(FORTUNE Magazine) – SOMETIME in the next few weeks, the 18 members of General Motors' board of directors will elect a chairman and chief executive to succeed Roger Smith when he retires August 1. The new man will face one of the most formidable challenges in corporate America: revitalizing the core business of the world's largest industrial corporation. GM concedes that its vast North American auto operations have barely broken even for three years, with profits on trucks offset by big losses on cars. The company must be trimmed down to run economically, reorganized to function effectively, and redirected to attract new customers. After nine years as CEO, Smith, 64, still has a firm grip on the wheel. In recent weeks he has reminded audiences that GM remains the biggest automaker on the globe. He says the company's competitiveness is at an all-time high and predicts that its U.S. market share will reach 37% this year for the first time since 1986.

GM does have remarkable strengths. It sells six brands of cars through five divisions; a sixth division, Saturn, will begin selling GM's new small cars in the fall. Its plants use the most modern machinery of the Big Three, and the company has launched more redesigned models since 1987 than Ford and Chrysler combined. Its technical and research skills and its cadre of experienced executives are unmatched anywhere in the world. But GM's North American auto operations, which include Canada and Mexico, misfired in the 1980s (see charts). Ford and Chrysler lately have slipped too. At Ford, new CEO Red Poling is struggling to discipline new-model development programs that are behind schedule and over budget. And Chrysler, facing fresh competition for its best-selling minivans and Jeep Cherokees, ultimately may be forced into a merger with a Japanese or European partner. GM, however, is the biggest of the Big Three, and its slide has been deepest and longest. The company sells nearly one-third fewer cars now than it did in 1979 -- and analysts say that in North America it lost close to $1 billion on them last year alone. GM is Detroit's high-cost producer; model for model, on average, its cars each cost $200 to $300 more to build than Ford's and $750 more than Japanese models made in the U.S., according to Ronald Glantz, a well-regarded auto analyst at Dean Witter in San Francisco. Nearly all the market share gains made by the Japanese during the last decade have come at GM's expense. Ward's Automotive Reports, which publishes the industry's most relied upon statistics, says GM lost 11.6 points -- from 46.3 in 1979 to 34.7 at the end of last year -- while the Japanese gained 10.4 points. Ford picked up nearly two points, to 22, and Chrysler lost almost three, to 10.3. (By its own reckoning, GM has 35.1% of U.S. car sales. That's because in calculating the total size of the market, it excludes 115,000 Japanese vans and sport utility vehicles that Ward's counts as cars.)

GM's new CEO will have to act fast to slow the damage and regain momentum. It is acknowledged within the company that unused factory capacity amounting to one million cars must be written off -- and the plants closed -- at a one- time cost of more than $1 billion. GM's huge white-collar bureaucracy must ^ be substantially reduced to get costs down and efficiency up. And negotiations will have already begun with the United Auto Workers, whose three-year contract expires September 14. As usual, the bargaining promises to be difficult. Over the past three years, the UAW says, the average GM hourly employee has received a total of $304 in profit sharing, vs. $7,900 for Ford workers. Among other grievances, the auto workers contend that GM has broken its word by indefinitely laying off workers at three plants without putting the workers into a job bank, where they receive full pay and benefits. GM says the issue is in arbitration. No one -- not even its competitors -- can take any pleasure in the plight of General Motors. It is a global institution -- a huge employer and buyer from countless suppliers, an enormous force in the world economy and in the many communities in which it operates. What's bad for General Motors is bad for the U.S. If it can return its North American car operations to competitiveness, GM will serve as an inspiration and example to other U.S. companies. THE QUALITY of GM cars has already gotten dramatically better. That is evident from data compiled by James E. Harbour, a Michigan manufacturing consultant for automakers and suppliers worldwide. In 1980, says Harbour, GM had an average of 740 defects per 100 cars, second worst in the industry after Chrysler with 810 per 100. (Defects are problems that customers notice, such as poorly fitted body panels, squeaks and rattles, and broken parts.) By 1989 the number had fallen to 168 per 100, though GM still lags behind Japanese manufacturers, which have an average of 121 defects per 100 cars, and Ford, which has 149. Over the next two or three years, GM will also break new ground in styling. On a recent tour of the company's design studio, Chuck Jordan, the Hollywood- handsome vice president of design, whisked the dust covers off nearly a dozen new cars parked in the hallways. Jordan's team has created an imposing new look for the 1992 Cadillac Eldorado and Seville that is as grand and distinctive as any of GM's luxury classics of the 1950s. And it redesigned the slow-selling Chevy Lumina and Pontiac Grand Prix for 1993 and 1993 1/2 with a futuristic cab-forward look that some analysts believe could catapult GM to the leading edge of automotive design. Except for its flagship U.S. car business, GM is thriving. Trucks earned more than $1 billion last year. GMAC, the finance group, contributed another $ $1.1 billion, and Ross Perot's old Electronic Data Systems kicked in a record $423 million. GM's overseas car operation is healthiest of all; it made $2.6 billion last year, mostly in Europe. Altogether, earnings were $4.2 billion, giving GM its third most profitable year ever. As recently as 1986, GM Europe was a money loser. But after it went through a vigorous cost cutting and developed some snappy new models, the market turned hot and cash started gushing. Unfortunately its success cannot be easily exported. GM Europe sells 60% fewer cars than the North American operation and markets them under just two brands, Opel and Vauxhall. Because it is less vertically integrated, it can buy more parts at competitive prices from outside suppliers. Managers operate with more autonomy than their American peers. And competition isn't as stiff. The Japanese account for only about 11% of European car sales, vs. nearly 26% in the U.S. Fairly or not, Roger Smith has become a lightning rod for criticism about GM. Smith insists that his changes will ultimately pay off for the company. As he told FORTUNE a year ago, ''Please come back in five years and see what I did.'' But many GM executives want to move faster to shake up the car operations, and they are eagerly awaiting the arrival of the new CEO and his team. There is even some admiration for Ford Motor, where CEO Donald Petersen recently retired early, enabling his successor to forge ahead. Says a consultant who works closely with GM: ''I don't know how many GM vice presidents have told me they can't make big changes until Roger leaves.'' Some of the company's institutional investors are frustrated. In January, New York State controller Edward Regan, a trustee of pension funds holding over 5.1 million GM shares, wrote to the board of directors noting that GM's stock had markedly underperformed the S&P 500 over the past decade (GM's stock rose 69%, vs. 227% for the S&P 500). Concluded Regan: ''We are very desirous of better performance from one of our most major holdings.'' California controller Gray Davis, a board member of the state employees' pension fund who is also co-chairman of a group of about 60 large pension funds, urges the board to cast widely for Smith's successor. ''The company needs new leadership, and it's the outside directors' obligation to mount a worldwide search for it,'' says Davis. ''Tradition is not the answer; it's the problem.'' Members of the GM board have pledged not to talk about the CEO selection process, but there are no signs they plan to recruit an outsider. Anne Armstrong, the former ambassador to Britain who is a 13-year GM director and a member of the nominating committee, said in a brief telephone conversation: ''We have discussed the possibility of looking at outside candidates. I'm not going to go any further than that.'' SMITH'S all-but-certain successor appears to be President Robert C. Stempel, 56, who rose through the ranks from transmission design engineer. After 32 years at GM, the indefatigable Stempel knows the car business ''from the road to the roof,'' to use a favorite phrase of his. As the first career engineer to run the company, he would end GM's domination by its financial staff. Under Stempel the company could be expected to concentrate at first more on products than on short-term profits. That would be popular within GM and, surprisingly, on Wall Street, where analysts and large investors view Stempel much the same way they do Petersen -- a car enthusiast who rebuilt Ford's product line and guided it to record profits. Stempel is a reformer, not a revolutionary. During his 2 1/2 years as president, he has been the ideal second-in-command. In January he flew 1,500 miles from a dealers' convention in Las Vegas to Chicago, where he listened to Smith give a 30-minute speech, then returned to Las Vegas that afternoon. If Stempel has any radical plans for GM, he is keeping them private. Says he: ''We made our decision to go with a long-term view, and it's paying off overseas, in trucks, and in our acquisitions. We just have to stay the course in North America.'' Like Smith, Stempel is sensitive to criticism and is known to study problems at length before reaching a decision. Also like his boss, he is confident the company is on an upswing. He told FORTUNE that GM will operate at 100% capacity by 1992, vs. 76% in 1989 (Ford ran at 97% capacity last year), regain 40% of U.S. car sales by 1994, and make Saturn profitable by 1995. The son of a New Jersey banker, Stempel played football at Worcester Polytechnic Institute in Massachusetts and got an MBA at night from Michigan State. At GM he ''moved through the chairs'' with unusual speed, winning 11 promotions during one 18-year stretch. As head of Opel in Europe in the early 1980s, he began the consolidation of GM's car operations that finally paid off in record profits last year. No buttoned-down executive in either style or demeanor, Stempel favors black jodhpur boots and poker chip-size gold cuff links bearing his astrological sign: Cancer. SEVERAL OTHER GM executives are poised for promotion. Depending on how the new titles are parceled out, their elevation could signal a greater sharing of responsibility among the company's top officers than during Smith's term. One candidate is executive vice president William Hoglund, 55, a lanky, likable Princeton graduate. He helped rejuvenate Pontiac in the early 1980s and then Cadillac in the late 1980s after succeeding Stempel as head of the Buick- Oldsmobile-Cadillac group. Another man who seems likely to move up is executive vice president John F. ''Jack'' Smith, 52 (no relation to Roger Smith or to another executive vice president, Alan Smith). The quiet, efficient Smith returned to Detroit last year after a highly successful term running GM Europe, where he launched several popular new models. Smith and Hoglund both started in finance but have won strong credentials as operating executives. As described by the legendary Alfred P. Sloan, GM's mission is ''to cut metal and in so doing add value to it.'' Historically, economies of scale have paid off big in the auto business, and nobody is bigger than GM, with its 755,000 employees, more than 200 plants and other facilities worldwide, and annual sales of $127 billion. But as more competitors have sprung up and GM's market share has shrunk, its avoirdupois has become a drag instead of a lever. The most telling indication is contained in a profitability analysis of GM's North American car and truck operations given to FORTUNE by a former GM employee. According to him, the analysis was prepared by GM's corporate finance staff. A GM spokesman refused to confirm or deny the document's authenticity. Analyzing any company's internal financial reports is difficult without knowing the assumptions and accounting principles behind them. GM's methods are more complicated than most because the prices charged by the company's in- house suppliers can inflate the booked costs of the cars. Moreover, budgets can be used for internal strategic or economic purposes that distort the final results. But two knowledgeable security analysts who independently were given copies of the document say it is genuine and presents an accurate picture of GM's U.S. car business. And when asked about some of the forecasts, four high GM executives did not dispute the data. Dated December 5, 1988, the report projected sales for every car in the GM line for the 1989 model year, which ended September 30. It estimated GM's variable profit, based on direct manufacturing costs, and the company's operating profit -- what is left over after capital costs, overhead, merchandising expenses, and sales rebates are deducted. The analysis actually understated GM's losses, because 1989 production ran 10% below projections and cash rebates averaged about $1,000 per car, higher than anticipated. GM says it paid $5 billion in rebates in all of 1989, up from $4.2 billion in 1988. Manufacturing and security analysts with whom the document has been discussed say the projections highlight GM's excessive overhead and inefficient plants. The company's biggest losses are coming from its newest cars: the midsize coupes and sedans of the GM-10 line, such as the Buick Regal and Pontiac Grand Prix. They were introduced in the 1988 model year as a belated attempt to compete with the Ford Thunderbird and Taurus. The cars have been a major disappointment, and company officials concede that their four GM- 10 plants run at only 60% capacity. GM, for example, budgeted a loss of $1,697 on the Grand Prix. When startup costs for the Chevrolet Lumina were added in, the budgeted operating loss of the GM-10 line came to $2,117 per car in the 1989 model year. According to the document, the company figured to lose $1 billion on the GM-10s. Says a well-informed Detroit automotive consultant, who often develops similar cost analyses on his own: ''These numbers are not surprising. They provide a good representation of the financial challenge in front of GM.'' Unused plant capacity also made losers of the Buick Riviera, Oldsmobile Toronado, and Cadillac Eldorado. GM spends only $4,000 more building a $29,900 Eldorado than it does a sporty $14,100 Pontiac Firebird. But the trouble- plagued Hamtramck plant where the luxury cars are assembled runs on only one shift because of slow sales. So, according to the analysis, GM estimated it would lose $1,250 on each in the 1989 model year. Two low-volume, expensive two-seaters designed as ''image cars'' were particularly costly. Budgeted operating losses on the $52,500 Cadillac Allante approached $8,000 per car. And GM figured to lose about $17,000 on every $26,700 Buick Reatta. GM continues to make strong profits on light trucks and several older lines of more traditional big cars: the Oldsmobile 98, including the Touring Sedan, and similar Buicks and Cadillacs. They were budgeted to produce average operating profits of $3,685 each in model year 1989. The plants where they are built operate closer to capacity, and the tooling, which is amortized over four years, is paid for. The same is true for the gargantuan $29,300 Cadillac Brougham, built with vintage body-and-frame construction. The document said it would generate a $5,257 operating profit per car. Heavy rebates made profits scarce for both American and Japanese automakers in 1989, and that trend has carried over into 1990. The car companies have had a particularly tough time earning money on small cars because of intense price competition. Ford concedes that its subcompacts, such as the new Escort, lose money and that its compact Tempo and Topaz are only marginally profitable. GM has never acknowledged losing money on any model. But the internal document shows that it budgeted losses on all three of its small-car lines in model year 1989. The biggest losses -- $522 per car -- came on a line of older cars that includes such models as the Pontiac Grand Am and Buick Skylark. Even with its hefty margins on trucks, GM's profitability is declining more than its competitors' because it is losing so much on cars. A study by Harbour & Associates, James Harbour's consulting firm, shows that Ford's earnings on cars and trucks peaked at $706 a vehicle in 1987 before falling to $396 last year. GM's profitability per vehicle in North America reached a high of $588 in 1984 and has been headed downhill since, reaching $12 last year. Overseas the direction has been exactly the reverse; GM made $215 per vehicle in 1979 -- and $1,150 in 1989. At home GM's sinking car sales are eroding its base of loyal customers, with potentially costly long-term consequences. Satisfied owners are the best source of future sales and those who remain loyal for a lifetime generate a revenue stream for 40 years or more. Says Stempel: ''A guy who has bought anywhere from 15 to 25 vehicles over his lifetime is a great contributor to our bottom line. We lost some in the 1980s.'' Ordinarily GM could expect to replenish its owner base with aging baby- boomers. But according to a 1989 survey of 35,000 car owners across the U.S. by J.D. Power & Associates, the highly regarded California research firm, customers under 45 prefer Japanese cars almost 2 to 1 over GM cars. Ford fares somewhat better, Chrysler a little worse. Says Christopher Cedergren, senior analyst at J.D. Power: ''The baby-boom and baby-bust generations buy 55% of all new cars, and they rarely shop GM products.'' GM recently completed a confidential study called Adopt/Retain to determine whether the baby-boomers are likely to retain a preference for Japanese cars for years to come or whether, as they age, they will adopt their parents' habit of buying GM cars. Among the findings: Older buyers do like the company's big, roomy cars. FOR NOW, market analysts expect GM's sales to keep sliding. DRI/McGraw Hill sees the company's share sinking to a low of 31.5% in 1993 -- despite the addition of 240,000 Saturn sales -- and then recovering somewhat to 32.1% in 1994 as new designs catch on. Analysts at J.D. Power posit two scenarios. In the most optimistic, Ford goes into reverse, the Japanese stall, and GM accelerates to 40% of the market. Stempel says that's his goal, and a couple of hot models could put the company close. But instead of pushing all out for a blockbuster, GM has temporarily stopped work on some new models to save money in the first half of 1990. Power's Cedergren thinks GM's goal is ''far-fetched.'' A more likely outcome, he says, is that the Japanese keep growing, Ford holds on to its 22% share, and GM dips to 29% by 1995. In this scenario, after projecting sales for every model in GM's lineup, Power determined that Oldsmobile and Buick combined could be selling only 770,000 cars in 1995, down from more than 1.1 million last year. That would make it tough to support the current network of 6,000 Oldsmobile and Buick dealerships, though Stempel insists, ''I do not plan to get rid of any divisions, and, no, we do not plan to merge any of them.'' During the 1980s GM spent $77 billion on modernized plants, automated equipment, and new models. Yet a 1990 report by Harbour & Associates says GM reaped only minor improvements in productivity. The report was compiled after a year-long analysis of publicly available data from the Big Three and their foreign competitors; it was published as a sequel to a similar report by the firm in 1979. Since Jim Harbour did not have access to internal company reports, he could not account for such factors as overtime or absenteeism. While he cautions that his statistics are approximations, nobody, including GM, has disputed his overall analysis. Harbour says GM's car and truck assembly efficiency improved only 5% during the 1980s, vs. 31% at Ford. In 1989 it took 39 hours to build a vehicle at GM's plants, vs. 37 hours at Chrysler's and 26 at Ford's. Productivity at the , recently modernized factory in Doraville, Georgia, that assembles the midsize Oldsmobile Cutlass Supreme has fallen 6% since 1979, according to Harbour, partly because the plant is running well below capacity. Doraville requires twice as many workers per car as Ford's comparable midsize-car plants. GM invested several billion dollars during the 1980s on high-tech press lines that stamp out sheet metal body parts. Yet Harbour says GM's stamping productivity improved only 14%, while Ford, which spent a fraction as much, got a 45% boost. He says GM designers were unable to create new car models that capitalized on the presses' capabilities. GM responds that productivity at Ford plants appears higher because employees work more overtime. Since 1984, GM has committed more than $2 billion to train and educate its employees. Even so, it admits that it underestimated the skills needed to use the sophisticated new equipment. The inefficiencies that followed, combined with falling sales, hurt GM's productivity and its cost structure. Concludes Harbour: ''If GM were as efficient as Ford, it would require 60,000 fewer workers ((GM had 419,000 employees in the U.S. in 1988)), and it would gain $3.3 billion in annual pretax profit.'' THE MOST TROUBLESOME element in GM's overhead remains its army of 100,000 white-collar workers in North America. ''GMers look at the way they do business as a sacred cow and don't realize that they aren't functioning internally,'' says an outside management consultant who does work there. ''More time is spent on meetings than anyplace I've ever seen, but less is accomplished.'' GM has cut 40,000 people in four years from its white-collar work force. Yet, says David Cole, director of the Office for the Study of Automotive Transportation at the University of Michigan, ''GM still has overhead coming out its ears.'' Stempel acknowledges that more cuts are necessary but says they won't be drastic. Eugene Jennings, a professor emeritus at Michigan State's business school and a longtime student of GM, makes a downright draconian recommendation: 40,000 to 50,000 white-collar workers ought to be let go immediately, he says, and the organization compressed to remove excess layers of management and redundant functions. Marketing and engineering, for example, are staffed at the corporate, group, and division levels. More and more it is apparent that GM's 1984 reorganization injected Novocain into the system instead of adrenaline. By creating two new car groups to take over the functions of planning, engineering, and manufacturing, GM, in the words of a management consultant who works at the company, ''introduced a whole new layer of bureaucracy and further diffused responsibility. People make group decisions: If something fails, everyone takes the blame.'' The reorganization is likely to be GM's most ambitious attempt to grasp its historic Golden Fleece: leveraging its enormous scale. By centralizing engineering and manufacturing in two car groups, it believed that it could eliminate some unique models and build a few basic designs that each division could tailor for its own customers.

BUT GM was already huge. Before the reorganization, Chevrolet, Pontiac, and Oldsmobile each sold more cars than Chrysler. The reorganization not only made it less nimble but also went against the trend toward smaller-volume, custom- tailored models. Says a high-ranking executive at a Big Three competitor: ''We were all brought up to believe in economies of scale, but the days of pumping out 300,000 versions of a single car are gone.'' GM's product development system still fails to allow the separate car divisions to sufficiently differentiate themselves. All models are styled by GM's central design staff, a fiefdom of its own that reports up through two layers to the chairman of the board without going through any car executive. Division heads must negotiate with the design staff on styling for the cars they sell. ''Not all of them are all that clear on what they want,'' says design boss Chuck Jordan. Building cars by committee produces cars only a committee can love. Since the reorganization, GM has not had a single smash hit like the Ford Taurus or Honda Accord. Consumer Reports recently rated Chevrolet's newest car, the Lumina, beneath Ford's five-year-old Taurus in design and performance. And the American Automobile Association's automotive engineers were quoted in an AAA newsletter saying the Lumina ''easily improves on the nine-year-old Celebrity but still lacks features on competing cars. Lumina, although new, still feels old.'' That is not a surprise since, according to Jordan, the Lumina was originally designed as a 1982 model. It was delayed, incredibly, for seven years because of development problems and disarray from the reorganization. The new system presses tightest on GM's three divisions in the middle of the market that already had the greatest tendency to overlap: Buick, Olds, and Pontiac. Some 80% of their cars vary only slightly in price and engineering. The styling and image, though different, is not sufficiently distinctive to keep sales from falling. Buick's dropped 7% in 1989, Pontiac's 12%, and Oldsmobile has gone from selling 1.1 million cars in 1984 to just 600,000 in 1989. Ask an Olds executive what makes his cars different from Buicks or Pontiacs, and he will hand you a laminated three-by-five card with the ''Oldsmobile mission statement'' on it. To jump-start sales, GM gave Olds a minivan and a four-wheel-drive sport utility vehicle to sell, but even those were originally designed for other divisions. Some Olds dealers are so angry about their declining fortunes that they stormed out of a meeting with general manager Michael Losh at Las Vegas in February after he refused to authorize factory rebates to dealers for the new Olds minivan. While GM executives continue to praise the reorganization in public, bits and pieces keep coming undone. When Cadillac's unique identity started to blur under the pressure of corporate homogenization, the company agreed to give its most prestigious -- and profitable -- division its own engineering team. Cadillac embarked on an expedited project to make its cars longer and heavier, and Ward's says it was the only GM division that did not lose sales last year. Analyst Maryann Keller, a managing director of Furman Selz Mager Dietz & Birney and GM's most authoritative and vocal critic, has long argued that GM can return to glory only if it gives the same kind of autonomy to its other car divisions as it did to Cadillac. No fan of Keller's, Stempel refuses to consider the idea. Says he: ''One reorganization is enough for anyone's lifetime.'' His strategy is to continue to differentiate models through design and extra features -- which the customer can see -- while cutting costs under the sheet metal by using more common parts. Seventy years ago GM went through a similar crisis. When Alfred Sloan was confronted with seven overlapping brands in the 1920s, he jettisoned three money losers -- the now forgotten Oakland, Scripps-Booth, and Sheridan lines -- and imposed strict price segmentation on the rest. The strategy worked for half a century. Sloan, of course, never had to cope with the Japanese, and Ford was still in the hands of its eccentric founder. But he set a good example. GM urgently needs to slim down, focus its resources, and sharpen its aim. While it lumbered through the 1980s preoccupied with its own concerns, the world changed. Soon, with new leadership, GM will have a chance to change with it.

CHART: NOT AVAILABLE CREDIT: SOURCE: WARD'S AUTOMOTIVE REPORTS; HARBOR & ASSOCIATES THE WORLD OF GM In so vast an enterprise, getting all the components moving in the same direction is a major achievement. Since 1979, GM has made big strides overseas. In North America quality is up, but factory gains are scant, and profits and market share are persistent problems.

CHART: NOT AVAILABLE CREDIT: NO CREDIT CAPTION: INVESTOR'S SNAPSHOT GENERAL MOTORS