The Report—and Some Reactions
James E. Harbour is a proverbial industry veteran. When he retired from Chrysler as director of Manufacturing Engineering in 1980, it was after 28 years in the industry, most of them with Chrysler, which he joined in 1964, and some with Ford. His background, knowledge and interest in issues related to manufacturing and productivity were parlayed into a Troy, Michigan-based consultancy, Harbour & Associates, which he founded after he left Chrysler. He is now chairman of the firm. Ronald E. Harbour, his son, is president of the firm.
What the name "Harbour" undoubtedly is most well known for in the executive suites of auto manufacturers—the world over, but especially in Detroit—is for an annual productivity study of the North American auto industry that's titled, quite simply, "The Harbour Report." It details which companies are doing well in terms of their vehicle assembly and engine and transmission manufacturing in North America; it goes right to the plant level, indicating which are stellar—and not-so-stellar—performers vis-à-vis their peers in the industry. (A European study is done, but it is a private, not public document.)
According to Jim Harbour, the methodology is straight-forward (which isn't surprising to anyone who has ever talked with him—he invariably cuts to the chase with little meandering). He and his associates hold a meeting that's attended by representatives from the companies that choose to participate in this benchmarking exercise. The information sought by the Harbour group is explained. The folks from the car companies then agree to what will be provided. The companies send the information to the office in Troy. "They provide the data. We analyze it," Harbour explains. Of the approximately 160 assembly plants in North America, about 40 are visited by Harbour personnel between November and February to verify the data.
Does anyone at the car companies ever get upset with the resultant report? "No, not with what we present," Harbour answers, adding, "But sometimes they question whether people at other companies have accurately reported what they are doing."
The participation by the auto companies occurs, Harbour explains, "Because we are an independent source. They do not pay us to do the study—they don't fund us with one dime.
"They are interested in the study because it allows them to know what the benchmarks are and what they have to do to get as competitive as the benchmarks."
Some people, apparently, are miffed by the findings. A case in point that Ron Harbour makes is the United Auto Workers, which went on strike at the GM Oklahoma City plant for 53 days this past spring. One reason for the strike was a demand for additional workers. The new Chevy Malibu and Olds Cutlass are produced in that plant, and while there are, Harbour admits, fewer workers building those vehicles than there were building the previous models, according to the findings, the GM Oklahoma City plant requires 3.41 workers per vehicle, which is 43% more than the benchmark, the 2.38 workers per vehicle required to build the Nissan Altima at the Nissan plant in Smyrna, TN (the Nissan plant is the overall benchmark for the study). In other words, the trade union wants more people to work in the plant. GM, for reasons of competitiveness, thinks there should be fewer people, and the Harbour Report indicates that the corporation is right.
"GM is doing a good job in design for manufacturing," Ron Harbour says. "They are designing easier-to-build vehicles." In addition to the Malibu and Cutlass, he cites the Pontiac Grand Prix and Olds Intrigue as examples of new cars that have been designed for manufacture. Jim Harbour elaborates: "What we see GM doing today is that on every new product they have a team of people work on it so that it is designed for ease of assembly. This is something that Ford has done for a long time." One of the consequences of better engineering up front is that it typically means fewer people are required to build vehicles, which leads to problems—such as those experienced at the GM Oklahoma City plant—with the unions (see "The Indirect Issue.")
The Major Findings
The big kicker: Nissan Motor Manufacturing in Smyrna, TN, is, for the third year in a row (according to the "Harbour Report"), the most productive assembly plant in North America. At the plant the subcompact Sentra (and derivative 200SX), the compact Altima (which received the best rating in its class in the J.D. Power and Associates 1996 Initial Quality Study), and the small size 4x2 and 4x4 pickups are produced. An average 2.23 workers are needed to build one vehicle. One thing that's interesting to note is that even though it is the best, the plant actually slipped by 6.6% compared to last year, when its number was 2.09 workers per vehicle.
Second place, again, went to Honda of America Manufacturing, with 2.51 workers per vehicle. The best of the traditional domestics is Ford, which requires 3.09 workers per vehicle. GM requires more workers per vehicle than any of the other major builders—3.47—but compared with the others, its rate of improvement is best: 4.8% better than last year. It has made major strides since 1992, when it needed 4.47 workers per vehicle.
Diving down to the vehicle segments, the plants, vehicle benchmarks and their workers per vehicle are:
- Subcompact: Nissan Smyrna; Sentra, 200SX; 2.22 workers per vehicle
- Compact: Nissan Smyrna; Altima; 2.38 workers per vehicle (estimated)
- Midsize: Toyota Georgetown #1; Camry; 2.50 workers per vehicle
- Large: Ford St. Thomas; Crown Victoria, Grand Marquis; 3.01 workers per vehicle
- Luxury: Ford Wixom; Lincolns; 4.17 workers per vehicle
- Sports car: Ford Dearborn; Mustang; 3.14 workers per vehicle
- Midsize sport utility vehicle: Ford Louisville; Explorer, Mountaineer, Ranger; 2.66 workers per vehicle
- Full-size sport utility vehicle: Ford Michigan Truck; Expedition, F-Series; 3.52 workers per vehicle
- Small pickup: Nissan Smyrna; pickup; 2.25 workers per vehicle
- Full-size pickup: Ford Kentucky; F-Series (previous generation); 2.83 workers per vehicle
- Minivan: Ford Oakville; Windstar; 2.85 workers per vehicle
- Large van: Ford Lorain; Econoline; 3.28 workers per vehicle.
It's Not All Factory Floor
One thing to realize about the numbers is that they are not isolated indices. The productivity ratings should not be considered separately as they are part of a larger infrastructure. In a very real sense, production is penalized by what can be considered failures in other functions, such as in design, engineering and marketing. Simply put: an ugly vehicle that's ill-conceived and/or an egregious sticker price is not likely to speed out of showrooms at more than a glacial rate. Consequently, if an assembly plant is capacitized to produce x vehicles, both in terms of manpower and equipment, and it is producing x - y vehicles, and must retain staffing levels due to a negotiated contract, then even if everything else is done to the extent that the late Drs. Ohno and Shingo would nod approvingly, the plant in question is not going to score well at all.
Instances where poor sales have resulted in poor numbers for facilities are the Auto Alliance International Plant (Flat Rock, MI), which is where the Mazda 626 and MX-6 and Probe were produced, and the CAMI Automotive (Ingersoll, Ontario) plant, which builds the Metro and Swift subcompacts and Sidekick and Tracker small sport utes. These plants are running at half volume. So the Auto Alliance average is 3.11 workers per vehicle (which is better than Chrysler and GM) and CAMI is 3.97 workers per vehicle.
From the standpoint of what actually makes the major difference on the factory floor itself it is, Ron Harbour suggests, a lot of what is often considered to be the so-called blocking and tackling—but realize that while these may be the fundamentals of good manufacturing, they are fundamentals as being practiced by people with world-class skills. Excellence in execution is essential. "A lot of it is basic industrial engineering," he says. "It's getting the most from each worker." To do that, it is important to design and engineer the vehicle correctly so that it is comparatively easy to produce, and have the facility setup so that the operator spends most of his or her time actually putting parts on the vehicle—adding value—rather than searching around for components to put on. "It's a lot of simple things to make things better," Harbour says. He says that technological solutions tend not to make a "huge impact" in many cases—he points out, for example, that in engine assembly plant piston stuffing automation is being removed because the downtime being caused is more hurtful than the help that the equipment might provide—but that technology is beneficial in situations where it can contribute to consistency in build, such as in the paint shop.
He cites the "limousines" that are employed in the Nissan Smyrna plant. These are simple platforms that move along the assembly line with the vehicle. The operator and the necessary parts ride on the limo. That way, the tasks at hand are truly at hand—as are the tools and the parts needed to do the work. He says that many plants have their own versions of limousines, but at Nissan, they have done a great job with them.
Another important factor: Line balancing, making sure that the work that needs to be done can be done by all of those who are responsible to do it.
The Indirect Issue
Although direct labor contributes about 15% to the price of a vehicle, it isn't simply the direct labor that the "Harbour Report" assesses in its calculation of the labor cost per vehicle. Indirect labor—including skilled trades—are also included in the figuring, which brings the labor contribution to cost to just above 20%, to about 22 to 23%.
Jim Harbour points out that when it comes to direct labor, the Big Three are becoming increasingly competitive with the Japanese-owned plants. But he maintains that when it comes to the indirect labor—everything from tool and die makers to sweepers, from QC personnel to material handlers—the Big Three plants have too much of it. It's not that there aren't these functions performed in Nissan, Toyota or Honda plants. But the difference is, the Harbours suggest, that the indirect people are (1) fewer in number at the Japanese-owned plants and (2) the people working in those roles are much more cross-functional in what they do than is the case in most U.S. facilities.
"The Big Three engineers know how to do a line and how to process a car for ease of assembly, but the indirect labor is providing to be more difficult," Jim Harbour says.
Commonization Is Coming
Jim Harbour maintains that design for assembly (DFA) can always help a company improve its productivity rating. The competitors keep getting tougher, but the leverage that DFA provides is always beneficial.
And he points out that simplification and commonization, two related aspects of DFA thinking, will have significant impacts on the shape of the overall industry—OEMs and suppliers alike. He cites, for example, Ford's stated intention to go from 32 vehicle platforms to 16. Think of what this means to stamping plants in terms of just having half the number of different floor pans. (Harbour notes that with more stampers running full body side panels today, there's a lot less assembly and welding occurring now). The 16 platforms mean a relational reduction in what's needed with regard to such things as suspensions, brakes, and powertrains—just to name a few items—which will have an immediate effect on suppliers of these things. The implicit message: Those plants that are the best in their category are likely to be the ones that will keep the business. As for the others, well...(see "A Warning.")
Harbour says that presently there are some 600 different seating systems available in the U.S. auto market. He suggests that this number could go to 12. This doesn't mean that customers will go into a given car dealership and look into, say, 6 cars and see half of all the available seating options. The exteriors will be different, but as will be the case with the reduced number of vehicle platforms, beneath the skin there will be commonality. So what he is suggesting is a case where there will be a dozen different mechanisms. Consider the ramifications of this on companies including Lear and Johnson Controls.
Commonization will result in a different landscape of builders and suppliers.
The Cost Differences
Nissan is the benchmark. Based on assembly, stamping and powertrain requirements, the "Report" reckons that the labor hours per vehicle (LHPV) for Nissan in Smyrna is 28.32. Then, assuming $43.00 per labor hour (which Ron Harbour says is high for Nissan, Honda and Toyota and therefore gives the Big Three an artificial advantage in the calculations), the labor cost per vehicle (LCPV) at Nissan is $1,218. Given these numbers, things really get scary for the other manufacturers—especially the Big Three—when annual production volumes are factored into the calculations.
Toyota's numbers, for example, are next-best: 29.54 LHPV; $1,270 LCPV. This means that it costs Toyota $52.00 more to produce a vehicle in North America than it does Nissan. By multiplying its 1996 production volume (483,000 units) by $52.00, the "Annual Cost Penalty to the Benchmark" (ACPB) is $25-million. And there is even a calculation of the "Excess Workers to the Benchmark." In Toyota's case it is 313 workers.
The Harbours stress that companies have different strategies in terms of what kinds of vehicles they make and have different sourcing strategies that can ameliorate some of the differences in the overall profitability picture of which labor costs are just a piece. This warning is necessary because the figures for the Big Three—especially GM—are astronomical in a place where smaller is better.
- Chrysler: 40.53 LHPV; $1,743 LCPV. It costs Chrysler $525 more to build a vehicle in North America than it does Nissan, which means based on a production volume of 2,767,000, the ACPB is $1,453,000,000 and it has 17,968 excess workers compared with the benchmark.
- Ford: 37.59 LHPV; $1,616 LCPV. This provides a lower cost penalty than Chrysler's: $398. Given its 4,271,000,000 annual volume, it has an annual cost penalty to the benchmark of $1,702,000,000 and 21,062 excess workers to the benchmark.
- GM: 44.59 labor hours per vehicle. $1,917 labor cost per vehicle. $700 more for GM to produce a unit compared with Nissan North America. Multiplied by 5,039,000,000 units, this means an annual cost penalty of $3,525,000,000 and—hang on—43,610 excess workers to the benchmark. Still, GM is doing a heck of a lot better today than it was a few years ago, and within a few years it will be doing even better.
Ron Harbour puts it plainly: "Those plants that aren't competitive are going to die."
The explanation: "Capacity is going up in the U.S. Do you see sales volume going up? No. A plant that doesn't have a competitive product, a competitive workforce, and a competitive level of quality isn't going to survive."
Clearly, the burden is with the folks at many of the Big Three plants. The question posed to Ron Harbour, then, is whether they can do what's needed to survive against what's sometimes perceived to be "foreign" competition. He points out that Honda, Toyota and Nissan facilities in North America are essentially managed and staffed by North Americans, just like the Big Three plants, so there's no big difference there. The difference is the way they do what they do. The amount of value-adding activities makes all the difference.
The choice seems to be simple: Get more competitive in product, productivity and quality. Otherwise, Ron Harbour warns, "They won't have a decision. They won't be around."
Interested in Getting the Full Report?
To obtain a copy of the latest "Harbour Report," which is available for $395.00 per copy, you can write, call, fax, or e-mail:
Harbour & Associates, Inc.
4967 Crooks Road, Suite 130
Troy, MI 48098-5812
Telephone: (248) 641-2854 in Southeast Michigan
(800) 208-1353 toll free
Fax: (248) 641-0886 E-mail: firstname.lastname@example.org