The early 1990s were lean years for many US corporations, particularly for airlines. In the first two years of the decade, airlines had a total combined loss of over $9 billion dollars, and there was no end in sight to the continued fiscal slump. Although it was a financially sound company, Delta Airlines found itself caught in the trap of costs that seemed to be spiraling out of control, thanks in part to its obsession for becoming a dominant worldwide carrier. For many airlines, revenue streams were either increasing or stagnant, but costs for programs, policies, and procedures that were remnants of the old regime of intensive government regulation remained in place, even a decade after the massive deregulation of the industry.

Delta was no exception to the lavish expenditures of airlines. Considered paternalistic, Delta offered such benefits as seven weeks of vacation to senior employees, and provided all-expenses-paid health care to the vast majority of its workers. In return, Delta enjoyed a high level of morale and dedication among its employees, although many thought this might have been too high a price to pay to attract the best work force. Delta had invested substantially in the growth of the airlines industry, right when the economy started to falter. It had absorbed Western Airlines in 1987, and the demise of Eastern Airlines in 1991 left Delta with almost monopoly power on flights in the southeast, particularly to and from its Atlanta hub. Delta also purchased from Pan Am all of its European routes, and absorbed 7,000 Pan Am employees in 1991. The Pan Am investment was a disaster, losing the company more than $1.65 billion over three years.

This resulted in one of the highest cost structures in the industry, 9.26 cents to fly one passenger one mile, which is the yardstick for measuring costs in the airlines business. Lower cost carriers were as much as 20% below Delta's expenses per passenger mile, and Delta was feeling the squeeze as many coach class passengers turned to other airlines that offered cheaper fares. Faced with the horrendous double whammy of declining market share and increasing costs, on April 28, 1994, Delta announced its Leadership 7.5 program. Leadership 7.5 embodied Delta's goal of cutting costs over a three-year period to the measure of 7.5 cents per passenger mile, which is on a par with low cost carriers such as Southwest or ValuJet. To attain these cost-cutting goals and prove to its stockholders that it was serious about reducing costs, Delta immediately eliminated 15,000 positions, and several thousand more over the coming months, a total of 20% of its workforce.

Lifetime employee Ronald Allen (no relation to CEO and downsizing disciple Robert Allen of AT&T) held the three most important positions at the company: Chairman of the Board, President, and CEO. He had spent seven years in that position at the time of the downsizing, and was crucial to its implementation. Allen hoped to cut over $2 billion from Delta's annual operating costs. Delta offered generous severance packages to several senior managers-along with other write-offs related to the downsizing, the carrier charged $1.65 billion in extraordinary expenses to the income for that fiscal year. Delta rallied over $2 (4.3% of the stock price) on a day during which the Dow fell more than 30 points (see Figure VII.1). This jump in the stock price was over two standard deviations above the mean return for the past 100 days on Delta.

The market's ebullience was augmented by a news release from the union which represented Delta's pilots: as a sign of good faith, the pilots' union announced that it would go to the bargaining table early for the upcoming expiration of its contracts. CNN reported that Delta wanted to cut more than $350 million of the unionized pilots' payroll as part of its strategy to cut $2 billion in costs, which amounted to 16% of its operating expenses.

Over the following weeks, Delta announced that it was going to completely revamp its services to rid itself of anything possibly superfluous. It eliminated its famous "Frequent Flyer" program, and replaced it with the more miserly "SkyMiles" program to reward loyal customers. Although many costs were reduced by small overtures, like putting less lettuce in the salads it served, the majority of the first round of expenditure reduction came by the swift decision to eliminate 20% of its workforce. The following chart tracks the value of Delta and the S&P 500, both normalized to 100 on the day before the announcement. In the 100 days following the implementation of Leadership 7.5 Delta's equity value rose more than 8%; the S&P 500 rose by 4% in contrast (see figure VII.2.).

By 1995, Delta was on-track with respect to its interim cost-cutting projections, and had still maintained a reputation for stellar customer service. Yet several problems emerged. With one fell swoop, Delta had eliminated managers with years of experience, as well as service personnel with extensive firm-specific skills. The carrier seemed completely unable to cope with the growth of airline traffic that soon followed the downsizing. Delta found itself having to hire back some of the laid-off workers at a premium-since skilled labor was in demand-and therefore failed to hire as much as it should have. This mistake caused Delta's customer service to plummet, as the part-time workers hired to replace the downsized full-time workers did not have the skills or the experience to deliver the level of customer service that Delta had previously offered. Delta's on-time performance went from the top to near the bottom of the industry; many dissatisfied business customers switched their loyalties. The Department of Transportation registered the least complaints for Delta than any other airline in 1993, the year before the downsizing. Delta plummeted to the fifth-worst airline after the downsizing took place. A Delta spokesman confessed to The New York Times that Delta's customer service had "deteriorated [because we] lost a lot of experience when we encouraged early retirements." Said another company spokesman to The Record: "There is a learning curve you have to go through with part-time people."

The spirit of paternalism soon vanished, and was replaced by resentment and frustration. Internal surveys obtained by The New York Times reveal that only 22% of Delta's workers considered the management "effective." Employees complained about the lack of innovation and strategy, and about the company's unwillingness to let employees make decisions. Delta could not hide the failures of its austerity program from its shareholders for too long. The market soon considered Leadership 7.5 more of a quick fix and less of a veritable strategy to improve Delta's operating margin. By 1997, Delta's costs were 8.75 cents per passenger mile, less than halfway to its stated goal. Enthusiasm for the stock waned as a lawsuit filed on behalf of disgruntled downsized employees contested the cut in benefits to laid-off workers. Delta abandoned its Leadership 7.5 program and replaced it with a "Balanced Strategy," which targets the carrier's operating margin, not simply its costs. Earnings fell short of Wall Street estimates, and net income was sluggish. Still, Delta was able to keep pace with the S&P 500 over the three years following the implementation of the program, perhaps due to a rosy economic outlook for airlines in general (see Figure VII.3).

As a result of the sufficient, but unimpressive results, longtime CEO Ronald Allen and staunch advocate of downsizing was himself downsized. Leadership 7.5 had clearly succeeded in cutting costs (by Allen's ouster, Delta's operations were the cheapest for any similar airline-8.5 cents per passenger mile), and in eliminating the financial slump brought about by the purchase of the Pan Am routes. Yet many wondered if such success had to be achieved with the stifling miserliness which Delta imposed on itself, to the detriment in the morale of its workers.

In a way, it seemed more costly to fire 20% of the carrier's workers than it did to provide them with limitless health insurance and two months of paid vacation. In his assessment of airlines, industry analyst Samuel Buttrick of Paine Webber noted that "morale fell a lot further than costs." The general obsession with cost control had led Delta to miss out on many investment opportunities in new technology. Its reservation system worked slower than many of its competitors', and it was one of the last airlines to offer electronic ticketing. Although it ousted a CEO, Delta's board did not out his ideology. The interim CEO to replace Allen, Executive Vice President Maurice M. Worth, told The New York Times that since Delta was in a "downsizing and cost-cutting mode" and that it "almost had to go too far to insure that [it had] gone far enough." With the recent boom in air traffic, Delta has found itself adding workers and offering service on new routes. Whether or not the fiasco of Leadership 7.5 really benefited the shareholders of Delta will be explored in the next section.

To actually determine if downsizing was in the long run a productive strategy for Delta, in terms of improving shareholder value, the scientific method would require us to have a control airline-let's call it Delta Prime-with the exact same cost structure and problems as the real Delta. The only difference between the two companies is that Delta Prime contemplated downsizing too, but opted not to. The test would be to see if Delta fared better than Delta Prime as a result of downsizing. Unfortunately for economists, such real-world control variables don't exist, but there did exist a real-world airline that almost perfectly matched the profile of a Delta Prime; that carrier was United Airlines.

In the summer of 1994, United and Delta would appear to the casual observer as very similar companies. They were more or less the same size, had the same pricing mechanisms, flew similar planes, had just began to expand their international routes, and were experiencing extreme difficulties in keeping costs under control. Both were highly worried about fierce competition from low-cost carriers such as Southwest and ValuJet (now AirTrans Corp.). By 1994, US carriers had lost a combined $14 billion over the course of the decade, with United and Delta each losing $1 billion of the aggregate. Since Delta downsized, and United did not, their example offers an interesting controlled experiment on the benefits of downsizing. Instead of downsizing, United offered its employees a deal: accept major wage cuts (25% of base salary) in exchange for a substantial equity stake (55%) in the firm. Also, if the offer were accepted, United promised that there would be no layoffs whatsoever over the next three years, in exchange for no strikes over the same time period. If the offer were rejected, United would downsize as well, laying off "tens of thousands" of workers, according to CEO Stephen Wolf. This union-led buyout allowed the management to impose wage concessions of over $4.9 billion in return for 55% ownership of the company. This stock began trading on the New York Stock Exchange on July 13, 1994, less than three months after the Delta downsizing.

United had also attempted a smaller-scale downsizing in early 1993, cutting 2,300 of 83,000 jobs after announcing large losses. The layoffs were part of a program to reduce costs by cutting flights and cities served, retiring aircraft, and eliminating the workers needed to maintain these operations. According to CEO Stephen M. Wolf, "the fundamental flaws in the [airline industry] threaten the long term financial health [of United]." At the time of the announcement, Northwest Airlines had just laid off 1,000 workers, and Continental and TWA were in bankruptcy. The stock of UAL, the parent company and sole owner of United Airlines, slid less than 1% for the day.

It might have been easier for United to offer its workers an equity stake instead of a massive downsizing in 1994, as its workers were slightly more unionized compared to those at Delta, and were therefore harder to eliminate. In contrast to the Delta plan, United's buyout was not well received by the stock market. Similar attempts by Eastern Airlines had failed, forcing the company into bankruptcy. The market at first boosted the newly issued stock due to satisfaction about the buyout plan's no-strike clause, but enthusiasm waned after the flight attendants' union rejected the plan. Figure VII.4. displays the 100 trading days following the announcement of the program, with the value of the S&P 500 and the stock of UAL Corp. both being normalized to 100 on the first day of employee ownership.

The volatility in United clearly demonstrates the breadth of uncertainty that the plan generated; it had the potential to fail, since United was not cutting costs, but rather deferring them. The wage cuts would help the firm's bottom line, but the firm had offered a major chunk of equity at no cost to the employees (except the theoretical opportunity cost of receiving reduced instead of full wages) and the equity would have a non-negative payoff no matter what. Even if the stock slid from $23 per share (the offering price) to $0 per share, the absolute worst that the employees could do under this plan was to make the reduced wage. The reduced wage in exchange for equity agreement constituted a legal liability of United's and in the eyes of the law, remained more senior than preferred debt in order of the claims against the firm, should it go into bankruptcy or reorganization. If the plan had been rejected, the worse any individual worker could do would be to make nothing (get downsized). This appears to be a classic example of risk aversion on behalf if the union. It chose a sure reduced wage instead of a full wage with a 20% probability of no wage (According to senior management, had the plan been rejected, United would have downsized 20% of its workers, just like Delta did). As an added incentive to choose the reduced wage, United added equity participation to the deal. This angered those who were already shareholders, as the plan did not entail a substantial decrease in costs. Moreover, their individual claims on the value of the firm were diluted by the huge increases in shares outstanding. In some ways, this made United a riskier company than before.

Perhaps the market did not first respond favorably to United's plan since the cost savings did not appear instantaneously as with a downsizing. Instead of saving money by eliminating employees, which is easy to measure, United chose to cut costs by reducing salaries somewhat, and also eliminating much of the so-called "agency costs" in the economics of United. Briefly, agency costs are losses in profit or utility associated with monitoring the employees so that they fulfill the wishes of the shareholders. Alternatively, the agency cost is the shareholder value lost by the employees due to divergences in the utility of specific actions to shareholders and to employees. Since the employees of United became its shareholders, they should theoretically have had the same interests as the shareholders, and have an incentive to do a better job, since they had acquired a long-term financial stake in the company. The examples of the reduction in agency costs abound; for example, sick time dropped 20% over the year, as did workmen's compensation claims. Many pilots left other airlines to join United, which now employs over 10,000 of them at a salary that is less than many of United's competitors'. One pilot told a New York Times reporter, "I have one major interest: I want to work for an airline that will still be around when I retire." Grievances filed against the company by union pilots fell a staggering 75%, and the union appeared more flexible to negotiate with the company on issues such as round-the-world service.

Eventually, the market realized the benefits to the reduction in agency costs associated with employee ownership, and the firm's value began to rebound. Additionally, since United did not fire workers, it was able to cope much better with the economic upswing than Delta. Its employee morale (and therefore productivity) did not suffer on a level of Delta's either. Over the year following the change in ownership, United appreciated significantly relative to the S&P 500, despite an initially poor start (see Figure VII.5.).

Hindsight is 20-20, and the question remains why the investors punished United relative to Delta at first, only to reward it later on. Perhaps the answer lies in the risk aversion of the plans each airline presented. Delta seemed intent on minimizing the worst-case scenario; United wanted to maximize the expected value of its business. Each plan has its merits: Delta wanted to prepare itself to deal with an onslaught of low-cost carriers and a reduction in business; the downside is that it would be completely unprepared to handle a boom or upswing in airline traffic or demand for its product. United, on the other hand, wanted to maintain its strength, and took the risky move of prolonging current expenditures by expensing them in the future (instead of paying cash salaries now, it chose to offer stock, which it will pay off in the future, if it ever does at all). This strategy was decidedly risky, since if demand for air travel fell, United would be stuck with replete payrolls and unable by contract to lay off any workers. Bankruptcy would have been a certainty.

The equilibrium of the game presented in Section VI.3. explains why firms of the same type (strong firms) might be observed making different choices in the marketplace. It certainly allows identical firms like Delta and Delta Prime (United) to make the different choices in equilibrium. A close examination of the payoffs will explain why this is the case. A strong Player 1 is faced with the decision to either downsize or not, given Player 2's strategies.

If he maintains the workforce, the market won't know if he is strong or inefficient, and chooses to dump the stock. After United announced that it was not going to downsize, investors didn't know if the company was strong or inefficient. A strong United would not downsize since it realized that it would be powerful within the airline industry, or face a healthy demand, and therefore keep on workers. An inefficient United would have simply been incapable of firing its workers since it is so highly unionized and requires too many people to make the company run. It is quite plausible to believe that in mid-1994, no one had any idea if a firm like United was strong or inefficient. Earnings per share and other financials in the airline industry are less meaningful, since clever accounting can manipulate the depreciation of aircraft, or the payment structure for new planes, or the expenses related to certain pension agreements. With inventive procedures for recognizing payables and receivables it is quite possible for an inefficient firm to fool the market into thinking it is strong for a short while. The management of United receives a payoff of 1 for making the right labor choice, and knows full well that its stock will not be popular with the market, and will be sold.

If the management downsizes, the market will not know whether it is weak or strong, and therefore will buy the stock only half of the time. The market randomizes, and half of the time, it will bid up the stock of the airline, giving it two points (the management gets no additional points since it made the wrong labor decisions). Even though the labor decision was wrong, the management is nicely compensated since the stock traded upwards. This could explain why a firm like Delta chose to downsize. If the market sells the stock (the other half of the time), the carrier's management is in the worst of both worlds since it was sold in a strong state, and made the wrong labor choice. In this case, the payoff is zero. Since each scenario is equally likely, the expected payoff is 1, and an airline will be indifferent between downsizing or not.

Therefore, it is possible to create situations where identical firms choose different employment policies as part of an equilibrium. When United offered to its workers the opportunity to either accept paycuts and equity or be downsized, we can assume that the management was indifferent between these two proposals, so it didn't care whether the firm downsized or not. Certainly, this is only one possible interpretation of United's decision. It could be that they offered this proposal to the union as a way of measuring the union's preferences over risk and utility. The obvious proof of indifference in this game is limited to the assumption that workers do not have any say in how the firm will be valued. An extension to this game could add labor as a player with the strategies of accepting or rejecting a contract. But for the simple illustrative purposes of interactions between management and owners that this thesis discusses, the outcome is sufficient to describe what actually happened. This is close to the theoretical strong Player 1 being indifferent between cutting workers and not cutting workers, and therefore determining what strategy he will pursue randomly.

Changing the payoff structure or the probabilities which nature chooses a firm type slightly will allow the strong type to never want to downsize, while allowing the weak types to randomize. The outcome of the game is clearly dependent of the arbitrariness of the payoffs I have chosen, but they correspond well to both executive compensation contracts and risk aversion among investors. Further research could impute a crude utility function of money for management and for investors that yields this result, given that the strategy choices and monetary payoffs are knows.