(first posted 8/19/2011) We often take General Motors to task in this space for its failure to learn from its mistakes. But today’s tale involves the opposite problem for GM – Failure to learn from success. After a quarter century of building troublesome (Vega), sometimes dangerous (Corvair), and flaccid (Cavalier) small cars, by the autumn of 1983, GM was finally ready for a new approach to building cars that could compete with the finely crafted entry level products that were coming from Japan.
The car that emerged from the company’s joint venture with Toyota could have been the fork in the road that led GM back to the forefront of innovation and design excellence. Instead, internal company politics and poorly understood market dynamics made the Nova an historical curiosity that failed in its primary mission and left no lasting legacy with the parent company that desperately needed it.
The very admission that the worlds largest automaker needed help to design and build a class-competitive compact spoke volumes about the internal rot at GM. By the early 80’s the company had become dangerously dependent on mid size rear wheel drive sedans and pickup trucks to generate the capital needed to stay competitive across its model lines, pay dividends to stockholders and make the investment needed to head off the resurgence of Ford and Chrysler. The latter had struck gold with the K-Car and its derivatives and the former was gaining traction with its “Quality Is Job One” campaign coupled with the introduction of the ground breaking new “aero” T-Bird .
We need to remember that the X-Car debacle was still fresh in the minds of the top brass at GM. Indeed,the quartet of troublesome compacts had arguably done more damage to the company’s reputation than the Corvair/ Vega disasters of the previous two decades. It just seemed like the worlds largest automaker couldn’t build a small car that was any good. Out in flyover country, keeping up with the Joneses didn’t seem so smart when the Joneses’ Citation had to be towed in for warranty work a couple of times a month.
The subcompact J-Car program was a similar disappointment for the management at GM. The cars themselves were light years ahead of the Vega /Astre / Monza / Chevette “bad old days” efforts of the 70’s, but were clearly not up to the high standards of fit, finish, and reliability of the concurrent Civic, Corolla or 210/ Sentra. GM was making progress but losing ground. What to do?
One person who thought he knew just what the company needed was fairly new in the corner suite, but his tenure at GM would mark the era when the company lost its old confidence that it had the talent, the capital and the know how to build class leading cars in every market that it chose to compete in. That man was Roger Smith. His initiatives would define the corporate drift at GM in the 80’s and the Nova would be the first project that bore his unique stamp.
It’s less remembered today that the Toyota/GM joint venture proposal that would build the Nova was viewed with fear and suspicion in Detroit in 1984. GM still had 43% or so of the U.S. market and Toyota owned Japan, so the announcement of a joint venture between the worldwide rivals was not met with universal acclaim. Ford and Chrysler made threatening noises about the antitrust implications of such a marriage and the smaller Japanese makes complained in vain to Tokyo about the un-level playing field being even more tilted against them. In the U.S., the justice department looked the other way in these years of benign neglect and laissez faire and Tokyo took the whole thing “under advisement”. Clearly, both companies had friends in high places.
The corporate form that would screw together the Nova would be a completely new car company with grouchy old workers and an fairly old factory. New United Motor Manufacturing Incorporated would be the empty vessel that GM hoped would turn out a world class product utilizing a veteran, though restive workforce. The company would forthwith be known as NUMMI and the factory that GM contributed as its share of the project would be its facility in Fremont, California.
Asked to choose the “least likely to succeed” plant to build the Nova , smart GM watchers would have chosen Fremont. The plant was old (it had built cars and pickup trucks since 1960) , inefficient, and had one of the most militant labor forces in the industry. And, it was mothballed. GM had shuttered the plant in 1982 after the workforce had turned in some of the worst quality scores ever seen on GM vehicles in the previous two years. Absenteeism was endemic. Drinking , drugs and on the job sex was rife. But the Toyota side of the venture thought that the problem was not the workers, but their management.
Neither company came into this marriage of convenience with totally pure motives. Toyota faced import restrictions on U.S. sales and needed to learn how to build cars here and deal with American workers. GM wanted a peek behind the curtain at Toyota’s secret of building small cars that ran well and made money. That only one partner would succeed was a story that took time to understand.
Roger Smith decreed that GM managers would use NUMMI as a classroom to learn the Toyota way of building small cars. They would then take this knowledge back to their home plants and divisions and GM would turn Toyota’s biggest guns against it and beat back the import threat once and for all. That was the theory, anyway. They would come to be called the “NUMMI Commandos”. In the event , the GM managers assigned to the venture would learn their lessons well, but time would prove that they were unable to teach them to a company that didn’t want to learn.
Smith (and to be fair,the rest of GM’s senior management ) truly believed that there was a valuable secret that the Japanese were hiding behind their inscrutable facade of company unity and shared purpose. Smith himself was the ultimate GM company bevel gear; he had spent virtually his entire working life in the financial bowels of the mother ship. He could have been by no stretch described as a “car guy”. He was a “numbers guy” and as such, he believed that if you stayed up late enough and studied hard enough, every problem could be solved by application of cold hard statistical control.
It never occurred to Smith that maybe the Japanese just hired better workers and listened to them when they had ideas or concerns. He would be aghast at the idea that a common line assembler could become more knowledgeable about a product than the engineer that designed it. He would be apoplectic if an assembler could stop a production line because of a defect. But that is just what Toyota taught its workers to do.
Thus Smith pushed the NUMMI Nova through the sclerotic bureaucracy at GM. The timetable for the car was ambitious: Both companies wanted a salable product by the fall of 1984. This meant that the car that emerged could not possibly be a clean sheet design. Almost by definition, the car would be a badge job, a car that could ramp up production quickly with existing technology. Because the whole point was to build a better small car than GM could build on its own, the resulting product would be more Toyota than GM , but plugged into the vast Chevrolet dealer network. Chevy had clout; it had long been one of the largest buyers of network television ad time. Suppliers had relationships with Chevy that had endured for decades. The make had dealers in virtually every town in America, thus parts and service would not be an issue. The Nova would be the “no excuses” small car that customers demanded.
Somehow, it all worked. The first cars that came down the Fremont assembly line in December 1984 (as ’85 models) were among the highest quality cars that GM had ever built. They were statistically indistinguishable to the Corolla built in Japan that the car was based on. The cars’ fit and finish were impeccable. The interior didn’t squeak or break. The engine bay was a model of logical, tidy, well machined poetry. Wires were well routed inside neat, dedicated looms , the battery was accessible and oil, belt and hose changes were a snap for weekend tinkerers. It looked, for a moment anyway, like GM had found the way to reinvent itself and ensure its prosperity for decades to come.
But the reality was different than what customers and stockholders saw. First, the Nova was nobody’s baby at GM. It was a “top down” program that created resentment in just about every corner of the company. Design studios resented the obvious badge engineering of a Toyota with a bowtie on the grille. Engineering thought that the company already produced the worlds best engines and transmissions (although the X-car program should have put paid to that idea) and finally, there were the dealers.
When a new Nova was sold, it was so reliable that it was rarely seen again in the shop for repairs. Warranty billing was almost non existent. Also, there was very little trade -in potential because customers ran them till the wheels fell off or had willing buyers through word of mouth, thus depriving dealers of a second bite of the owners’…apple. And since GM and Toyota split the profits from each car sold, even accounting questioned the wisdom of building a car with razor thin profit margins.
The only constituency that loved the car were the wildly satisfied owners. The solid construction, bulletproof 1.6 litre Toyota engine and indestructible 5 speed transmission (or 3 and 4 speed automatics) made believers out of buyers that had not visited a GM showroom in years (like the author) . The car was the best part of the NUMMI experiment and the one longest remembered.
The legacy of the Nova is rather mixed. The car itself stayed in production until 1988 , when GM decided that forthwith the NUMMI products would sell better under a different brand and thus the GEO nameplate was born. The Nova would give way to the Prizm later that year.
Roger Smith would retire at GM in 1990 and would be noted as one of the greatest destroyers of corporate value that ever lived. Before he gave up control , Smith placed a few nickel bets on GM’s foreign partners Suzuki and Isuzu which yielded the Spectrum and Sprint captive imports in the mid/late 80’s (below) and a huge , bet -the- farm wager on Saturn, which would never return the capital invested in its creation. Today, Smith is seen as a fossil from another age, an age when GM didn’t have to be concerned with market dynamics because it pretty much was the market.
As for the NUMMI Commandos, they would become prophets without honor within GM. The mistake that the company made was to try to take a shortcut to quality by stealing “secrets” that didn’t really exist. The core group of quality evangelists were dispatched throughout GM where jealous division managers made sure that the lessons they learned would stay locked away without serious attempts at implementation. The Toyota system could not be grafted onto the GM way of doing things unless everybody bought in. And in a company the size of General Motors, buy in wasn’t happening.
NUMMI soldiered on until 2010,when the last Corollas and Pontiac Vibes rolled off the line. Till its last days, Fremont built a world class car that stood comparison with any model in production anywhere. The plant is now a facility owned by Tesla Motors.
Toyota, meanwhile, learned that it could build cars in America that were as good as any built at Toyota City. It later opened a plant in Georgetown, Kentucky to build Camry’s. It builds them there to this day. Toyota is now bigger than erstwhile partner GM. As for GM itself, the lesson learned from the NUMMI experience is that you can’t teach people that have no desire to learn. GM eventually went bankrupt. Shareholders were wiped out and the U.S. Government eventually took control of the company.
My young family’s experience with the NUMMI Nova was a happy one. We bought a 5 speed 4 door saloon in March 1986 and drove the car for 4 years and put over 125,000 miles on the clock. We paid just under $7300 and the car never went in the shop once. The only repair it ever needed was a muffler at 70,000 miles. I gave the car to my brother for a ham sandwich and he put almost another 100,000 miles on it and sold it for $625 in 1994.