., how to route HVAC ducts and plumbing pipe without chopping the framing to pieces). Nailing down those details in advance helps everyone finish on time and start on time, and that reduces your build cycle.

If European offsite construction is in fact revolutionary, this is the reason. Forget about R2D2 on the assembly line—the big hairy deal in all this is that you get a squeaky clean building plan and a dried-in shell handed to you on a silver platter.

That's worth 3.5% and you can get more if you want. All you have to do is manage the rest of the project from start to finish a whole lot more aggressively than most builders do it these days.

Yes, that's a little like saying my nutritional supplements will improve your health if you also exercise regularly, eat more fish and less beef, quit drinking, stop smoking, and get at least nine hours of sleep each night.

But construction is complicated. There is no magic panacea to the housing crisis. The big benefit of European offsite construction is that thorough planning and preparation are baked into the process so those steps can't be skipped.

That's a good thing, but to whatever degree those steps have or have not been ignored over the years, builders are still making significant progress reducing build times.

In 2017, average start-to-completion days for a single-family home were roughly 3% above their 45-year average. That's not terrible. Also not impressive—until you adjust the numbers to account for the fact that the average single-family new home is 58% larger than it was in 1973.

In 2017, average start-to-completion time per 1,000 sq. ft. built was 18% below its 45-year average and a whopping 32% faster than a single-family home in 1973.

Yes, offsite played a role. Declining cycle time during the 1980s and 90s appears to correlate pretty closely with growth in the use of roof trusses.

But the majority of those gains were achieved without abandoning stick framing: Tools got better, delivery and material handling equipment got more sophisticated, and across virtually every category, products got simpler, more durable, and easier to install.

Moreover, while this was all happening, offsite construction systems—modular, panelized, and precut packages—were losing more than half of their market share.

Recent analysis from NAHB found that offsite accounted for just 3.3% of the single-family market in 2017. "In 1997 and 1998," noted chief economist Robert Dietz, "7% of single-family completions were modular (4%) or panelized (3%). This marked the largest share for the 1992-2017 period."

Proponents of offsite construction would almost certainly say that new technology—BIM, production and logistics software, and automation—will turbocharge efficiency and reverse that trend. No argument here. It seems obvious that we're only scratching the surface of what construction technology—"contech," as Silicon Valley calls it—can do.

But there is also a turnaround to be had in stick framing. Pretty much everyone agrees that jobsite productivity has been deteriorating. The closest anyone has come to putting numbers on the decline was in Reinventing Construction, a 2017 study by the McKinsey Global Institute. By MGI's calculations, dollar output per person in the framing trade fell at a rate of more than 3% per year from 2002 to 2012. Industry insiders might say it was in free fall long before 2002.

Most insiders would also agree on the reason: Carpentry has become a deeply unpopular career choice. Maybe it's just demographics. Maybe it's the fact that well-paid jobs with benefits and pensions for framers have disappeared with the decline of unions.

Whatever the reason, you have to really like working with your hands and really really like working with wood to want to be a carpenter. There are people out there like that, but for the most part, framing houses is a job you take only if you can't find anything else.

So far there doesn't seem to be a solution in sight. That doesn't mean there never will be, and that has to make you wonder: If carpentry was a viable career choice that attracted ambitious young workers, how productive could stick framers be?

The late Larry Haun was a framer, author of The Very Efficient Carpenter and numerous other books on building, and a long-time contributor to Fine Homebuilding magazine. At KeepCraftAlive.org, Haun recounts his experience as a union carpenter in the early 1950s, framing homes with his brother Jim in Los Angeles’s San Fernando Valley:

“For $90, we would frame a 900 sq. ft. house on a slab. It had two bedrooms, one bathroom, and a gable roof. A three-bedroom, 1,100 sq. ft. version went for $120, which included setting door jambs and window frames, putting siding on the front, and making it ready for stucco on the outside and plaster on the inside. My younger brother Joe joined us and soon the three of us were framing one of these houses every day (my italics).”

In SBCA’s 2015 demo, it took 152 jobsite hours to erect a 2,900 sq. ft. panelized home with a four-man crew. Assuming ten-hour days, that’s 763 sq. ft. per day or 191 sq. ft. per crew member per day. The 3,200 sq. ft. home Entekra built in 2017 was built in three days with a three-man crew. That works out to 1,067 sq. ft. per day or 356 sq. ft. per crew member.

Keep in mind that in both cases we're only counting time spent on the jobsite, not the time in the factory designing and building components.

Haun and his brothers took delivery of a pile of lumber without a truck-mounted forklift or a crane to place the material. There were no pneumatic tools; everything was hand-nailed. The only power tools were a drill and a Skil 77 worm-drive saw that weighed more than 20 lbs. Window and door frames were fabricated on site. Interior walls were prepped for plaster, not drywall. The floors, walls and roof were all sheathed with 1x8s, not CDX or OSB.

Yet the Haun brothers routinely framed houses at a rate of 1,100 sq. ft. per day, or 367 sq. ft. per crew member per day.

Coming in Part 3: Offsite promises to make homes more affordable. Promises are easier to make than to keep. CLICK HERE to be notified when Part 3 is published.

Greg Brooks is editor of LBM Executive and moderator of the Executive Council on Construction Supply. He is a steering committee member at the Harvard University Joint Center for Housing Studies and a 49-year veteran of the construction supply industry. 303 845 4880