Saving Chrysler: the use and non-use of accounting information by the US Congress

Introduction

Government loan guarantees for the financially troubled Chrysler Corporation during the early 1980s constituted a unique episode in the history of US capitalism. A giant corporation on the verge of bankruptcy was saved by explicit government intervention into the credit market. Although loan guarantees had been a timehonoured tradition for the federal legislature, never before had they been offered to a single company totally engaged in the for-profit, private sector. Although short of direct subsidisation, government backing for Chrysler debt proved the difference between corporate success and failure.

Sometime during the 1970s, Chrysler management realised that corporate survival was no longer within their control. Like the rest of the automobile industry, Chrysler had been rocked by the OPEC oil embargo and the growing realisation that business as usual could not continue. Large and noisy bankruptcies had occurred in other sectors, however, not in the industry that in many ways exemplified US manufacturing preeminence. Chrysler began an orchestrated lobbying effort to obtain federal government assistance. An important element of this was recruiting Lee Iacocca, a high-profile Ford Motor Company executive. The Chrysler decision became the story of 1979, actually elevating a business event in a way that was unusual for its day. In December 1979, shortly before US embassy personnel were taken hostage in Iran, the US Congress approved loan guarantees for future Chrysler borrowing, thereby assuring that the company would have the credit market access that a large capital-intensive manufacturer needed.

With the advantage of hindsight, the Chrysler bailout, as it has colloquially become known, appears as a great victory for both government and private enterprise. Chrysler Corporation had been a profitable and viable corporation for many decades, offering employment to thousands and contributing to US competitiveness in the auto industry. Its restoration at this moment in time allowed it, and its successor entity,1 to continue today. The outcome determinative logic might even boost the incident to a legend in the emergent cooperation between polity and business. Although the Chrysler case now exists in the wake of much larger and much more expensive bailouts of entire industries,2 it still provides an interesting moment in governmental action.

The purpose of this paper is to study the role of accounting information in the Chrysler decision made by the US government. A very broad conception of accounting information is taken for this purpose. Because accounting information competes for the attention of decision-makers with other information, an attempt has been made to place accounting information within the context of a multifaceted decision. The totality of information needs to be considered so that accounting information will not be over-emphasised. Information, accounting or otherwise, may be useful to actual choice or it can provide a post-hoc patina of rationality to a decision made on other grounds. In either case, the information itself is a matter of interest to those interested in accounting processes, formal organisations and governmental action.

This paper uses two major sources of data. First, the popular press of that day was surveyed to provide a general understanding of the issues and the agenda for action. Secondly, deliberations contained in the Congressional Record were examined to glean factors pertinent to the actual decision.3 Admittedly, these sources are imperfect and incomplete. The print media has a bias toward the wishes of their readership and a tendency toward over-dramatisation. Members of Congress, if assumed to be partially motivated by the furtherance of their careers, can use this forum to speak to their key constituents and therefore may overrepresent certain influential interests. Nonetheless, the discourse discovered in these sources illustrate much about accounting and its ability to be a meaningful part of public sector debates.

The paper is divided into three main parts. The first part establishes those dimensions of the decision that could not be affected by accounting. Many of these facets are contextual or manifestly political. The second part considers elements that did use accounting information. Here, more attention is focused upon how this information was used. The last section develops areas of the Chrysler decision that could have benefilted from accounting information. The paper concludes with a discussion about public sector decisions involving corporate crises and the potentialities of accounting information and measurement.

Where accounting could not go

The use of precedent

Governmental actions should always be interpreted in the context of that which came before. Past legislative outcomes create an ex post legitimisation for the continuation of similar outcomes. Aid to Chrysler must be contextualised within its perceived place with past programs and its projected bearing upon future governmental actions.

Proponents of aid pointed out the extensive activity of the Federal government in loan guarantees. A detailed listing of such involvement was entered into the Congressional Record to convince skeptics. Although the size of the Chrysler request made it noteworthy, the government had already intervened on behalf of many specific companies and industries believed to have strategic importance or widespread economic impact.4

However, precedent, or more precisely its interpretation, cut both ways. Opponents distinguished the past and suggested that the Chrysler package was without a true forerunner. The support that Chrysler needed was far more extensive than previous cases. Furthermore, the Chrysler request presented an instance where very risky underlying business fundamentals were involved. The problem was neither short-term nor readily repairable. Instances of past refusals of assistance also served as precedent. The aggregate number of business failures to any given year (Proxmire, 1979) provides evidence that government inaction is the modal public response to private economic distress.

Thus, the lessons of past governmental programs of aid were unclear at the time of the Chrysler decision. With evidence on both sides, the decision upon the merits of the Chrysler case could be said to be in conformity with a historical pattern no matter which way it was resolved. The absence of accounting information is particularly startling in this matter. Successful programs of assistance were ones that resulted in a continuing company. No attempt was made to parse out the actual contribution of the assistance in ways that would have involved cash flows, solvency assessments or even profitability. Thus, any lessons were muted ones.

Legislators were perhaps more concerned with Chrysler as a precedent for the future than they were with consistency with the past. The spectre of a governmental inability to refuse any company in distress in the wake of a Chrysler bailout was proffered, albeit in an intentionally overstated manner (Panetta, 1979). The absence of decisional criteria made conversations of subsequent bailout candidates and susceptible industries range wildly. The epitome of this perspective characterised the Chrysler proposal as a push toward the replication of British subsidisation of industry of this era. Many worried that the problems of private industry would become an inexorable drain upon the public coffers. Again, accounting was not seen as a source of reliable criteria by which governmental assistance could be responsibly provided.

In sum, the decision makers struggled to put the Chrysler case into a historic frame of reference as an initial means of understanding its unique merits and its historical significance. Throughout, however, the process was very equivocal because the argumentation never rose above the anecdotal. For these purposes, accounting evidence may have been packed within these examples, but it is debatable whether this was sufficiently well understood to have any direct role in the contribution of the past or the projections of the future.

Philosophical positions

The violation of the separation of public sector and private sector that the Chrysler situation made visible brought with it heated controversy over the essence of private enterprise. The bailout proposal gave the government a rare opportunity to see itself against the background of the prevalent economic system. In this way, this debate may have paved the road to the more systematic interventions into the economy by the US government during the 1980s and 1990s.

Opponents of the bailout proposal attacked any program of special relief as a violation of the central tenets of US capitalism. One Congressman, with near religious fever, suggested that difficult times should be when government cleave closely to capitalistic principles (Ichord, 1979). In this view, free enterprise resembled a natural law that should not be contradicted. The integrity of strongly held beliefs that were central to the American way of life was undermined when an economic albatross such as Chrysler was permitted artificial support (Buckley, 1979).

The core of the principled arguments against a bailout rested within the notion that failure was a confirming experience for the exposure of private ventures to the test of the marketplace. Since the wellbeing of all was enhanced by the failure of the inefficient and ineffective (Paul, 1979). A governmental bailout was tantamount to a direct rejection of private enterprise (Friedman, 1979). Moreover, special assistance demonstrates a preference for a few private interests over the more diffuse public interests (Stevens, 1979).

Proponents of the bailout refuted the ideological conclusions of the free market thinkers and offered their own. The business version of the survival of the fittest was characterised as an outmoded depiction of the US economy now that the interconnections between elements of the system have increased (Nelson, 1979). Nothing sacred or novel was said to be involved when competition was already less than pure. Every large company already had an on-going array of relations with the government that were not at arm's length. A more assertive stance proposed that temporary intervention was consistent with the finest spirit of capitalism (Wylie, 1979). Along these lines, government's role in preventing market failure was called essential for the smooth functioning of the modern high-risk market (Eagleton, 1979).

Unlike bailout opponents, advocates indicated the desirability of an ad hoc analysis. Philosophic approaches about the proper function of government without sensitivity to the costs and benefits of particular circumstances were seen as selfdefeating. To the extent that ideological debate could be displaced, new types of information would take on increased importance. For example, accounting information could fill the breach once an inquiry into particular costs and benefits had been legitimated. However, there did not seem to be an appetite for the exploration of more technically rational questions about the efficiency of the allocation process. Arguments about philosophic predispositions could not set the tone for the introduction of accounting information.

Political process

The bailout of Chrysler did not occur through government initiated action. A strenuous lobbying effort by Chrysler proved to be very important to the final outcome. The process of pursuing government guarantees required the employment of skilled professional lobbyists deployed as early as 1978. For some, Chrysler's systematic efforts to obtain guarantees became a factor against assistance being extended (for example, Proxmire, 1979). However, for most, the rallying of support in this manner was business as usual in Washington DC.

Chrysler attempted to impress Congress with the urgency of its case. Delay in the deliberation could cause the company to deteriorate to the point that any aid would be a less consequential package. Many Congressmen faithfully relayed the message that time was of the essence (Hart, 1979). Those that saw the bailout as inevitable construed delay as a factor that would only increase its cost to taxpayers (Riegal, 1979). Others perceived delay as a factor that would render Chrysler unsalvageable (Newsweek, 1979). Opponents of the bill felt railroaded by the haste urged by its proponents, with at least one Congressman urging for more painstaking deliberations (Stockman, 1979). Ceteris paribus, this pace weighed against the assembly of tailor-made accounting information.

Public opinion was another force in the process of the bailout. Whereas Chrysler shareholders were silent and the other major automakers were divided, only unions were unambiguously sympathetic to government assistance to save jobs. However, external parties could not add specific information to the debate. They could clutter the process.

The multifaceted aspects of the process of attaining governmental support for Chrysler create a context that crowds out the use of real accounting information. In political processes, information is not neutral but instead exists within a system of resource mobilisation and interest alignment that itself is bounded by temporal constraints. In this instance, the accounting information could only be a small part of a much larger package that was manipulated and moderated in a short time frame by a host of parties with much to win or lose.

Party politics and maneuvering

Most of the political attention was focused upon the precarious position held by President Jimmy Carter. Despite early opposition within the Cabinet, the President was in favour of a governmentally assisted rescue well before other governmental groups were convinced. This type of involvement promised considerable political exposure. Federal interventions of this sort were "no-win" propositions (Time, 1979a). Carter's liability was exacerbated by the proximity of the 1980 election year and the political importance of the most affected state (Michigan) to reelection prospects (Chapman, 1979). The President's situation was kindly described as trapped between the political power of the unions, that dwelling upon the employment issue favoured the bailout, and the logical conclusion of his own populist image, that abhorred anything that suggested corporate welfare.

The political dimension of the bailout decision cast Chrysler in a swirl of many other issues only marginally related to the company. The company produced a fleet of "gas guzzlers" at a point when the consequences of petroleum dependency were painfully apparent. Broader reforms of capitalism were also vogue with ideas that ranged from small degrees of worker board representation to complete public ownership (Nicholson, 1979b; Brill, 1980). The Chrysler episode was used as a bully pulpit for these agendas.

The Chrysler bailout was greatly facilitated by the failure of an opposition to coalesce. Support for loan guarantees, although sponsored by a Democratic president, was not limited to his party. The bipartisan nature of the emerging program was assisted by the pro-business ideology of the Republican party and by the absence of a narrow regional benefit. As Cameron (1979) puts it, "No one is anti-Chrysler". Congressmen that supported the Chrysler bailout did so without much adverse political exposure for their position. Unlike some issues, the Chrysler decision was sufficiently multi-dimensional so that support could not be interpreted in a black and white manner. Although not lacking the usual political tussles, the Chrysler decision did not have the sort of battle lines that required the son of strong and apparently neutral type of information that accounting systems were believed to produce.

Interested parties sacrifice

The idea that those who stood to gain should first sacrifice toward the goal of saving Chrysler was given great currency. However, little agreement existed on how much private sacrifice would be enough. Several individuals suggested that the concessions already made were small and would have to be substantially augmented (New York Times, 1979; Hinson, 1979). Others delighted in detailing the voluntary losses accepted by those who sought governmental aid as a good reason for public assistance.

Compensation package concessions made by the UAW figured prominently as a condition for the bailout. There was an undercurrent of feeling that Chrysler workers were overpaid by national standards and should not be allowed additional cost of living adjustments until their company's time of crisis was ended. The early agreement obtained from the UAW on the delay of benefits provided needed momentum for the assault on Congress (Pasztor, 1979).5 State governments had an obvious interest in maintaining Chrysler on their tax rolls. Not surprisingly, Michigan was first to offer tax concessions, followed by tentative promises from several other states. In such a delicate balance, some arm-twisting was possible through the imposition of necessary conditions on the federal aid component. Throughout, the prospects of non-cooperation which would threaten the entire arrangement was always present. Although these concessions may have been predicated on some calculus of affordability, they were explicitly normative in origin and symbolic in importance.

New management

Management's involvement with the legacy of economic losses at Chrysler's made the bailout difficult to sell. If nothing else, the credibility and objectivity of the request was at issue. Furthermore, the safety of federal funds in the hands of those that had so much experience losing money was problematic. Pan of the process of demonstrating worthiness was the jettison of this negative imagery and the creation of a more suitable corporate persona.

The purge of Chrysler management prior to the debate over the bailout created a symbolic desire to disassociate the corporation from its past. This would not be complete or effective without the withdrawal of John Riccardo, the man who oversaw the years of record losses. His resignation, officially for health reasons, was done with one eye toward the bailout prospects (Business Week, 1979b). This departure not only reduced the stigma of bad management, but it also allowed new management the ability to conduct a more thorough housecleaning. Several firings of long-tenured executives and their replacement with external talent soon followed. As part gesture and part reality, the turnover at the top of Chrysler was often noted as a factor that made the bailout decision more palatable to Congress.

The replacement of past management would not be consequential without the accompanying belief that new management would be capable of more desirable outcomes. Chrysler was successful in implanting the belief that positive steps already had been taken that would reverse the company's fortunes. Unfortunately, this was partially undermined by the long lead times required by the nature of the productive process in this industry (Stuart, 1980). Two significant moves that were performed quickly were the restoration of reliable financial controls and the elimination of excessive bureaucratic procedures that had surrounded critical tasks. However, the monetary value of these changes in terms of better information and more timely processing of it was not estimated. Softer changes, such as the addition of the word "New" to the Chrysler Corporation's name was as much aimed at Congress as it was at the car buying public.

The personification of the resurrection of Chrysler stands out as a truly exceptional moment during this anonymous era of capitalism. The embodiment of the bailout was Lee Iacocca. Stepping into the breach with unequalled style and confidence, Iacocca came as close to a charismatic leader as any figure in American industry since the age of the robber barons. Much should be made of his infectious risk taking and calming presence. The abandonment of his pension rights from Ford Motor Company in order to take the Chrysler position, combined with his waiver of his Chrysler salary during the bailout request period added a heroic dimension to the drama. Although this was partially a carefully constructed public relations effort, it engineered a turnaround wherein the persons comprising Chrysler management were changed from a liability to an asset. In this age of disbelief and skepticism about corporate leadership, an apparent exception was made for Lee Iacocca.6 He moved quickly on the marketing front. Iacocca's television presence, plus some unprecedented efforts to generate volume sales (Time, 1980), at least demonstrated that something extraordinary was being attempted. The effort made progress toward the restoration of faith by a shaky dealer network. Although the promotion of the belief that well-made cars were coming from the New Chrysler preceded its actuality, movement in that direction started at the top of the company. Marketing made the belief that sales would return tenable, even in the absence of actual revenue.

Summary

Much about the Chrysler decision could not, by its nature, entail accounting information. The dimensions of the situation that were explicitly political were beyond the influence of this worldview and expertise. The political realm, at least as involved here, verged toward binary discriminations that could not be rendered more fine-toothed. Another domain of relevant behaviour was heavily symbolic and motivated as a means of distancing the past from the present. This domain struck another blow against the possible relevance of accounting information.

Accounting in action

Loss magnitudes

The red ink bled by Chrysler during the years that surrounded the bailout was legendary for its time. The interpretation of the bad news summarised by this bottom line, as well as its major components, can be safely suggested to be a major factor in the Chrysler decision. Unlike most other elements, the evidence of losses was considered to be fairly unambiguous. Furthermore, the implications of these numbers also were agreed upon by the media and Congress. Almost without exception, commentators and decision-makers believed that this portion of the accounting information spoke for itself. Losses suggested unsustainability.

A variety of loss figures were recited. The record shattering third quarter of 1979 was most often mentioned. This particular result stands out because it was announced in the midst of the bailout deliberations and because it exceeded the most pessimistic expectations of the time. Perversely, this outcome was believed to enhance Chrysler's position that its plight was neither slight nor self-reversing. The third quarter report occurred around the time that the federal administration proposed to double Chrysler's initial loan guarantee request. A causal connection between these events was imputed but never officially recognised (Nicholson, 1979a; Wall Street Journal, 1979). Third quarter losses of $460 million, a record for American firms,7 made the previous quarter's sizable loss of $207 million appear small in comparison. The tale of 1979 showed exponential acceleration in losses from the modest $53 million of the first quarter to a projected total year loss exceeding one billion dollars. Moreover, projections of the future were equally bleak. The official estimate of $482 million of losses for 1980 was roundly jeered as hopelessly optimistic. One subsequent analysis that reconstructed these numbers suggested that Chrysler faced at least $2.5 billion in future losses at the time of its first plea for aid (Stuart, 1980).

The little accounting information with direct consequences for future profitability did not provide any more cause for optimism. Chrysler profit margins were small on the few models with good sale forecasts (Time, 1979c). Market share was at a post-World War II low of 10 per cent and projected to go still lower.8 Already stiff competition was expected to get sterner as GM introduced its much anticipated, highly praised X-car series (Economist, 1979b). Summer sales in 1979 were down by nearly 40 per cent from 1978, a fact which gainsaid Chrysler recovery scenarios. Shortfalls in net working capital also strained the credibility of short-run capital spending plans, and forced Chrysler to cut production. However, the debate over Chrysler did not painstakingly attempt to assess the chances for a recovery with meaningful forward-looking accounting information. If anything, the historical cost information weighed heavily as unproblematic predictors of the future that required more cash to be thrown at.

Chrysler's past profit history also provides a context for its 1979 results. Chrysler's 1979 losses reflect earlier decisions made under a severe short-run rationality that had ultimately unravelled. The company's immediate pre-bailout history revealed a rapid sequence of crises, each more narrowly averted than the last. The previous dozen years provided inadequate average annual profits to pursue an adequate modernisation policy (Schuyter, 1978). Yet only for 1975 were dividends suspended. Chrysler could never stay ahead of the profit plowback necessary in an industry where continual expensive retooling is a predictor of competitive success. Chrysler's 1979 crisis capped a 25 year period where return on sales had been improved over the previous year a scant four times (Stockman, 1979). The company's lack of success in the market merits special note since the costs of dwindling sales are asymmetrically large in the auto industry (Heinz, 1979).

Walking such a thin line, Chrysler's credibility concerning their understanding of their own fortunes was doubly tested. Past failures to realistically estimate their position, together with the occasional bursts of false optimism, made it more likely that Chrysler's condition was worse than that revealed by the conventional accounting information.

The nature of costs

Chrysler's size prohibited it from exploiting all the economies of scale that existed in production. Per unit production costs consistently exceeded comparable costs at Ford and GM by 10 per cent or more. Marketing costs per unit were inflated by high warranty and rebate costs (Norman, 1980). The former was attributable to the lack of quality that had crept into design and production processes. The latter was necessary in order to induce buyers to purchase a product that no longer was competitive in non-price terms. Labour costs, maintained by one of the strongest US unions, were high, even relative to other companies in the auto industry. Since these arrangements were collectively bargained under long term contracts with the Big Three auto makers, they were less reflective of worker productivity and specific company ability to pay. In this time of fiscal crisis, Chrysler's labour costs were less reducible as a result. Executive wages also were exorbitant if weighed by company performance.

For years, Chrysler produced cars in advance of dealer orders and, as a result, amassed a gigantic finished goods inventory. Weekly inventory costs of two million dollars (including parking, maintenance, moving, record-keeping and advertising charges) relentlessly accumulated. Hundreds of millions of dollars of sunk costs were invested in the dubious belief that sales would eventually catch up with production. As one model year was overtaken by another, inventory depreciated rapidly. Perhaps more telling was management's previous unawareness of the situation or their unwillingness to confront it. A recently installed system of cost control gave management the ability to measure some costs for the first time (Dizard, 1980).

Capital needs

In an attempt to appear profitable, Chrysler had deferred maintenance programs and only fitfully participated in research and development. Faced with an unprecedented retooling imperative, the company's need for capital was much larger than at any previous time. Unfortunately, dwindling profits from sales precluded a strong internal contribution to capital formation, leaving Chrysler dependent upon external sources. Debt service on past borrowing drained away a large portion of liquid funds.9

Exactly how much Chrysler needed was not often clear. The dilemma for the company was to seek loan guarantees sufficiently large to raise enough capital to make a difference, yet not ask for so much that the company's needs appear infinite. The choice was made more difficult because the governmental resistance to repetitive requests was apparent ab initia. John Riccarclo. prior to the bailout campaign, started a $7.5 billion capital drive to enable the company to meet the minimum 1984 EPA mileage standards. Projected capital expenditure needs for 1979 alone exceeded one billion dollars, or $10.55 per share (Business Week, 1979c). Chrysler originally requested $750 million in loan guarantees. This low figure seemed astronomical until more of Chrysler's operating results for 1979 became known. Some analysts termed even one billion dollars a stop gap measure and suggested as much as seven billion would be more realistic (Anderson, 1979b). Even according to the company's own spending plan, Chrysler was to be outspent by both GM and Ford (Time, 1979d). Many found it ironic that the Carter administration, by doubling Chrysler's loan guarantee request to $1.5 billion appeared to understand the company's situation better than management.

Summary

As the definitive historical record, nothing exceeded the ability of accounting information to reveal the plight of Chrysler in 1979. If the losses were not clear, the trajectory made them so. The explanations of the numbers revealed the endgame of previous accounting abuse (in an era before earnings were appreciated as pervasive) and the fundamentals of a business that did not have much hope. Not much about the numbers suggested that a turnabout would be successful. The more Chrysler invested in systems to discover their position, the bleaker it became.

Where accounting could have helped

The role of regulation

If the costs of regulatory compliance contributed heavily to the company's present financial distress, a bailout could be legitimated as an attempt to restore governmental neutrality. Governmental assistance would be almost compensatory for a network of regulations that had run amok. However, dysfunctional regulatory overload also may have been only a convenient scapegoat to blame for other problems about which the US government would be less sympathetic. Regulatory standards mandated an aggressive timetable for innovations in process and design that were unprecedented for this industry. Chrysler's reaction was indicative of the US automakers of that time, seriously discounting the possibility that car buyers actually wanted safety and gas efficiency.

Those that were disinclined toward aid for Chrysler did not assail this argument directly. They admitted that regulation was expensive but stressed its social desirability. They suggested that in Chrysler's case its impact was not as large as the company purported. No evidence existed that the regulatory burden alone made Chrysler unprofitable (Byrd, 1979). However, Chrysler's regulatory burden may have been disproportionate to other firms. Regulation tends to impose fixed costs upon firms. If these costs cannot be spread over a large volume of sales, the amount of cost per unit will be high. According to some estimates, regulation cost per car for Chrysler was nearly double that of GM in 1978 (Economist, 1979d).

Although the regulatory issue was pivotal, no systematic attempt was made to deal with the accounting information that might have detailed the extent of this problem. Regulatory costs were very grossly dealt with as large aggregate numbers were bantered about in the debate over Chrysler. Tracking regulatory costs through the supply chain would have exposed more about the unfunded mandated nature of regulation than politicians could afford to do.

Macroeconomic impact

The case for a Chrysler bailout was supported by the system-wide macroeconomic impact of a Chrysler failure. The shock value of the cumulative consequences coexisted with the detailed examination of selected sectors of harm. Collectively, these numbers tell a story that depends upon many implicit assumptions about economic behaviour.

Total GNP loss attributable to a Chrysler bankruptcy was estimated at $32 billion as a worse case scenario and $18-20 billion under the assumption that a healthy redeployment of resources would occur (Riegal, 1979). A large portion of this would be lost tax revenue, estimated at $10 billion (Brill, 1980) or less pessimistically, $6 billion (Riegal, 1979). The loss of tax revenue, plus the extra cost of unemployment benefits were estimated at $16 billion (Bohr, 1979). Less precisely denominated estimates of overall economic loss of between $10 billion and $2.7 billion (USNWR, 1979b) were offered. Any loss would add to a federal deficit, which at this time was just beginning to garner concern. Estimates of this incremental worsening included $2.75 billion for 1980 and 1981 (Whitten, 1979) and $11 billion over a longer time frame (Riegal, 1979). Revenue losses for state and local governments were estimated at $266 million (Whitten, 1979). The various estimates cannot be harmonised or compared due to the reckless, unsupported and casual manner they were presented, invariably providing no clue as to what assumptions were utilised. Nonetheless, the rhetorical power of these numbers was not dependent upon their ability to clarify and inform the decision.

The collapse of Chrysler was expected to deal a blow to the capital markets primarily through the worthlessness of that company's common stock and the uncollectability of over a billion dollars worth of debt. Most of the risk of this event had already been impounded into the current value of the securities at the time of the bailout. Chrysler stock historically had been more sensitive to decreases in sales volume than its American competitors (Nevens, 1978). Moody's downgraded both Chrysler and its subsidiary Chrysler Financial Corporation twice during a four month period. Chrysler's sale of preferred stock in 1977 had met with considerable reluctance within the investment banking community. One member of the distribution syndicate refused orders, while others sold only on an unsolicited basis (Nevens, 1978). Thus, expert judgement on Wall Street seemed to discount the likelihood of a successful turnaround.

Although the bulk of the macroeconomic impact predictions forecast the impending doom in the wake of a Chrysler failure, others suggested an adversity would be ushered in by an attempt to save Chrysler. Inflation would be aggravated by this unnecessary governmental expenditure if Chrysler loan guarantees had to be honoured, or even if not, by the large expansion of the aggregate amount of new credit (Paul, 1979). Friedman (1979) labelled the proposal a pure deviation from the principle that resources should be allocated based on productivity, and therefore was certain to raise interest rates as an unintended consequence. Those that predicted a Chrysler failure-induced recession were attacked as reckless statements that underestimated the recuperative powers of the economy (Time, 1979d; Panetta, 1979).

The primitive economic analysis of isolated facts was another context for the Chrysler decision. The political process was unable to deal with a general equilibrium analysis which required a much more sophisticated set of tools than most that had voice in this debate could appreciate. For these purposes, the science of economics built a general structure that was not accountable. Therefore, more precise accounting data could not be brought to bear on the question of economic impact and the fairness of distributional consequences.

Governmental transfer payments

Government's role us the provider of last resort was highly involved in the Chrysler deliberations. Claims made against governmental transfer funds by adversely affected parties revealed a secondary public cost if Chrysler declared bankruptcy. If Chrysler could be saved and these claims prevented, all taxpayers would benefit.

Unemployment benefits would be the right of those whose jobs would be lost after a Chrysler bankruptcy. The most often quoted cost estimate of this element was 1 .5 billion dollars. Large as it was, that figure failed to include the loss of tax revenue from the wages that were not earned. Unlike wages, unemployment benefits were not subject to federal income tax in 1979. This magnified the figure to $2.75 billion (Wylie, 1979). This large shock, incurred nearly at once, threatened the solvency of the unemployment system which was still reeling from the mid-1970s economic recession (Riegal, 1979).

Less obvious was the impact of a bankruptcy upon the Pension Benefit Guarantee Corporation. Chrysler's bankruptcy would almost certainly entail default upon its mostly unfunded vested pension benefits. The estimated cost of 1.1 billion dollars exceeded the total assets of the Guarantee Corporation (Broomfield, 1979). Restoring the solvency of this entity following a loss of this proportion would require that fees collected from employees be tripled (Byrd, 1979). A collapse of this magnitude threatened the effort of the government to assure employees about the safety of their pensions from private sector employment.

Although the secondary macro-level impacts were every bit as speculative as the primary ones, they also revealed the extent to which organisations were tied together in ways that could not be shown by the conventional accounting measures. The collective interest in the success of enterprises belies the entity basis of modern accounting but certainly is a legitimate element for the polity.

Employment

Perhaps the most salient issue in the Chrysler decision involved the government's role in saving jobs that would be lost in the event Chrysler was allowed to fail. Although it is part of the macroeconomic impact, employment merits a more selective investigation. The employment consequences, perhaps because of their often nonquantifiable human cost, were a highly politicised concern. In fact, the importance of saving jobs in the Chrysler controversy lead one writer to suggest it was the only real issue (Chapman, 1979).

The most widely quoted direct loss estimate of a Chrysler failure was 500,000 jobs, although upwards of 600,000 was also mentioned (USNWR, 1979a). The objectivity of these estimates was challenged since they included certain assumptions about employment by suppliers and dealers (Vitullo-Martin, 1979) and the redeployment of productive facilities (Stockman, 1979). Although the agreement over the unemployment figures was small, it exceeded the consensus over a proper interpretation. Nelson (1979) thought job loss was a mandate for governmental action. In that vein, others projected the total citizenry adversely affected, given family structures and other support relationships, at 2 million people (Patterson, 1979). Rodino (1979) interpreted the employment shock as even more severe, due to the high unemployment levels that already existed in the late 1970s. Others saw the employment consequences as regrettable but inescapable (Shumway, 1979; Kelly, 1979). For these congressmen, employment concerns had to be weighed relative to other goals for government.

During the Chrysler deliberations, the consulting firm DRI issued macroeconomic projections estimating the ripple effect of the unemployment that would occur if Chrysler went away. Great resultant unemployment was suggested, occurring mostly from the failure of Chrysler's many suppliers. Broomfield (1979) cited this report in the estimate that 80 per cent of Chrysler's national network of auto dealerships would close within two weeks of a Chrysler bankruptcy. The small business status of many of these operators was mentioned as deserving of special relief (Vanik, 1979). Yet the multiplier effects of an initial employment shock were not subscribed to by all. Some characterised DRI's work as overstated and speculative (Byrd, 1979) and as wrong and exaggerated (Bethune, 1979).

The belief that saving Chrysler would automatically save jobs was shared by most. No congressional notice existed that a substantial amount of employment reduction on a permanent basis was inevitable. Large losses should have indicated to more people that competitive conditions could no longer afford a Chrysler Corporation that supplied as many jobs as it did. No discussion existed on the need to reengineer the fundamental way that Chrysler did business. The job loss prospects were predicated on the assumption that the corporation could be restored in toto to sustainable profitability. An examination of labour costs, workforce productivity and alternative mixes of capital and labour would have provided a much different discourse.

The premise that job loss would result even if Chrysler was provided loan guarantees was better recognised in the media than in the halls of Congress. One estimate indicated that Chrysler employment would be halved even with a bailout (Chapman, 1979). Layoffs were a matter of economic conditions unalterable by government-guaranteed liquidity. Evidence for this proposition was indirectly given by the extent of the Chrysler layoffs that occurred prior to the request for aid. Chrysler's experience in other countries also was relevant to expectations about employment prospects. Canadian subsidies had failed to ensure continued employment (Economist, 1980b). Massive layoffs in Australia were also attributed to the inevitabilities of the marketplace (Economist, 1979a). The employment of individuals in outmoded and inefficient facilities was just not viable under any financing arrangement (Brown, 1979). Government loan guarantees were unable to address the structural problem of long run job insecurity in the auto industry. Stockman (1979) criticised the bailout plan because no guarantees of job retention were included.

The Chrysler decision could never be successfully decoupled from the employment issue. The company's problems were symptomatic of a macro-scale economic transition triggered by the new realities of foreign competition in the sector. The employees who would become short-run adjustment victims in this shift took centre stage in the debate although their fate was predestined. At this time, the fact that many quality jobs simply have no American future was relatively new. When specific accounting information on worker productivity asset turnover and product sales were left unconsidered, a bailout for the rank and file could front for a bailout of bankers, stockholders and management. The former was a palatable image projected by many (see Metzenbaum, 1979; Vanik, 1979). Very few saw that the glory of saving jobs made other more difficult redistributions tenable.

Other economic issues

Although Chrysler was a national company that was an important part of an industry closely aligned to the national identity of the time, its demise also posed a concentrated threat to specific regions of the US. Members of Congress that represented these areas could not be indifferent to these specific effects and therefore belaboured the Congressional Record with estimates of job loss, multiplier effects, and harm to the property tax base (for example, Hillis, 1979; Bayh, 1979; Oaker, 1979). Lacking meaningful accounting for state and local governments, these effects were difficult to assess and therefore, could not be systematically understood.

At the same time, the Chrysler decision identified for others the larger decisions that would be needed to chart a course out of the economic malaise of the day. Saving one company made it possible to deny that the economy in transition might require a new form of government-industry cooperation. The Chrysler bailout was effectively decoupled from larger problems that might call for the government to provide selective but periodic assistance to private sector enterprises. The information that would be necessary to identify what this responsibility might be and what should be done about it either did not exist or was not used.

From the perspective of the national economy the entity called Chrysler was unnecessary if a way could be designed to preserve the jobs and productive capacity that it currently represented. However, lacking public sector courage and creativity, these bold ideas garnered no support. Without high quality information, the bailout proceeded as a very conventional solution. Accounting, organised on an entity basis, had very clear boundaries for measurement that were favoured by the bailout process. Accounting historically had defined the box about which no one dared to think outside.

Stimulating productivity

One of Chrysler's many problems was poor productivity. For years, the work force had been noted for its apathy, its lack of concern with product quality, its absenteeism and its high wages. One proposal that billed itself as more far sighted than business-as-usual was an attempt to build worker incentive through stock ownership. Expanding the workers' stake in this manner not only might improve productivity, but also lessen the extent to which federal assistance would work as a windfall to current stockholders (Lundine, 1979).

The general belief that Employee Stock Ownership Plans (ESOP) would facilitate productivity increases underlie the early notion that the creation of one at Chrysler was a vital condition of the bailout (Business Week, 1979a; 1980). Senator Russell Long and other long-time ESOP advocates saw the Chrysler incident as an opportunity to further worker ownership. The ESOP idea called for the creation of convertible preferred stock that would contain dividend restrictions until federally guaranteed loans were repaid. The initial ESOP proposal called for worker ownership of up to 40 per cent of Chrysler stock, and would require a 15-20 per cent dilution of current shareholdings (Bloomington Herald, 1979). As the debate continued, the ESOP proposal gradually whittled down to a barely noticeable level before being completely forgotten.

Whereas loan guarantees made loaning Chrysler more money easier, some reasoned that Chrysler needed more equity, not more debt (Lugar, 1919; Lundine, 1979). Any infusion of capital would bring interested parties with demands for input into managerial decision-making. Much speculation existed about potential mergers, despite the fact that Chrysler did not present an attractive acquisition target (Time, 1979b; Economist, 1979c).

Productivity was approached very obliquely in these proposals. Providing workers with a stronger incentive or expecting this incentive to come from new owners essentially bypasses the more fundamental issues of how the work is organised and what makes a task worthy of performing. Both bypass reams of important data more closely connected to production and distribution.

Re-making the company

The most visible acts undertaken by Chrysler before the bailout was the sale of foreign operations. Under its previous leadership, Chrysler had become a multinational automobile producer. Although much money and managerial effort were invested, non-US operations never lived up to expectations. Sale of productive capacity abroad produced some liquidity. However, these transactions' largest contribution was to eliminate debt from the balance sheet (Business Week, 1979c). Other sales were made domestically that reduced Chrysler to an undiversified automotive maker and therefore presented the image of a "pure play" for Congress that would otherwise have been tainted by a diversified conglomerate structure.

The shortening of the Chrysler product domain had already begun. The company had begun to close out models in the large car market. This was an inevitable reaction of a smaller company to the change in the structure of domestic demand (Economist, 1980a). During this period, Chrysler production also became less integrated as marginal processes were contracted out so that more assets could be sold (Hunter, 1983). However, to make decisions like this because of an externally imposed radical restructuring of the company, as one of the terms of the bailout, would be quite a different proposition for the company.

The conventional wisdom about Chrysler was that full-scale competition with the then gigantic and powerful GM was an impossible dream that had contributed to the company's decline (Anderson, 1979a). Accepting a limited number of profitable niches seemed a more viable strategy than any scenario that required market share to be increased across all product lines. This option was especially attractive in the context of bailout proceedings because it limited the amount of investment capital Chrysler would need (Stockman, 1979) and would allow the firm to sell some of its no longer necessary full-line assets. However, the transition costs of drastically altering the present product mix would be high because a nearcomplete retooling would be necessary. The focus on smaller cars that some suggested was resisted by the company as not sufficiently profitable, given its current deployment of fixed assets (Blanchard, 1979) and the already overcrowding in this area by other manufacturers.

Congress had little appetite to impose on Chrysler any theory about how firms should be organised so as to be best positioned to survive. Although accounting could detail how flawed previous strategy had been, it would not be consulted to select a set of choices for the future. Although competitive information existed, using it would invade the prerogatives of management to react as necessary. This privilege was strong, even in a disgraced company.

Forecasting the future

The advisability of federal loan guarantees could only be assessed with some governmental ability to imagine the future of Chrysler. The best information available at that time was generated by Chrysler itself, and thus its substance and its credibility were at issue.

Chrysler projected that it would return to modest profitability in 1981, first by halving its 1979 losses in 1980. This was based on a belief in growing market penetration across all segments over those years. Cost savings (both fixed and variable) of over one billion dollars would be necessary to achieve the projected targets. Sales were projected to increase by 42 per cent or three times the industry average. Robust macroeconomic growth and low inflation for the US economy were assumed for the 1980-1985 period.

This scenario was greatly at odds with those assembled by other parties. These issues centred on the plausibility of building market share in a time of product mix shifts and a recessionary economic climate. Chrysler's projections were assailed as untenable (Byrd, 1979), ridiculous (Panetta, 1979) and naive (Panetta, 1979). Poor financial and accounting controls made any prognostication inherently unreliable. Lacking a coherent information system made the past, which Chrysler was attempting to break from, itself more uncertain. Congress also complained that financial projections prepared by Chrysler were sloppily done and insufficiently documented. Assumptions were often undisclosed and when they were, very little sensitivity analysis was performed around them (Stockman, 1979). The company was reluctant to prepare worse case scenarios and was less than open about how it could achieve the future it desired. Congressmen continued to demand more detail and longer run projections (Wall Street Journal, 1979) than Chrysler had itself thought to provide.

Although the calls for accounting to be forward-looking were yet to be made, such information did exist. The projections that were made, lacking supporting detail, seemed to be little beyond naked hopes. Accordingly, they presented no defense against their critics. That they could be made, and apparently be believed in, may have been more important than their accuracy. The information that existed may have been withheld because it did not point in the direction that those that possessed it wished.

Previous mismanagement

The mismanagement of Chrysler was a fact well within the awareness of both the media and Congress. The salience of this can be illustrated by Congressional statements suggesting the need to probe further into the extent of management's culpability (Anderson, 1979; Hinson, 1979). Those that opposed relief dwelled upon the contribution of mismanagement to Chrysler's condition (Economist, 1979d) as a reason for government inaction (Panetta, 1979). CEO Lee Iacocca was forced to admit a past of pervasively weak management at Chrysler that he estimated created between 50 to 75 per cent of Chrysler's woes (Vitullo-Martin, 1980).

The specifics of mismanagement amounted to a litany of horrors. However, these were countered by arguments that excused the history of bad decisions. For example, conditions such as inflation and foreign competition reduced the margin for management error, and were part of the conjuncture of inopportune events that impaired Chrysler's cash position and justified government assistance (Albosta, 19/79). This reasoning portrayed Chrysler as an unfortunate economic adjustment victim that lacked the critical mass to internally cope with its problems (Moorhead, 1979).

The accounting system at Chrysler, both in its importance and in its ineptitude, contributed to the charge of mismanagement. Poor financial controls made the symptoms of organisational distress difficult to diagnose. A surprising lack of cost controls made production decisions impossible to evaluate. Ironically, Chrysler's forte had been financial accounting magic. Three of the four CEO's before Iacocca were trained as public accountants, and contributed to a "tone at the top" that was before its time in its obsession with achieving financial targets. The privileging of financial over managerial accounting undermined the corporation's ability to learn about itself, as it devoted significant effort to preventing others from doing so. But these were the 1970s, and the accounting information was still seen as much more invariant and absolute at that time than now.

Summary

This section shows the many areas in which accounting information could have been deployed to make an intelligent policy decision. In many cases, this could have tempered the wild ranges of possible impact that were otherwise recklessly offered by advocates. In other occasions, such information could have been useful to suggest the feasibility of future actions. The volume of these areas of potential clarification suggests the need to pose the question why further efforts to develop and use this information were not employed.

Discussion and conclusion

Congress approved the loan guarantees sought by Chrysler shortly before the 1979 Christmas recess. Signed promptly by President Carter, the bill became law and Chrysler eventually was saved. Access to the credit markets allowed the company to buy time until it could rebuild a product line that customers actually valued. Although equity valuation did not rebound for some time, those that believed in the company ultimately were rewarded. The loan guarantees not only did not need to be invoked by creditors, but they earned a small fee for the government. The company continues to this day as a major part of the automotive industry even though, reflective of the progressive rationalisation of manufacturing, it now provides many fewer jobs.

Burchell et. al. (1980) stipulate four roles for accounting information in organisations and society. These functions include accounting as an answer machine, a learning machine, an ammunition machine and as a rationalisation machine. Considerable use of this classification scheme has occurred over the last twenty-five years to better understand contemporary decisions in organisations. This template can also be used in a historical perspective for choices made in the public sector. In fact, the hindsight enabled by history allows these roles to be clearer than they were when the circumstances were more contemporary. The use of accounting by government also facilitates the examination of more decisional dimensions.

Accounting provided few answers for Congress in the fall of 1979. If accounting was a predictable time series, the losses accumulated by Chrysler might have suggested that there was no hope for this company. Beyond the bottom line lay staggering capital needs and lousy fundamentals that could not have led to a bailout. The accounting information at Chrysler also failed to function as a learning machine. The company had no shortage of evidence on the results of operations yielded by strategies that did not work. Its attempts to distance itself from its past also implies that learning from the past is mostly moot.

The examination of the accounting and related information evoked in the bailout of the Chrysler Corporation in 1979 demonstrates the considerable role of the ammunition and rationalisation functions spelled out by Burchell et al. (1980). These roles tend to overlap in that they both presume that a decision was made apart from the logic of the accounting information. Congressmen in particular bombarded their colleagues and their constituents with a mountain of facts, most of which could not be proven or contextualised. This ammunition may have facilitated the persuasion of the undecided. To the extent that the speaker has assembled these facts with the purposefully one-sided intent to support a pre-existing position, rationalisation is involved. Few speakers acted as if they had not already decided their vote.

The prevalence of these roles for accounting in this case does suggest that it is a necessary function for this information. However, this needs to be discounted by the circumstances of this particular case. The unprecedented nature of direct involvement by the government in the private sector created a difficult political issue for the government As such, the debate rested upon first principles held by elected officials. Nonetheless, the need to mask these dispositions and to make them appear to be rational existed on both sides. Hence, rationalisation or legitimation appears to be a dominant motif. Secondly, the historical context of the case allows us to read the conclusion and outcome into the events in a way that would not be possible for situations that have not yet run their course. In such a vantage, the rationalisation role of accounting may appear more conclusive.

The dominance of a particular type of "machine" requires a closer look at its principal mechanism. Cost/benefit analysis, when applied to the Chrysler case could be seen to have produced indeterminate results. However, this may be exactly what it is intended to do. The inability to precisely measure tended to collapse attempts into a generalised preference for doing something or one for inaction. Those that favoured the former would accept any cost to avoid a world without Chrysler. This included the subsidisation of the owners of a private venture. Those with a taste for the latter magnified the costs of intervention. Benefits, even in the most human terms of preserved jobs for minority groups, were downplayed.

Support for Chrysler was typically hedged by "best-of-two-evils" reluctance. Everyone would have preferred a strong company making money for its shareholders, repaying its debts and providing employment to many. Unable to have this, decisions makers made mysterious yet explicit references to cost benefit procedures that would render what would be the next best alternative. Even the most strident supporters indicated that they were not pleased with having to make the decision in favour of governmental assistance. All saw a slippery slope that would drag the public sector into a precarious place of marketplace risk.

Those who saw the scales balanced in the other direction also referred to the pseudo-science of costs and benefits. Here, a well defined philosophy weighed in heavily as a cost if a breach had to be considered. Their rhetoric suggested that in addition to this philosophic cost, the prospects of furthering inefficiency and calling forth unwanted production would weigh heavily. Unlike bailout advocates, the position of these individuals was not equivocal. Instead of picking the lesser of two evils, proponents of this position saw no credible rationale or benefits that would outweigh propping up the Chrysler Corporation. On balance, this was a more technically accurate reading of the accounting information.

Following Schumpeter's dictate of the creative destruction role of capitalism, the typical response to massive financial trouble is bankruptcy and reorganisation. A Chrysler reorganisation under Chapter 11 of the Bankruptcy Code would be one of monstrous proportion, but eventually would have sorted out the property rights in a satisfactory manner. So, what information caused such a major departure from the logic of the market? Embedded within many elements of the scenario is the special nature of the automotive product. Unlike other purchases, a car puts the consumer into a long run relationship with the manufacturer. A very important part of the transaction for the car buyer is the faith that warranties will be honoured and that parts will be available. A Chrysler bankruptcy would erode this confidence, no matter what set of countervailing assurances were given (Time, 1979b). The 71 per cent leap of sales in the 10 days following the loan guarantee legislation provides circumstantial evidence for this proposition. More generally, that which is special about automobile industry is its very long supply chain. Pulling out a manufacturer radiates consequences backward toward many component part and raw material providers, and forward to the sales network of dealers and other major users.

In many ways, the quarter century that has elapsed since the Chrysler bailout has not changed much. Government still does not have many good options. Chrysler, by virtue of its size alone, was considered such a special case that quasi-reorganisations tailor-made for this case were proposed. More imaginative and flexible approaches than what the bankruptcy laws contemplated could have been designed and applied. A selective yet fundamental reorganisation would have been a middle ground between bailout and bankruptcy. Mutual sacrifices for all interested parties could thereby be coerced in the name of putting the company back on firm footing. But nothing like this was done then or since. Perhaps we lack the compelling accounting data to support such a resolution.

The worst possible scenario that was not faced was a bailout that did not work. If Chrysler was given loan guarantees yet failed despite them, a large transfer of public money to private parties would have had to occur. In addition, all the adversity the government sought to prevent, would have been visited upon the economy. Despite a substantial possibility that this downside could materialise, very little attention to it was commanded in the debate. Much depended upon that which could not be known at the time. Perhaps the accounting information that was used, or that could have been used, would have not sustained such a leap of faith. That might be why we have politics.

In addition to its historical appeal, the Chrysler bailout episode instructs about the role of financial information in a corporate collapse. The marketplace is not the court of final appeal as long as government intervention powers exist. The Chrysler case provides some insight about the importance of the financial in this domain. On balance, accounting information is not at the centre of the universe. If anything, accounting information enters into the fray as a powerful part of the rhetoric. Although there are several avenues whereby expansions of the information set and improvements in measurement might lead to more influence, there is also a nonpenetrable core where no systematic information seems consequential.