A quiet and unassuming man, John Houlihan was ushered into our business by the charismatic Dave Chambers, and together they built the foundation for today’s labor management practices. It helped that Chambers’ plant, Imperial Components, was close to the site of the Hawthorne Study, where one of the most intensive worker evaluations was conducted, and that Chicago was one of the hubs of industrial engineering practice. It also helped that Chambers had a burning desire to increase his plant capacity to expand his share of the booming demand for trusses in the top housing market in the country in the 1970s.
Chambers had done everything he could to boost production, in both capital investment and personnel. He replaced the wood tables in the plant with one of the earliest Sanford gantries, and he brought in the first semi-automated DePauw component saw. He promoted his night shift leader, Don Hershey, to run both shifts. As a final step, he brought in John Houlihan, an experienced industrial engineer who had earned his stripes in much heavier industrial venues. And from then on, labor management in the truss business would never be the same.
Houlihan apparently realized that Chambers’ business was so tightly tied to construction that he would need to apply a recent adaptation of the Scientific Management Theory called Short Interval Scheduling (SIS). Then he spent several weeks on the plant floor observing Imperial’s operations to develop reasonable expectancies (REs) for each task. For example, in truss assembly, he based REs on span and truss type, during a time when trusses were mainly common, scissors, mono, or flat. REs were then aggregated by job and recorded on a two-week Forecast and Master Schedule. While this was a highly useful summary of what the plant was expected to produce, its implementation was the key to its effectiveness.
Houlihan then applied the key finding of the Hawthorne study, that workers performed better when they knew they were being observed, by requiring hourly pickups at each work center. If the work center was behind schedule during a complete shift, a schedule miss was reported to management, along with the reason. At the end of a two-week schedule, REs could be adjusted to reflect actual performance, or other actions could be triggered to remedy the schedule miss.
The most important part of Houlihan’s onsite work, and where he spent most of his effort, was in the establishment of REs, which depended on plant layout, equipment, worker abilities, and other factors. After Houlihan’s work at Imperial’s very compact and enclosed plant in Chicago, he developed REs for Chambers’ open-air, TruTrus roof truss line in Phoenix, which ended up being markedly different, despite having the same equipment.
The most valuable product of Houlihan’s work was the Forecast and Master Schedule, because of its built-in feedback mechanism. The plant personnel, production management, and sales staff knew what to expect, and they could readily determine if there were any openings or over-bookings in the schedule. This feedback extended even to Houlihan’s oversight, as he received a faxed copy of the completed schedules at least once a month, and he would determine if return visits or other follow-up actions were necessary.
While the fundamental values embodied in Houlihan’s system were and are sound, some aspects of SIS were modified after Houlihan completed his work, likely at the behest of Don Hershey. Houlihan measured work and performance in terms of scheduled units, with 100 units per hour, apparently so that workers would be less aware of their production goals and would theoretically work at a steady pace. This was changed to man-minutes so that measurement and communication was improved at all levels. And finally, as trusses became increasingly complex, time studies were redone to determine REs as a function of pieces and joints rather than truss types and spans.
Houlihan’s methodology spread slowly at first, mainly through Dave Chambers’ leadership in TPI’s Component Manufacturers Council (CMC), precursor of WTCA. The early head of the Council, Charlie Barns of Barns Lumber in Dallas, and his partner, Dick Rotto at Trussway in Houston, were two of the first SIS converts, each with his own unique implementation and equipment. Yet, having John Houlihan install SIS at other plants was a hard sell for most manufacturers, since his services could cost as much as a component saw in the 1980s. As a result, only about 35 of 1500 component plants contracted with Houlihan directly for his services.
In an indirect way, though, Houlihan methodology gradually reached hundreds of component manufacturers, via their truss engineering software, especially as in-house computers became commonplace. Alpine was one of the first to incorporate Houlihan-like values, greatly broadening the use of these factors by the early 1990s. At TruTrus, for example, the expected man-minutes for each truss were extracted from the Alpine output and were annotated on the shop drawings. When a given job was completed, the actual man-minutes consumed were handwritten next to the expected. At the end of a shift, those figures were accumulated, and the all-important expected-over-actual ratio was determined. According to Houlihan, plants should operate at a 90% level or higher, a crucial performance metric often given high visibility within the walls of a business. Embodied in that numerical grade is the true genius of Houlihan’s work, a single metric that everyone could understand and be proud of or be motivated to improve.
John Houlihan’s work was crucial to the development of sound labor practices in the truss industry, and still is widely used, though in modified forms. That Houlihan visited a relative handful of plants over a relatively short time span, more than 30 years ago, attests to the lasting value of his work.