Volatility and Persistence in the Automobile Industry: Persistence as a Signal of Brand Loyalty.

• Author: Andrews, Anthony Paul

INTRODUCTION

In recent years, there have been several product recalls by automobile firms that previously had stellar reputations in terms of quality control. For example, Toyota, which was known for its production quality control, was viewed by consumers and industry watchers alike as having been able to contest the U.S. automobile market with its attention to quality, design, and customer service. The quality of a particular automobile brand can most easily be determined by the customer's experience. If product recalls tend to signal customers regarding product quality, then are recall shocks persistent or are they short lived? In other words, does a recall impact the sales and/or the market share in the months following a major recall and, if so, for what duration?

Prior research has provided several salient approaches to view the impact of product recalls with respect to consumer behavior. Some approach the impact of recalls on the seller's financial value (Ni et al., 2013), while others consider a behavioral approach based on the misconceptions of consumers for the probability of failure of particular products prior to purchasing the product (Spence, 1977). Nevertheless, others have examined the product recall process with respect to the impact on the firm's financial viability, consumer misperceptions, and brand loyalty (Souiden & Pons, 2009; Mannering et al., 1991). Few studies, however, have examined the impact of product recalls in the automobile industry with respect to market structure and, in particular, the intertemporal impact of product recalls on the market share of the firm. This study addresses the latter point while shedding light on each of the other approaches. In particular, all the research streams that are addressed in prior studies can be unified in the theoretical lens of the signaling theory, which is useful for describing the behavior between the two parties and for reducing information asymmetry between them (Connelly et al., 2011; Spence, 2002). As the signaling theory has been used to generate positive and negative signals between the parties, the concept of signaling was utilized to investigate consumer responses to automobile recalls and to determine if the signal elicits a positive or negative response to recalls. In addition, this study attempts to measure the recall responses as signals with respect to the degree of persistence in a time series context. In other words, this investigation focuses on how a major recall affects, in particular, the market share of an automobile manufacturer over time.[7] By linking signal theory with the market share in a time series [7] Persistence is defined as the impact of a particular shock on a system that does not easily die out (Patterson, 2000, p. 210). context as a model specification, this paper introduces a new approach in the study of product recalls pertaining to the U.S. automobile market. For the context of this analysis, the 2010 Toyota recall was studied because a sufficient period of time has elapsed to examine the intertemporal impact over a time horizon. In this paper, the signaling theory is linked with consumer preference responses to a recall on the basis of their information content for the product quality and brand loyalty. Thus, the firm's signal of product quality and the consumer's information content of brand loyalty should determine the attenuation in sales over time, as demonstrated in the massive recall that was experienced by Toyota in 2010.

A BRIEF HISTORY OF THE 2010 TOYOTA RECALL

The 2010 Toyota crisis began with a fatal car crash in Southern California in 2009 due to a 'sticky gas pedal problem', which caused the car to accelerate. It is believed that Toyota's drive to increase its market share through continuous quality improvement, lean manufacturing, and long-term supplier relations as the 'Toyota Way' led to the company's fast expansion. This forced Toyota to buy more parts from suppliers outside of its supplier association (Andrews et al., 2011). It is believed that this caused several issues, which resulted in the deterioration of product quality. This started to appear in 1999 with additional reported automobile accelerations in the 2002-2004 models. Further quality issues were exposed in October 2007, which resulted in Toyota losing its automatic recommendation from the Consumer Report's magazine and its number one position in the J. D. Power Customer Retention Survey (Train & Winston, 2007). The fatal crash of an off-duty police officer in 2009 ultimately led to Toyota's massive recall. During that period, there were more than 5,000,000 automobiles in the U.S. market alone (Andrews et al., 2011).

Toyota's competitors have also experienced recent and significant recalls. For example, Ford had recalls in 2009 and again in 2014, where the company recalled 1,390,000 of its SUVs in the U.S. market. In February 2014, General Motors recalled over 20 million vehicles with faulty ignition switches. Moreover, due to the penalties paid by Toyota, more auto companies may be addressing recall issues more promptly, as several major automobile companies have since registered recalls with the National Highway Traffic Safety Association (NHTSA). Although these later recalls, especially the recent General Motors recall, may be larger and more significant than the 2010 Toyota recall, sufficient time has not elapsed to examine consumers' response with respect to declining and/or later recovery of sales. However, the time lapse for the 2010 recall has been sufficient to generate consumer responses to the recall and allow for a test of their response to a significant automobile recall. In addition, the adjustment can be examined in the context of intertemporal persistence of a major automobile recall in terms of Toyota's and its major competitors' market share. Thus, this study attempts to measure the impact and persistence of a recall on consumer and firm behavior in the U.S. automobile market in an intertemporal context.

THEORY AND HYPOTHESES DEVELOPMENT

Several studies have investigated consumer response behavior to recalls from many viewpoints. Hoffer et al. (1992) assessed the impact of recall-specific variables in terms of the automobile recall response rates. They determined that U.S. owners of Japanese and European automobiles were significantly less likely to respond to a recall than owners of American vehicles, which is consistent with previous studies. Spence (1977) examined recalls in markets in which the products are sold with regulatory intervention. Spence noted that when insurance is involved, the problem of a moral hazard arises. In another investigation, Souiden and Pons (2009) noted that some researchers (Wynn & Hoffer, 1976; Harrison et al., 1982) determined that recalls did not negatively affect the purchase responses of consumers from the same automobile dealers. However, other researchers (Pletzman, 1985; Weinberger et al., 1981) presented evidence that indicates adverse effects on the sales and market share. Their review also identified several factors that may influence customer responses to recalls. For example, they cite the strength of the brand (or reputation), the structure of a firm's crisis management of a recall, and the media's and/or government's influence on the recall response. Their research also supports the results that indicate that voluntary recalls by particular types of firms may have a positive impact on future sales. Meanwhile, firms that refuse to acknowledge recalls suffer a significant negative impact on future sales.

In a more empirically based context, Mannering et al. (1991) developed an econometric model of brand loyalty on the decline in the market share of American automobiles. They found that the decline in brand loyalty explained 35% of the 3.74% decline in the market share during the period between 1979 and 1989. The remaining loss is attributable to vehicle attributes and socioeconomic changes. Rupp and Taylor (2002) found that the government automobile recall regulatory agency, NHTSA, was more likely to initiate larger, less hazardous recalls that involves older models, financially weak firms, and reported injuries. In addition, they concluded that firms conduct recalls when the benefits (reduced expected liability) exceed the expected cost of the repairs.[8] Barber and Darrough (1996) documented significant and negative market reaction to automobile recall announcements and 'long term direct and indirect costs, with respect to the value of the firm.' More recently, Gokhale (2014) estimated that the 2010 Toyota recall would have resulted in a 19% decline in the company's cumulative abnormal returns.[9] Further, Hartman (1987) used a hedonic pricing model for 1980 automobile sales. Hartman determined that the post-oil embargo period that reflected the introduction of GM's high-quality X-cars had a diminished resale for these cars, but it did not affect the value of other automobiles (cross-product effect).

According to the theory of perceived risk, consumers in non-reversible purchase decisions (e.g. automobile purchases) search for consonant versus dissonant information (Frey, 1986), which points to the fact that consumers are more likely to trust brand strength and/or loyalty as a way to search for information on a particular automobile. Train and Winston (2007) suggested that consumer preferences and, more importantly, brand loyalty is intricately linked to a firm's development, maintenance, and protection of its market share. In addition, they found that GM's loss in the market share during the 1980s could be explained by the '...[increasing] intensity of Americans loyalty for Japanese [automobiles].' They further infer a significant positive relationship between brand loyalty, market share, and market structure. Andrews et al. (2011) determined that Toyota's recent efforts to become the top selling automaker in the global market place might...