Roots in the Mud

Sonic testing on foundation concrete deep beneath San Francisco detected an alarming anomaly. David M. Wilson, vice president at the builder Clark Construction, was looking at proof of a void in the bedrock socket deep underground. A lowered camera confirmed an open gap, and subsequent checks turned up further flaws, requiring structural repairs across 42 supportive pillars. From the builder's perspective, Wilson later pointed to those underground hurdles as a major contributor to a 6-month construction delay.

A massive high-rise cannot simply rest on dirt; crews have to bore concrete shafts deep into the ground and secure them to solid rock. That subterranean snag points directly to a question from a curious listener. Even with satellite guidance and digital modeling, raising buildings upward takes longer than ever, driving up the cost of new urban space. To explore that mystery, we look at why Salesforce Tower reportedly took from 2013 to 2018 and over 2.5 million man-hours according to its official website, whereas the 102-story Empire State Building debuted after a reported 18 months.

At first glance, that timeline seems like a staggering loss of efficiency, but the underlying system has fundamentally changed. Around 1930, crews of 5,000 workers erected 4 floors each week to rein in loan interest and capture rent, even tacking on extra stories mid-design to edge out the Chrysler Building. Modern towers instead face soils that can liquefy during earthquakes and must comply with demanding life-safety codes. How builders nearly a century ago stacked steel floors faster than modern machinery can even dig its footing starts right under their boots.

Geology explains a large part of that contrast. In midtown Manhattan, earlier crews had to shovel through only a shallow layer of dirt before hitting solid bedrock. By comparison, those foundation hurdles in San Francisco meant anchoring Salesforce Tower through the soft, liquefiable mud introduced earlier. To keep the structure upright near active faults, engineers defending modern timelines point out that crews had to sink 42 load-bearing elements up to 70 feet into rock, reaching 310 feet below grade over a 14-foot-thick foundation mat.

The foundation contractor's president, Giancarlo Santarelli, noted that building each element reportedly took up to five days. Yet anchoring those shafts in stone was only the beginning of the engineering workload. Rather than sticking to standard prescriptive codes, the team used performance-based seismic design—running computer models of major earthquakes to show that a custom-designed skyscraper can survive them. Structural engineer Ron Klemencic explained that the sheer complexity of the approach led San Francisco's Administrative Bulletin 083 to mandate an independent review by an outside engineering team. Engineers also conducted advanced computer modeling to see how the tower, the ground, and the transit terminal next door would behave together in a violent tremor. Seeing months drain away into drilling rock and running digital simulations framed the debate over that central puzzle: are we actually building slower?

When the structure finally rose toward the clouds, modern wind precautions created delays. High above the bay, powerful winds can stop work in its tracks. Wilson reported that fierce gusts forced repeated crane shutdowns, costing what he estimated as "a month to a month and a half" on the higher exterior wall alone. Still, once the tower climbed out of the mud and into the sky, the mechanical systems packed inside made it an entirely different kind of machine.

Walk into an early skyscraper, and you were inside a hollow shell with steam pipes and windows you could open. As engineering specialists point out, tall buildings across North America through 1930 relied on simple steam heat, completely lacking modern ventilation. Today, a high-rise functions more like a living organism. Engineers must counter fierce drafts from the stack effect, where warm air rushes up a tall core like a chimney, along with voltage drop along vertical wiring and crushing water weight. To prevent that downward force from blowing out lower valves, builders use hydrostatic pressure zoning—splitting plumbing lines into separate vertical blocks roughly every 10 floors, or about 40 meters.

Contemporary life-safety mandates demand an enormous amount of physical labor on site. State rules in California compel crews to install seismic bracing on all elevated ducting, piping, and mechanical gear. The ventilation network also doubles as a smoke-control system engineered to pressurize exit stairs and isolate noxious fumes during a fire. Fire marshals must witness that machinery tested in person. If an acceptance test fails, engineering logs show occupancy can slip by 4 to 8 weeks.

Sophisticated software was supposed to speed things up, but pixels cannot assemble themselves. Trade analysts point out that digital clash detection lets teams spot when plumbing runs smack into a steel beam on a monitor rather than during construction. Yet as Federal Reserve Governor Michael S. Barr observed, building information modeling and computer-aided design mainly streamline project coordination. They do nothing to eliminate the grueling physical labor of threading conduit, ducting, and steel through tight spaces. Fitting all those mechanical systems into the clouds is only half the challenge. Securing permission to break ground has turned into a multi-year ordeal in its own right.

Long before excavators touch soil, paperwork eats up years. Economists call this the preconstruction planning lag—the gap between filing for zoning and permits and the day shovels hit the ground. When Federal Reserve researchers studied commercial projects, that cost-weighted lag averaged 28.6 months. Moving from the fastest approval areas to the slowest added seven months to planning, while Goldman Sachs estimated postwar rules placed a major drag on construction productivity.

Much of California's planning friction centers on CEQA, the landmark state environmental law that gives opponents legal standing to challenge urban infill projects. State legislative analysts found environmental appeals stretched timelines in major California cities by two and a half years on average, with 80 percent of those court challenges hitting infill development. As permitting specialist Amy Ablakat observed, local review boards face a compounding bind: “Multiple departments and agencies are often involved, environmental reviews can be extensive, policies and regulations change frequently, and many departments and agencies are understaffed and overloaded with projects.”

Inside city hall, paperwork splinters across separate departments. Securing a residential building permit in San Francisco required a median 280 days, compared to 91 days in Austin. Critics point to administrative paralysis when sixteen percent of applications go through repeated rounds of agency revisions. Permitting consultant Tanner Blackman spent years shepherding a single development through entitlement, even taking a fight over narrow dirt-hauling hours to the city council to avoid weeks of extra delay. As Blackman confessed, “In a perfect world, my job would not need to exist.”

Yet it is too simple to dismiss all oversight as pure waste, because structural peer reviews and fire-safety tests protect lives. That creates a real uncertainty in the historical record. Even though independent engineering reviews and smoke-control tests plainly add weeks, scholars have never separated the timeline impact of worker protections from the drag of municipal bureaucracy and mechanical equipment. It leaves an unresolved question: does this prolonged build schedule merely stem from city hall friction, or does it reflect an economy-wide slump in construction productivity?

Step outside the planning office, and economists see signs of a much wider slump. In a paper by Austan Goolsbee and Chad Syverson, value added per American builder sank roughly 40 percent after 1970. By 2020, factory productivity had grown more than 9-fold over that same stretch. Actual physical output per worker slid down toward late-1930s levels, with construction productivity shrinking across a majority of surveyed nations.

That drag leaves a deep mark on the rest of the economy. Chen Yeh, an economist at the Richmond Fed, showed that economy-wide output per worker more than tripled after 1948, yet construction barely budged. Work cited by Yeh estimated that this one industry accounts for about one-third of the country's entire postwar drop in trend growth.

Looking at what these structures actually do reframes whether we are actually building slower. Standard economic tallies treat an empty floor plate from 1931 the same as one inside Salesforce Tower, overlooking that modern towers run like self-contained green power plants. The building earned LEED Platinum, the highest environmental standard, by cutting energy use per square foot by 50 percent compared to the median office, while filtering nearly 8 million gallons of blackwater annually.

A lot of that statistical collapse also traces back to inflation math. The construction value-added deflator—the formula economists use to adjust for rising costs—climbed at 5.47 percent annually compared to 3.37 percent for broader GDP, mechanically making building output appear to shrink. Researchers at the Bureau of Labor Statistics using more dependable price indexes saw no sustained drop in productivity. Accounting for unmeasured building quality led Goldman Sachs analyst Elsie Peng to calculate that the annual decline narrowed from 1 percent down to just 0.3 percent. Yet even after crediting those upgrades, construction still trailed every other major industry, showing how broad economic forces collided with physical friction on the site.

Financial turmoil delayed the tower for years. San Francisco adopted its Transbay redevelopment plan in 2005, but the 2008 crash hit before documents were signed. Hines senior managing director Paul Paradis recalled the price looked too high and the schedule too short, leaving the project virtually shelved. Following 2012 approvals and the 2013 land transfer, Paradis reflected that development was tougher than construction—except for the foundation hurdles and that alarming void in the bedrock socket.

That explains why digital tools have not revolutionized job sites. As Barr noted, physical building remains labor-heavy. Presenting the case for stagnation, Elsie Peng at Goldman Sachs observed that machinery like bulldozers, concrete mixers, and cranes has evolved only incrementally since the 1950s, equipment's share of costs has stalled at 10 percent, and prefabrication dropped to 5 percent of homebuilding. Measuring modern speed against 1930 compares apples to spaceships: earlier crews stacked a hollow shell with steam pipes over shallow rock rather than building a self-contained vertical city.

Validation arrived days before the tower opened. In January 2018, a magnitude 4.4 earthquake shook nearby Berkeley. Inside the skyscraper, an accelerometric array recorded small drift ratios and zero damage.

Modern builders have not lost their speed; our standards simply demand an entirely different structure. The Empire State Building was an interwar sprint anchored into Manhattan rock with basic radiators. Today's high-rise is an energy-saving sanctuary engineered to flex through earthquakes and protect thousands of lives. Taking five years is not a technological failure. It is the price of building to last.

Spend some time with construction productivity and the Empire State Building. And follow the economy; we have more stories there, so you'll be set when the next skyline debate opens up. Thanks for listening to Angle. Until next time.