Agile Strategies

General Motors and the Chevy Cobalt Ignition Switch Crisis 1

General Motors and the Chevy Cobalt Ignition Switch Crisis

Randall D. Harris, Texas A&M University-Corpus Christi

W. Scott Sherman, Texas A&M University-Corpus Christi

“They are struggling,” said the letter in Mary Barra’s hands, “They need closure.”1

It was Saturday, March 29, 2014, and Barra, Chief Executive Officer of General Motors (GM), was reading an invitation to meet with the families of Chevy Cobalt crash victims in Washington, D. C. the following week. The timing could not have been worse. Barra was scheduled to appear before Congress on Tuesday, April 1st, to answer questions about GM vehicle safety. Just the day before the letter had arrived, on Friday, March 28th, GM had announced a third Chevy Cobalt recall. In total, 4.8 million vehicles worldwide had been recalled in connection with an ignition switch defect.2 The Chevy Cobalt, positioned as a small and inexpensive car, had been marketed and sold to many young, first-time buyers and their parents. There had been fatalities.

Barra read more of the letter:

For a moment, Ms. Barra, and respectfully, imagine this happening to your family. One morning, while driving your children to school a GM Cobalt loses power and suddenly drifts into your lane resulting in a head on impact. Imagine your heartache, your pain, your fierce confusion when you begin to understand that GM, an American icon, did this to you and your children and did it intentionally, knowingly and with a sense of corporate entitlement that is stunning in its enormity.3

It was Mary Barra’s 10th week on the job. Now Barra, the first female Chief Executive Officer (CEO) of a major automobile manufacturer, was in the middle of a crisis. Barra had called the GM Chairman of the Board, Tim Solso, and told him about the additional recalls. Solso replied that GM now had “a huge problem.”4 While other vehicle models had been affected, the problem appeared to be centered on the Chevrolet Cobalt. There had been problems associated with the 2005 Cobalt, in particular, almost from the time that they were released. Customers had initially complained about the Cobalt after the first 2005 models were sold. Yet the problem was never fully resolved. What had gone wrong?

----------------------------- Copyright © 2017 by the Case Research Journal and by Randall D. Harris and W. Scott Sherman. This case was prepared for the basis of class discussion rather than to illustrate either effective or ineffective handling of the situation. All materials in this case were drawn from publicly available sources. The authors would like to thank the Editor and reviewers of the Case Research Journal for their helpful comments. An earlier version of the case was presented at the 2015 North American Case Research Association conference in Orlando, FL.

NA0499

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

2 Case Research Journal  Volume 37  Issue 4  Fall 2017

GM had initially recalled model year 2005 to 2007 Chevy Cobalt vehicles beginning

in January, 2014. On February 7, 2014, only days after Mary Barra had become CEO, GM had informed the National Highway Traffic Safety Administration (NHTSA) that it had determined that there was a problem with the 2005 to 2007 model years of the Cobalt.5 In its report to the NHTSA, GM had stated that the problem appeared to be centered on the vehicles’ ignition switch. The ignition switch’s “torque performance” on these vehicles had not met GM’s engineering specifications. The switches were easily jostled, and when jostled, the switches would sometimes move out of the “Run” position to “Accessory”, causing the vehicle to stall. A number of fixes had been tried to keep the ignition switches from inadvertently moving out of the “Run” position, including a change in the design of these vehicle’s ignition keys, but these efforts had been ineffective in addressing the issue. GM further explained that, depending on the timing of the switch moving out of the “Run” position, the airbags would not deploy.6 A vehicle stall, combined with the vehicle’s airbags not deploying, had been a lethal combination.

Following the first Cobalt recall in January 2014 and GM’s NHTSA report in February, events had escalated rapidly. There had been a second recall in February of additional vehicles that contained the faulty ignition switch. Lawyers in nine states representing Chevy Cobalt accident victims had now organized into a class action lawsuit covering accident victims throughout the United States. The U.S. Justice Department had begun an inquiry, and was investigating whether there had been a criminal cover-up.7 The NHTSA had also announced that it was conducting their own investigation, opening up the possibility for fines and penalties. There were multiple articles daily in most leading U.S. newspapers and on television regarding the recalls, and social media sites were commenting continuously about GM’s shortcomings. There had been several interviews with the families of Chevy Cobalt crash victims in the media. The mother of one Cobalt crash victim said, “I lost my son for a measly part. My son was my world…and they just took him away from me.”8

Adding to the immense pressure, Mary Barra had been summoned to appear before the U.S. Congress on April 1st in Washington, D. C. Representatives from the NHTSA had also been summoned to appear with Barra. The NHTSA, GM’s regulatory agency, had been solidly and publicly criticized regarding their perceived lack of oversight, and were also facing a sharp backlash from Congress and the general public. It was probable that the NHTSA would deal with GM harshly. Announcing a third recall days before a congressional hearing was devastating. Now, there was this letter in her hands. Before Barra was scheduled to face a withering Congressional investigation, the families of Chevy Cobalt crash victims wanted to meet with her, face to face.

Mary Barra looked back at the letter and thought hard about what to do next. She was due in front of Congress in three days. (Exhibit A describes the timeline of events in this case study).

MARY BARRA – GM CHIEF EXECUTIVE OFFICER

When asked whether she had ever been discriminated against as a female manager, Mary Barra replied, “Just don’t go there.”9 Born in 1961 and raised in Michigan, Barra was a child of General Motors. Her father, Ray Makela, was a journeyman die maker at the GM plant in Pontiac, Michigan, where he worked for 39 years. A career-long member of the United Auto Workers Union, neither Ray nor his wife, Eva, went to

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 3

college. “My whole life growing up, I can’t remember my father ever not being at work,” said Barra, “He worked a lot of overtime.”10 As a child, Mary’s parents strongly encouraged education, and she always did well in school. “I liked math and science, and they encouraged me to pursue that,” she said.11 (Exhibit B presents a biographical sketch of Mary Barra).

After high school graduation, Barra committed to attend Michigan State University, but switched after a friend told her about General Motors Institute, or GMI (later renamed Kettering University). Barra was surprised to discover that she could work her way through school at GMI. “I wasn’t sure how I was going to pay for school,” she said, “My parents had saved for the first year, but GMI seemed very attractive to me because I could pay my own way.”12 Female students at the school made up about 32% of GMI entering classes in the 1980’s, when Barra enrolled. At GMI, Barra chose electrical engineering as her major. “I picked electrical engineering because a lot in the electrical world is based on math,” said Barra.13 While at GMI, Mary began dating her future husband, Tony Barra.14

Following graduation from GMI and a stint in the MBA Program at Stanford University, Barra advanced rapidly inside the GM organization. Her big break came in 2003, when she was tapped to become the plant manager at the Detroit Hamtramck Assembly Plant. Running a GM plant was considered a testing ground for GM top management. Gary Cowger, the President of GM North America at the time, considered the plant manager job to be the toughest job that Barra had taken up to that point in her career. Hamtramck was one of GM’s largest and most complex assembly plants. “We really threw her off the deep end with that one,” he said.15

Barra continued to excel in a number of key executive roles inside the company, and by 2013 had been identified as one of three finalists to take the position as CEO. When Dan Akerson announced his retirement on December 10, 2013, Barra was named as the new CEO. Commenting on her appointment, Akerson stated that Barra was chosen because of her qualifications and that she faced a very difficult job. “Mary will be viewed as an evolution of the company and I think over the next five, ten, twenty years you will see more women and minorities in these slots,” said Akerson, “She is a pioneer and she will do extremely well.”16

GENERAL MOTORS AND THE DELTA PRODUCT LINE

General Motors, headquartered in Detroit, Michigan, was one of the largest multinational corporations in the world. Founded in 1908 as the General Motors Company by William C. Durant, the company had been in business for over a century.17 In 2013, GM manufactured more than 9.7 million vehicles worldwide, producing revenues of $155.4 billion and a net income of $3.8 billion.18 General Motors employed more than 210,000 employees in 396 facilities. The company sold automobiles in over 30 countries under thirteen different brands, including the Chevrolet, Buick, GMC and Cadillac brand names. GM also had 21,000 dealers worldwide.

In the early 2000s, GM decided to develop a new line of small cars to complement its lineup of larger cars and trucks. This new line, or the “Delta platform”, included the Chevrolet Cobalt, Saturn Ion, the Chevrolet HHR, and the Pontiac G5. All of these models shared certain common parts and basic architectural design elements. A common ignition switch, for example, was used in a number of these car lines.19 Lori Queen, who was the Vehicle Line Executive for small cars starting in 2001, said that the Cobalt was “meant to be a statement that GM was getting back into the small car

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

4 Case Research Journal  Volume 37  Issue 4  Fall 2017

game.”20 The Cobalt was designed to compete against foreign models such as the Toyota Corolla and the Honda Civic. GM employees said that the Cobalt and the other cars in this small car line were “cost conscious vehicles” that were “produced on slim margins”.21

While the Cobalt was under development, it was occurring against a backdrop of declining fortunes for GM and for the U.S. automobile industry. All three U.S. auto makers faced financial problems in the early 2000s. From 2001 to 2007, GM’s global market share declined from 15% to 13.3%.22 Profitability at GM was elusive. In 2005, GM posted a consolidated net loss of $10.6 billion. One of the main effects on GM during this time was a concerted effort to cut costs. GM announced a major cost- cutting effort in November 2005. GM cut production, pressured their suppliers to lower costs, reduced health care and pension spending, and reduced their workforce.23 GM leaned particularly hard on their component suppliers by setting cost-cutting targets for individual parts. The use of common vehicle architectures was another way that GM attempted to reduce costs. It was against this stressed and cost conscious environment that the Chevrolet Cobalt was designed and built.

DESIGN AND APPROVAL OF THE COBALT IGNITION SWITCH

The ignition switch that eventually made its way into the Chevy Cobalt and other similar vehicles was called a discrete logic ignition switch (DLIS). GM Project Engineer Tom Utter first designed the DLIS in 1997. The design schematic, called the Component Technical Specification, laid out the technical specifications for the DLIS. One of the key technical aspects of the design was the specification of a force displacement curve of 20 Newton-Centimeters (N-cm) as the torque needed to turn the ignition switch. (Torque was the amount of rotational force needed to move the ignition switch from one setting on the switch to another setting.) This specification was a design specification only. After the actual ignition switch was manufactured by a GM supplier, the actual force displacement of the switch would be furnished by the component manufacturer subject to GM approval.24

The ignition switch was a vital component in the Chevy Cobalt’s operation. When a key was inserted into the ignition switch, the driver would rotate the key from the off position into one of several other positions to operate the vehicle: accessory, run and crank. Components within the ignition switch regulated the amount of effort, or torque, which was needed to rotate the switch from one position to another. A small metal plunger cap with a coiled spring inside, called a detent plunger, would sit in a small groove, or “detent” on a plastic disc inside the switch, and hold the switch in the off, accessory, run or crank positions. (Exhibit C shows a cutaway view of the Cobalt ignition switch with the detent plunger resting inside a detent). A critical issue in the design of the switch regarded the amount of force (torque) that was needed to rotate the switch into the various positions, and whether the switch would stay in the correct position while the vehicle was being operated. One way to increase the torque needed to rotate the ignition switch was to use a longer and more tightly coiled spring inside the detent plunger.25

The original design engineer, Tom Utter, transferred responsibility for the ignition switch specifications to Calvin Wolf, a Design Release Engineer, in 1998.26 Wolf stated that he passed off responsibility for the ignition switch specifications to engineer Ray DeGiorgio in September or October 1999.27 DeGiorgio stated that he did not recall taking over responsibility for the ignition switch design that early.28 Wolf recalled a meeting in September 1999 with DeGiorgio and representatives for the initial ignition

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 5

switch manufacturer, Eaton Corporation. Preliminary tests prior to the meeting on an early prototype of the ignition switch had showed that the switch was not meeting the rotational torque values in GM’s technical specifications.29

On March 22, 2001, Ray DeGiorgio “finalized” the specification for the ignition switch. This specification was to signal to GM’s suppliers for the ignition switch that GM did not anticipate making any additional changes to the ignition switch. The 2001 specification also finalized the force displacement curve, or torque, for the switch at 20 Newton-centimeters. Further, DeGiorgio added the specification that “Torque Curve allowable tolerance shall not exceed +/- 5 N-cm. Thus, the final allowable tolerances for rotational torque on the ignition switch was between 15 N-cm and 25 N-cm. Eaton Corporation, the original supplier for the ignition switch, was acquired by Delphi Mechatronics (Delphi) in 2001.30

GM and its supplier, Delphi, ran into problems with the ignition switch almost immediately during the development and testing phase of the production process. GM discovered in prototype testing during the summer of 2001 that the “Run” detent in the ignition switch was not well defined and that this allowed the key to settle somewhere between “Run” and “Accessory” when the car was started.31 While this problem was solved with a production tool change, Ray DeGiorgio stated that the ignition switch’s electrical design had “failed miserably” and that he essentially had to redesign the switch’s electrical concept.32 The ignition switch simply did not work when it moved from design into production.33 One GM report from 2001 noted that other subsystems in the car would lose power when the ignition circuit lost power, including the anti-lock brake system and the vehicle’s airbags.34

Also in 2001, DeGiorgio noticed problems with the ignition switch prototypes. DeGiorgio wrote in September 2001 that he recently learned that 10 of the 12 prototype switches supplied by Delphi “failed to meet engineering requirements” and that the “failure is significant.”35 Validation testing by Delphi in late 2001 also concluded that the ignition switch failed to meet the specified torque values (20 N-cm +/- 5 N-cm).36 In other words, the switch was “soft” and too easily moved out of position.

In February 2002, Delphi product engineer Erik Mattson e-mailed Ray DeGiorgio and others in February 2002 that the “Accessory” detent on the ignition switch was at “9.5 N-cm”, which was far below the design specifications.37 Mattson said that the torque in the switch “can be increased” but noted that such a change would risk triggering other issues.38 These risks included cracking of the rotor and impacts on the switch’s electrical functions, among other things. Mattson stated that the cost to make the changes was nominal, but said that it would take time to test and validate any changes to the ignition switch at that point. DeGiorgio replied, “If increasing the detent “Accessory” force by 5 N-cm will destroy this switch then do nothing…maintain present course. Under no circumstances do we want to compromise the electrical performance of this switch nor (approval) status.”39 DeGiorgio signed the e-mail after some further discussion, “Ray (tired of the switch from hell) De Giorgio.”40

GM approved shipment of the ignition switch through its normal parts approval process on May 3, 2002.41 However, a documentation package for the ignition switch approval was missing.42 GM’s Global Supplier Quality Manual stated that the onus was on the supplier to maintain the appropriate documentation. Delphi was unable to locate the ignition switch approval documentation. The process leading up to the approval on the Cobalt ignition switch was also in question. DeGiorgio stated that he had the authority to approve the Cobalt ignition switch without needing anyone else to sign off on the approval decision. Differences of opinion persisted among other

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

6 Case Research Journal  Volume 37  Issue 4  Fall 2017

employees whether GM required someone else to approve the switch and/or any deviations from parts specifications.43

CHEVROLET COBALT GOES INTO PRODUCTION

The Chevy Cobalt went into production in August 2004. GM tested the Cobalt prior to full production by having GM employees drive 80 to 90 pre-production vehicles, as it usually did with new designs. None of the GM employees involved with this pre- production testing reported any problems with the Cobalt.44 Joseph Taylor, the Program Quality Manager for the Cobalt, did not recall any official GM reports of ignition switch issues with the Cobalt, either. Mr. Taylor drove one of the 2005 Cobalt test vehicles and personally experienced moving stalls in the vehicle.45 Taylor estimated that the Cobalt he drove stalled three times during pre-production testing as a result of the ignition switch being inadvertently turned off.46 Taylor did not report the stalling incidents because he did not consider them to be significant, and he did not regard these stalls as a safety issue.47

Closely following the launch of the Cobalt, reports began to occur of moving stalls in the car. A journalist informed the Cobalt Chief Engineer at a press event in mid to late 2004 that he had turned off the car by hitting his knee against the key fob or key chain.48 The Chief Engineer followed up on this complaint by asking the GM Engineering Team to try and replicate this incident and determine a fix. The engineers regarded the incident at the time as an isolated event with no safety implications. Word of this report also reached DeGiorgio. DeGiorgio did not respond to internal GM e- mail inquiries.49 (Exhibit D shows a Chevrolet Cobalt Driver View).

GM opened a Problem Resolution and Tracking System (PRTS) report on November 9, 2004 to address the complaint from the press conference that the Cobalt could be keyed off with the knee while driving. The problem was described in the report as Level 3 (on a scale of 1 as most severe and 4 as least severe).50 The engineers who were assigned to work on the PRTS report came up with a number of solutions, several of which were considered and ruled out. One option that was ruled out was to change the design of the key head from a slot to a hole. Another option that was ruled out was to change the location of the ignition switch on the steering column to a higher mount. GM Engineer David Trush noted that the location change of the ignition switch was never seriously considered because it didn’t address the torque issue in the ignition switch and it would also have been an expensive fix.51

The PRTS was closed with “no action” on March 9, 2005, a few days after one of two committees working on the PRTS report had met. The closed PRTS report stated that it was closed pursuant to the directive of the Cobalt Program Engineering Manager.52 The Cobalt Program Engineering Manager at the time, however, stated that he did not have the authority to close the PRTS without action, and suggested that other senior Cobalt program team members had made the decision.53 One of these program team members said that the Program Engineering Manager did have the authority because the issue was a security level 3. The two other senior program team members did not recall closing the PRTS.54 The GM policy at that time was that the committee as a whole had to approve the closure of the PRTS report when no action was taken. The PRTS report also indicated that the committee did not regard the ignition switch stalls as a safety issue.

The PRTS closure report stated that:

All possible solutions were presented in (the two GM committees in charge of reviewing the PRTS), and that (a) the lead-time for all of the proposed

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 7

solutions is too long, (b) the tooling cost and piece price are too high, and (c) none of the solutions seems to fully countermeasure the possibility of the key being turned (ignition turn off) during driving. Thus none of the solutions represents an acceptable business case.55

Customer Complaints Intensify

GM received multiple reports of moving stalls in the Cobalt by May of 2005, and GM dealers were receiving buyback requests from dis-satisfied customers.56 The National Highway Traffic Safety Administration also began investigating reports about the Cobalt. A second PRTS report on the issue was opened on May 17, and this resulted in another round of meetings by the two committees responsible for the issue. Engineers working on the problem focused on two potential solutions: a short-term fix of changing the key head design to prevent customers from putting too many keys on their key ring, and a longer term fix of improving or replacing the ignition switch.57

Ray DeGiorgio was involved in the internal GM meetings working on the moving stall issue, and he also worked closely with Delphi, GM’s parts supplier. Engineers at Delphi described the problem in one e-mail as follows: “Cobalt is blowing up in their face in regards to turning the car off with the driver’s knee.”58 Following positive tests regarding the key head design change, GM approved this fix for existing customers. The basic premise of the key change was to lighten the weight on the key chain, and thus prevent the ignition switch from being jostled out of position.

A GM committee in charge of engineering oversight of the Cobalt rejected a proposal in September 2005 to replace the Cobalt ignition switch with another ignition switch model, an approach that Ray DeGiorgio had been advocating.59 In an e-mail exchange that included Ray DeGiorgio (among others) on September 28, 2005, an executive at GM explained the rationale for the decision:

I wanted to close the loop on the…attempt to bring a new ignition switch to the…vehicles. The con of the change is that the piece cost of the ignition switch went up around $0.90 and would require $400 (thousand) in tooling…The warranty offset for the new switch is in the $0.10 - $0.15 range. It was felt by the Delta team that the revision of the slot in the ignition key to a hole would significantly reduce the inadvertent offs and make any additional changes unnecessary. Consequently, the ignition switch…will remain…until the piece cost hit can be eliminated or significantly reduced.60

In other words, the cost of the replacement ignition switch would be an additional 90 cents per car, but the savings on warranty claims for the vehicles fitted with the new ignition switch was estimated at 10 to 15 cents.

GM issued a technical service bulletin in December 2005 to their dealer network explaining that they now had to install an insert into the key for the Cobalt and several other models to address the “inadvertent turning of key cylinder”. The bulletin also advised drivers to remove unessential items from their key chain. (Exhibit E presents the December 2005 GM Service Bulletin).

Ignition Switch Part Change

DeGiorgio, however, did not let the Cobalt ignition switch issue go. In late 2005 and early 2006, DeGiorgio discussed two changes to the Cobalt ignition switch design with engineers at Delphi, GM’s part supplier. This revised ignition switch design improved the electrical architecture of the switch, and also had an improved detent plunger. The changes to the electrical design of the ignition switch were intended to address electrical

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

8 Case Research Journal  Volume 37  Issue 4  Fall 2017

problems that would cause the vehicle not to start (an issue unrelated to the moving stall problem).

The mechanical change, an improved detent plunger, was intended to increase the torque needed to turn the ignition switch.61 DeGiorgio replaced the original detent plunger in the ignition switch design with a better detent plunger that had a longer and tighter spring. The improved detent plunger stayed in place better, and required the driver to exert greater effort to move the switch between the off, accessory, run and crank positions.

An engineer at GM’s supplier, Delphi, sent Ray DeGiorgio an e-mail in January 2006:

Per our conversation in December, I sent you the 24 Delta Ignition Switch samples for vehicle testing…These switches contain the new PCB (printed circuit board) design and also the stronger… detent plunger.62

Subsequent e-mails between DeGiorgio and Delphi continued to focus on these changes to the ignition switch design. The change in the detent plunger was described in Delphi documents as a request from GM to “be in specification according to the GM spec for torque forces.”63

DeGiorgio approved the redesigned ignition switch for the Cobalt and other models on April 26, 2006 on a Form 3660.64 Each Form 3660 was then linked to a master work order. The master work order for the new ignition switch documented the electrical changes to the switch, but not the change in the detent plunger.65 Also, there was no change in the part number of the ignition switch. An internal Delphi document stated that “Ray DeGiorgio agreed to implement change without changing GM (part number). He provides his approval…to implement both changes.”66 It was GM policy, however, that engineers were required to change a part number for any change that impacted a part’s fit, form or function.67

The changes approved by DeGiorgio were moved into production in 2006. The improved ignition switches were installed beginning with the model year 2007 Cobalts that were currently in production in 2006. These changes effectively solved the problem of low rotational torque in the Cobalt ignition switch and addressed the safety problem in future models of the Chevy Cobalt.68

THE ELUSIVE SEARCH FOR A ROOT CAUSE

GM engineers had worked on the problem of moving stalls in the Cobalt for almost two years by mid-2006.69 GM employees announced there was no safety issue with the Cobalt, and the company’s Product Investigations group had determined that no further investigation was warranted. It was around this time that Cobalt airbag non- deployment litigation began reaching GM’s legal department. The airbags should not have deployed in some of these cases, and in other cases the engineers had no solid technical explanation for why the car’s airbags had not deployed despite extensive analysis.70

Wisconsin Accident

Natasha Weigel and two other teenage girls were returning from a trip to Walmart in a nearby town in Wisconsin on October 24, 2006 when their 2005 Chevrolet Cobalt suddenly lost all power and steering. The car hit a raised driveway and went airborne for almost 40 feet before slamming into a telephone pole and two trees. None of the girls in the car were wearing seatbelts, and the airbags in the car did not deploy. Two

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 9

of the girls died from injuries received in the crash; the third girl survived with brain damage.71 Crash investigators noted that the Cobalt’s ignition switch was in “Accessory” mode rather than “Run”.72 GM was alerted to the crash by a call from a television reporter asking for comment. It was unclear how GM responded to this reporter.73

Outside Investigations

The Wisconsin State Patrol issued a collision report on another accident involving a Chevy Cobalt on February 14, 2007. The Trooper stated in this report that he believed that the vehicle’s ignition switch appeared to have been in accessory mode, and this prevented the airbags in the car from deploying upon impact with some trees at the scene of the accident.74 This report was saved in GM Legal’s electronic files on March 2, 2007.75

The National Highway Transportation Safety Administration (NHTSA) then told GM employees on March 29, 2007 that they had observed a number of airbag non- deployments in Cobalt cars.76 GM participants in that meeting stated that NHTSA did not make a formal request at that meeting, and did not ask them to report back to NHTSA on this matter.77

GM engineer Brian Everest stated that GM personnel were directed by the Product Investigations Group at GM to follow up on the NHTSA inquiry after the group returned to Detroit.78 Another engineer, John Sprague, began compiling a spreadsheet on Cobalt accidents as a result. Everest also stated that the Cobalt airbag non- deployment investigation was transitioned to the Product Investigations Group inside GM in April 2007.79 Some evidence suggested that an engineer named Eric Buddrius might have worked on the investigation at that point, but Buddrius stated that he had no recollection of involvement in the matter.80

The Indiana University Transportation Research Center issued an NHTSA report in April 2007 on the same crash that was studied by the Wisconsin State Patrol. Indiana University researchers noted in the report that the power mode of the Cobalt was recorded as “Accessory”, and that the deputy sheriff who had responded to the scene of the accident had found the ignition switch “jammed” into the “Accessory” position.81 No evidence existed that the Indiana University study was known at GM until 2012.82

GM Goes Into Bankruptcy

GM’s ignition switch supplier, Delphi, had been operating their company while in bankruptcy since October of 2005. GM’s financial strains had also accelerated, particularly in the wake of the global financial crisis of 2008. GM received emergency funding from the U.S. government in late 2008, and also in the first half of 2009.83 On June 1, 2009, GM filed for Chapter 11 bankruptcy, which halted all litigation against the company.84 GM listed assets of $82 billion and liabilities of $172 billion in the bankruptcy filing.85 As part of the bankruptcy filing, GM sold the majority of its assets to the U.S. Treasury, who then became the majority owner of what would now be called the General Motors Company. The total investment of the U.S. government in acquiring GM was roughly $50 billion. Critics of this move referred to the company as “Government Motors”.86

GM’s bankruptcy filing brought major changes to GM. More than 400 of GM’s 1,300 U.S executives were asked to resign or retire. Seven new board members were elected, most of whom were handpicked by the U.S. government.87 GM would emerge from bankruptcy in July 2009, and Delphi in October 2009. Following the trauma of

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

10 Case Research Journal  Volume 37  Issue 4  Fall 2017

bankruptcy and the subsequent restructuring, GM made an initial public offering for the new General Motors Company. Trading in the new GM shares began on November 18, 2010. Approximately 457 million shares of the new GM had been sold in the initial public offering at $33 per share. Following the launch of GM stock trading, the U.S. government sold almost half of its ownership in GM, and recouped $13 billion of their $50 billion investment in the company.88 (Exhibit F shows GM Financial Information from 2009 to 2013. Exhibit G presents GM’s Daily Trading Volume and Closing Stock Price from November 18, 2010 to March 28, 2014).

Critical Clues in the Investigation

Sometime after 11 pm on September 12, 2008, Joseph Harding and his friend, Zachary Shoenback, borrowed their friend’s 2006 Chevy Cobalt SS for a late night fast food run in Baroda, Michigan. There is no clear record for the next several hours, but the two may have gone to a Burger King. Joe, who was driving, did not have a driver’s license.89 The two, traveling at 87 miles per hour, ran off of a rain-slicked road at 3 a.m. the next morning and hit a very large tree. Describing the scene:

Zach was not wearing his seatbelt, Joe did but both still died. Joe’s body was not initially identifiable because his face hit the windshield so hard he broke his neck. The windshield was completely smashed. The medical examiner said the severely broken neck caused Joe’s death. Zach’s body was so badly mangled his leg was sticking through the floorboard and paramedics had to amputate it to get him out and to a hospital. He died of his injuries there.

The Police report says the car went airborne at 3 am, in the rain, and while 19-year old Zach was not legally drunk with tests showing a 0.07 blood alcohol content, Joe was 0.12 and legally drunk.90

Crucially in this accident, the airbags did not deploy upon impact with the tree. Also, due to invalid data in the car’s black box recorder, the State Police report on the crash did not indicate the position of the ignition switch.91 Commenting on the crash, Clarence Ditlow, Executive Director of the Center for Auto Safety, said, “The airbags should still work…The alcohol only goes to the cause of the crash, not whether you survive or not.”92

GM’s Field Performance Assessment engineer, Lisa Stacey, agreed. She thought this was an accident where an airbag deployment should have been expected.93 GM acquired the vehicle involved in this accident and began an analysis. A new PRTS report was opened on the Cobalt in February 2009. A number of meetings ensued, and engineers at GM initially focused on the car’s electrical system and a part called the body control module, or BCM, which was a part of the car’s electrical system.94

At some point during this round of investigations, John Sprague, a GM Field Performance Assessment engineer, noticed that the airbag non-deployment problem didn’t seem to occur in model year 2008 and later Cobalts. He became curious whether some change had occurred around these model years.95 Sprague met with Ray DeGiorgio in the summer of 2009 and they discussed the Cobalt ignition switch. Sprague stated that DeGiorgio told him that the model year 2008 change to the Cobalt ignition switch was a change only to the anti-theft system, and that there had not been a change that would have affected the power mode in the car shutting off.96 Sprague continued to collect data on Cobalt incidents, but was unable to make further progress in his investigation. Sprague handed his investigation off to Product Investigations (PI) investigator Brian Stouffer in August 2011.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 11

GM issued a power steering recall for model year 2005-2010 Cobalts in 2010. Also, litigation regarding the non-deployment of airbags in Cobalts increased at this time, and outside counsel for GM began to warn the company of the potential for punitive damages related to Cobalt litigation.97

The GM Nod

GM lawyers then requested a meeting with the Product Investigations (PI) group in January 2011. This meeting, held in July 2011, was for GM legal to “express to PI that certain Cobalt non-deployment cases represented a safety concern and that PI urgently needed to determine the root cause of the problem.”98

According to one of the lawyers present, it was an “unusual” meeting.99 There were numerous attendees at the meeting from both GM Legal and the GM PI group. While numerous witnesses stated that Douglas Wachtel, Senior Manager of Product Investigations, was present, Wachtel stated that he did not attend the meeting.100 Witnesses at the meeting stated that Wachtel was reluctant to investigate the airbag non-deployment issue because the incident rate was not high.101 Lawyers present at the meeting reported being frustrated by the general response of the PI group. According to persons present at the meeting, Wachtel was directed to put the Cobalt matter through the investigation process. A witness to the meeting stated that Wachtel appeared to understand and accept this order, and agreed to assign an investigator.102 Brian Stouffer was assigned the case. According to Stouffer, Wachtel never told him that the Cobalt investigation was an urgent matter.103 While further meetings were held inside the PI group, the investigation stalled at this point until 2012.104

This pattern of behavior, where GM meetings were held, everyone nodded in agreement to a proposed course of action, but then left the room and did nothing, was labeled by GM insiders as the “GM nod”.105 A similar phenomena, where GM employees would cross their arms in meetings and point outward toward others, indicating that responsibility belonged to someone else, was called the “GM salute.”106

Still can’t identify the Root Cause

In early 2012, the main investigation into the Cobalt was housed in the Product Investigations (PI) Group at GM. The Indiana University study eventually found its way also into the PI group’s investigations. Brian Stouffer (a PI investigator), however, disagreed with the Indiana University study’s conclusions regarding the ignition switch being the Root Cause of the vehicle stalls and airbag non-deployments. This was because the report could not explain a number of the incidents that had been reported where the airbag had not deployed and the ignition switch was recorded in the “Run” position and also why the non-deployments were limited to pre-model year 2008 Cobalts and other similar vehicles.107 Crash investigations in 2012 were also reporting “confusing” data, further hampering progress.108

John Sprague organized a trip to a Michigan junk yard on March 28, 2012, as part of the ongoing investigation. At the time, GM engineers were still working on the theory that there was an electrical root cause to the airbag non-deployments. While assessing a 2006 Cobalt’s electrical system, however, they noticed that the key turned quite easily in the car’s ignition switch.109 Using a fish scale purchased at a nearby bait shop, the team took measurements on the ignition switch and concluded that a driver could potentially hit a pothole and jar the ignition switch out of the “Run” position.110 The research team immediately reported these findings back to the Product Investigations group. A number of meetings discussed this and other findings, but no clear consensus emerged.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

12 Case Research Journal  Volume 37  Issue 4  Fall 2017

John Sprague took Brian Stouffer and others to another junk yard on May 22, 2012 to further test crashed vehicles. With better equipment this time, the group tested approximately 40 Cobalt vehicles. Their results were similar to their previous trip, and the team observed a noticeable dip in torque for model year 2005 and 2006 vehicles.111 However, Stouffer also noticed that there were fairly similar torque values for some vehicles from model year 2007 and 2008, and as a result, could not conclude that the ignition switch had changed in model year 2008.112

Late 2012

GM meetings on the Cobalt continued into late 2012, and the primary discussion centered on the vehicle’s electrical system. In one meeting, however, another solution was discussed, which was to change the ignition switch and increase torque. Brian Stouffer e-mailed Ray DeGiorgio regarding this idea, and asked for a proposal. DeGiorgio replied, and asked what torque value Stouffer wanted.113 Stouffer replied that he didn’t know the desired torque value, but wanted a high-level understanding of what it would take to create a new ignition switch. DeGiorgio replied that for the purposes of a high-level proposal, and not knowing what Stouffer’s requirements were, that he assumed that the new torque value should be 100 N-cm and estimated that the lead time for production would be 18-24 months from issuance of GM purchase order and supplier selection.114 (100 N-cm was five times the torque specification for the original Cobalt ignition switch). After this exchange, there was no request for any follow-up work or information about this estimate.115 After this meeting, GM engineers continued to pursue both electrical and mechanical solutions to the Cobalt airbag non-deployments. GM Legal also continued to handle a steady stream of litigation related to the Chevy Cobalt.

The Red X Investigation

At this point the Red X team was called in. The Red X team were regarded as master problem solvers and experts at determining root cause.116 Brian Stouffer e-mailed the head of Red X’s North America unit on September 6, 2012, to request “…the Red X team to examine changes on the Cobalt between 2007 and 2008 model years.”117 Dan Davis, GM Red X Global Lead, was assigned to the case. Davis took ignition switch torque measurements on five Chevy Cobalts, and concluded that his results were largely consistent with Brian Stouffer’s previous research.118

The next step in the investigation was to examine a crashed Cobalt whose airbag had failed to deploy. According to Davis, he requested such a car from Brian Stouffer, and Stouffer responded that he had to check with GM legal.119 Stouffer stated that he did not recall Davis ever asking him for assistance with obtaining a vehicle, and said that since Davis worked at Milford (a GM testing center), it would not be logical for Davis to ask Stouffer to find him a vehicle.120

Davis said that he eventually followed up with Stouffer, and that Stouffer said that he could not provide a crashed Cobalt because they had been quarantined by GM Legal because of litigation.121 Five separate members of GM Legal stated that they did not tell, and were not aware of anyone else saying to, an engineer at GM that a Cobalt could not be tested, inspected, or examined because it was “quarantined”.122 Stouffer stated that he did not recall telling Davis that all of the Cobalts had been quarantined by GM Legal.123 Davis took no further action, and closed the Red X Cobalt investigation in May 2013.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 13

ANSWERS START TO COME TOGETHER

By February 2013, Brian Stouffer decided that the Cobalt investigation was “stuck”. Internal GM committees were recommending solutions that did not appear to be able to solve the real problem with the cars, the internal Red X investigation had gone nowhere, and none of Stouffer’s work on the problem had, in his opinion, made any real progress.124 Frustrated, the Product Investigations Group requested that that a consultant named Subbaiah Malladi from the consulting firm Exponent be hired to investigate. There were a number of internal GM concerns about this request, including the cost of the consultant and various approvals that were needed, but Malladi was eventually retained and began work in May of 2013.125

There was considerable pressure for Malladi to conduct his investigation. GM Legal was preparing to defend a series of depositions related to Cobalt accidents. Ray DeGiorgio was deposed in April 2013.126 Litigation on a number of Cobalt cases was also underway, and GM lawyers were warning of “a substantial adverse verdict” in at least one of these.127

On a July 30, 2013 conference call, Malladi orally briefed a number of GM executives and also an external lawyer from a firm that was representing GM in one of the ongoing Cobalt litigations. At that meeting, according to one participant in that meeting, Malladi stated that the ignition switches installed in model year 2005 to 2007 Cobalt vehicles did not meet GM’s torque specifications.128 Another participant at that meeting stated that Malladi reported on “the differences in sizes of springs and switches uncovered through x-rays, and asked GM to obtain records from Delphi to confirm that past changes were made.”129 Malladi stated that he had no recollection of the July 30, 2013 conference call.130

Malladi presented his findings again on September 20, 2013. After that meeting, Brian Stouffer contacted Delphi, the parts supplier, to obtain additional information. Delphi provided these documents on October 29th. These documents confirmed that Ray DeGiorgio had approved a change to the spring and detent plunger in the Cobalt ignition switch in April 2006.131 Brian Stouffer presented his conclusions at a GM Investigation Status Review meeting on November 5, 2013. In his conclusions, Stouffer stated that “A review of selected Cobalt…crash events indicates some airbag non-deploys have occurred where the ignition switch was in accessory or off…The hypothesis is that during the off road event the driver’s knee is interacting with the keys and/or the mass of keys is causing the ignition to rotate.”132

After a series of e-mails, the Cobalt ignition switch issue reached the GM Executive Field Action Decision Committee (EFADC) on December 17, 2013. There are no minutes and no notes are taken at EFADC meetings.133 John Calabrese, GM Vice President of Engineering, was dissatisfied, though, with the presentation to the committee and believed that the presentation team had not yet identified a root cause for the airbag non-deployment.134 The December 17 meeting ended without a decision.

The next day, December 18, a GM contract worker ordered 500,000 ignition switches from Delphi, GM’s ignition switch supplier, in a rush order.135 The GM contract worker labeled the order an “Urgent Field Action”.136 The switches were ordered to fix defective ignition switches in the Chevy Cobalt.137

In late December, John Calabrese informed Mary Barra that the EFDAC was discussing a possible recall, and that the committee was awaiting additional analysis before a recall decision could be made. Barra told Calabrese to “get the right data; then do the right thing.”138 This was the first time that Barra had learned of the Cobalt ignition switch issue. At this point, Ms. Barra’s title was Executive Vice President of

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

14 Case Research Journal  Volume 37  Issue 4  Fall 2017

Global Product Development, Purchasing and Supply Chain. Barra had been named as the next GM Chief Executive Officer, but had not yet assumed the role.

BARRA BECOMES CEO

Mary Barra became CEO of GM on January 15, 2014. That same day, Barra held a meeting in Detroit of GM top executives. There were approximately 300 top GM executives in attendance at the meeting.139 After Barra gave a short presentation, she opened up the presentation for questions, and addressed many of the executives present by name. “Being a lifer at GM, it was really great to see the CEO be able to answer all the questions, because she had worked in all the areas of the business, and knew all the people by name,” said one GM executive that was present, “It really felt good.”140

Barra did not have much time to settle into her new role. On January 31, 2014, the EFDAC committee considered additional information about the Cobalt ignition switch. Calabrese concluded at the meeting that the presentation team had sufficiently established a root cause to warrant a recall. The rest of the committee concurred. The initial recall approved by the committee covered the model years 2005 to 2007 Chevy Cobalt.141 It was unclear whether fatalities from Cobalt accidents were discussed at the meeting.142 Mary Barra was informed of the recall later that day.143

On February 7, GM notified the National Highway Traffic Safety Administration (NHTSA), noting that the recall covered approximately 780,000 vehicles at an estimated cost of $32.3 million.144 The recall included all Model Year 2007 Cobalts because the company could not identify when the redesigned switch began to be installed in the cars.145

Top executives, including Mary Barra, discussed Chevy Cobalt fatalities on a conference call in early February.146 Subsequent top-level meetings followed, and executives began to discuss the need to discuss process failures at GM how to correct them. One executive at the time noted that it was difficult to conduct an investigation when there was litigation occurring regarding the issue, and noted a current lawsuit.147 Journalist interest in the problems at GM began to increase, and a number of newspaper articles began to discuss the recall. On February 20th, the New York Time published an article that criticized GM for the limited scope of the January 31st ignition switch recall.148

John Calabrese, GM’s Vice President of Global Engineering, began to dig into the data on the Chevy Cobalt ignition switch issue. Calabrese was troubled by the ignition switch issue and began to request more data.149 Calabrese and others from GM met on Friday, February 21st and over the following weekend. During these meetings, they reviewed the data on the Cobalt ignition switch and realized that the data that they were working with was out of date and incomplete. Calabrese then called for an emergency meeting of the EFADC on February 24th.

At the EFADC meeting on February 24th, GM expanded the recall to an addition 842,000 vehicles, increasing the cost of the ignition switch recall by an estimated $34.3 million.150 Then, on March 17th, CEO Barra announced another recall, this one unconnected to the Cobalt ignitions switch problems. GM recalled 1.7 million vehicles worldwide for a variety of problems, including faulty wiring on side-impact air bags. “Something went very wrong in our processes in this instance, and terrible things happened,” said Barra.151 As a result of the recalls, GM came under intense pressure from government officials, including the NHTSA, to explain the long delay in correcting the Cobalt ignition switch and other safety issues.152

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 15

Congress began a formal investigation. On March 20th, The U.S. House Committee

on Energy and Commerce Subcommittee on Oversight and Investigations scheduled a hearing for April 1st. The meeting, titled “The GM Ignition Switch Recall: Why did it take so long?” was to feature testimony from Mary Barra, as well the top NHTSA Administration official, David Friedman.

GM attempted to communicate with their customers about the recall using conventional means, such as through letters, a call center and the traditional news media. However, social media began to play a critical role in the recall communications. GM customers used social media to vent their frustrations regarding the recall. One customer on GM’s Facebook page stated that her daughter “used every penny she had to purchase her dream car and instead she is stuck with a death trap on wheels.”153 GM had a team of about 20 people located in Detroit to manage the company’s social media presence. “If they don’t respond to their customers, then those customers are going to continue to flame them,” said Roland Rust, a professor of marketing at the University of Maryland.154 GM also maintained a social media presence on Twitter.

Finally, on March 28th, GM announced a third ignition switch recall. In this recall, GM extended their two previous ignition switch recalls by 970,000 vehicles, and adding an additional $39.7 million in estimated costs beyond what had already been allocated. This recall came four days before Mary Barra was scheduled to testify before the Subcommittee hearing on Capitol Hill.

THE LETTER IN MARY BARRA’S HANDS

Barra looked down at the letter in her hands. The words were harsh:

The massiveness and depth of GM’s betrayal can only be understood through the consequence of the true human cost to its customers – customers who trusted GM to make safe cars that would protect their families. Accept their invitation – come and look into the sad and painful eyes of those who must now live with losses that are immeasurable. Please talk with them. Listen to them.155

Barra thought about what she should do. Should she meet with these families, and their attorneys? Barra wondered what GM’s own attorneys would advise her to do. Barra would be meeting with parents like Pam Harding, who had lost her son, Joey, in September 2008. “It destroyed my life,” said Harding, “My son was my world. He was my everything. And they just took him away from me.”156 A majority of the Chevy Cobalt casualties were young adults. The Cobalt had been marketed to them and to first time car buyers. Young drivers tended to be inexperienced. When the Cobalt had stalled, they had panicked.

The problems for Mary Barra and GM went far beyond the question of meeting with the grieving families, and it was mission critical that she get it right. There was a meeting in Washington, D. C. on April 1st. Barra did not expect that Congress would even attempt to be polite. The cameras would be rolling, and the glare of the spotlight would be fixed directly on her. There were powerful Representatives on the subcommittee that Barra would face. One, Fred Upton, R-Michigan, had deep experience with auto regulations, writing the legislation in 2000 that had required GM to report the fatal Cobalt crashes to the NHTSA. “We were very surprised with the revelations that came forward the last couple of weeks,” said Upton, “Everything is on the table.”157

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

16 Case Research Journal  Volume 37  Issue 4  Fall 2017

The Chair of the Subcommittee, Representative Tim Murphy, R-Pennsylvania, was even more direct. “The families who have lost loved ones…have every right to ask whether NHTSA and GM should have detected a trend and acted sooner in the interest of public safety,” Murphy said.158 How should Barra respond to the Senator’s questions? What was the right tone to take? Should she say nothing, stick to the facts, or something else? Did she dare show weakness in front of the cameras, in front of Congress? Further, how should she handle the media, both during and after the hearing? (Exhibit H details the composition of the Subcommittee on Oversight and Investigations, the subcommittee of the U.S. House Committee on Energy and Commerce that would be holding the hearing on April 1st.)

The urgent list didn’t stop with an irate Congress, grieving families and the media onslaught. There was an NHTSA investigation, a U.S. Justice Department criminal investigation, and a legal morass that included a nationwide class action lawsuit. The class action lawsuit, filed in San Francisco federal court, was seeking damages as well as attempting to expand the scope of the recall to another 2.6 million vehicles.159 How should GM handle all of these challenges simultaneously? Longer term, Barra also knew that she had a major problem inside GM. Analysts had pointed at the culture of GM as one of the possible culprits, but first GM had to correct the immediate problems with the Chevy Cobalt. Repairs on the cars were estimated to run into the hundreds of millions of dollars, and GM’s balance sheet would take a major hit. Then there was the customer. Negative publicity had tarnished the reputation of the company. How, in the wake of the negative publicity, was GM going to regain the trust of their customers?

It was Mary Barra’s third month on the job as CEO, and her company was in crisis.

Exhibit A: General Motors Ignition Switch Timeline May, 2002: Ray DeGiorgio approves production of the ignition switch

August, 2004: Chevrolet Cobalt goes into production

September, 2005: GM Committee rejects ignition switch change

January, 2006: Delphi sends DeGiorgio new ignition switches for testing

April 26, 2006: DeGiorgio approves new ignition switch for MY 2007 Cobalt

June 1, 2009: GM files for bankruptcy – litigation against GM stops

November 18, 2010: “New” GM IPO – GM opens for trading

September 6, 2012: The Red X Investigation

November 19, 2013: GM begins internal processes to commence a recall

January 15, 2014: Barra becomes CEO of GM.

January 31, 2014: GM issues first recall of 2005-2007 Chevy Cobalts.

February 24, 2014: GM expands Chevy Cobalt ignition switch recall.

March 20, 2014: Congress schedules GM Hearing for April 1st, 2014.

March 28, 2014: GM issues an additional Chevy Cobalt ignition switch recall.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 17

Exhibit B: Mary Barra Biographical Sketch

Source: Wikimedia Commons Name: Mary Teresa Barra Born: December 24, 1961, Waterford, Michigan. Education: B.S. - Electrical Engineering, Kettering University - 1985 MBA, Stanford University - 1990 Select GM Titles: Plant Manager, Detroit Hamtramck Assembly Plant - 2003 Vice President of Global Manufacturing Engineering - 2008 Vice President of Global Human Resources - 2009 Executive Vice President of Global Product Development - 2011 Executive Vice President of Global Product Development,

Global Purchasing & Supply Chain - 2013 Chief Executive Officer - 2014 Personal: Married to Tony Barra, Two children Favorite Cars: Chevrolet Camaro, Pontiac Firebird Memorable Quote: “No More Crappy Cars” Source: GM, Colby (2015)

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

18 Case Research Journal  Volume 37  Issue 4  Fall 2017

Exhibit C: Chevrolet Cobalt Ignition Switch

Source: HHRG-113-IF02-20140401-SD061

Exhibit D: Chevrolet Cobalt Driver View

Source: HHRG-113-IF02-20140401-SD048-U2

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 19

Exhibit E: GM Service Bulletin, December 2005

Source: HHRG-113-IF02-20140401-SD021

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

20 Case Research Journal  Volume 37  Issue 4  Fall 2017

Exhibit F: General Motors Selected Financial Information December 31, 2009 to December 31, 2013 (Dollars in Millions, except per share & employee data)

At or for the period ended

July 9, 2009*

Dec. 31, 2009

Dec. 31, 2010

Dec. 31, 2011

Dec. 31, 2012

Dec. 31, 2013

Income Statement

Total Net Sales & Revenues 47,115 57,474 135,592 150,276 152,256 155,427

Automotive Cost of Sales 55,814 56,381 118,792 130,386 140,236 134,925

Total Cost & Expenses 63,210 62,402 130,508 144,620 182,619 150,296

Operating Income (Loss) (16,095) (4,928) 5,084 5,656 (30,363) 5,131

Reorganization Gains, Net 128,155 ----- ----- ----- ----- -----

Income Tax Expense (Benefit) (1,166) (1,000) 672 (110) (34,831) 2,127

Net Income 109,118 (4,428) 4,668 7,585 4,859 3,770

Earnings per Share 178.55 (3.58) 2.89 4.58 2.92 2.38

Balance Sheet

Cash & Equivalents 22,679 21,061 15,499 18,422 20,021

Accounts & Notes Receivable 7,518 8,699 9,949 10,395 8,535

Inventories 10,107 12,125 14,324 14,714 14,039

Total Current Assets 59,247 53,053 60,247 69,996 81,501

Property & Equipment, Net 18,687 19,235 22,957 24,196 25,867

Goodwill 30,672 30,513 27,741 1,973 1, 560

Intangible 14,547 11,882 10,013 6,809 5,668

Total Non-Current Assets 77,048 74,913 71,313 79,426 84,843

Total Assets 136,295 138,898 144,603 149,422 166,344

Total Current Liabilities 52,435 47,157 48,932 53,992 62,412

Pensions 27,086 21,894 25,075 27,420 19,483

Total Liabilities 107,340 101,739 105,612 112,422 123,170

TARP – Preferred Shares A 6,998 5,536 5,536 5,536 3,109

TARP – Preferred Shares B ----- 4,855 4,855 4,855 -----

Common Stock 5 15 16 14 15

Add’l Paid In Capital 24,050 24,257 26,391 23,834 28,780

Retained Earnings (4,394) 266 7,183 10,057 13,816

Total Equity 21,957 37,159 38,991 37,000 43,174

Cash Flow

Net cash flow from operating activities

(18,303) 971 6,780 8,166 10,605 12,630

Net cash flow from investing activities

(21,134) 2,012 1,233 (12,740) (3,505) (14,362)

Net cash flow from financing activities

44,229 542 (9,770) (358) (4,741) 3,731

# of Common Stockholders ----- 4 185 276 319 403

Total Number of Employees ----- 209,000 202,000 207,000 213,000 219,000

Source: Mergent Online

*Note: Income Statement and Cash Flow data for July 9, 2009 are from GM’s reorganization. Balance sheet data is reported only for Dec. 31, 2009.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 21

Exhibit G: General Motors (GM) Daily Trading Volume and Stock Price November 18, 2010 to March 28, 2014

Source: Yahoo! Finance

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

22 Case Research Journal  Volume 37  Issue 4  Fall 2017

Exhibit H: 113th Congress, Committee on Energy and Commerce Subcommittee on Oversight and Investigations

Republican Subcommittee Members (in order of seniority)

Murphy, Tim, PA Chairman, Subcommittee on Oversight & Investigations

Burgess, Michael, TX Vice Chairman, Subcommittee on Oversight & Investigations

Blackburn, Marsha, TN Committee Member Gingrey, Phil, GA Committee Member Scalise, Steve, LA Committee Member Harper, Greg, MS Committee Member Olson, Pete, TX Committee Member Gardner, Cory, CO Committee Member Griffith, Morgan, VA Committee Member Johnson, Bill, OH Committee Member Long, Billy, MO Committee Member Ellmers, Renee, NC Committee Member Barton, Joe, TX Committee Member Upton, Fred, MI (ex officio) Chairman, Committee on Energy and Commerce

Democratic Subcommittee Members (in order of seniority)

DeGette, Diana, CO Ranking Member, Subcommittee on Oversight & Investigations

Braley, Bruce, IA Committee Member Lujan, Ben Ray, NM Committee Member Markey, Edward, MA Committee Member Schakowsky, Jan, IL Committee Member Butterfield, G.K., NC Committee Member Castor, Kathy, FL Committee Member Welch, Peter, VT Committee Member Tonko, Paul, NY Committee Member Green, Gene, TX Committee Member Waxman, G, CA (ex officio) Ranking Member,

Committee on Energy and Commerce

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 23

NOTES

1 Hilliard, R. (March 29, 2014). Open letter to GM CEO Mary Barra. Retrieved at www.hmglawfirm.com

2 Bennett, J. (March 29, 2014). “GM recalls more vehicles.” Wall Street Journal. Retrieved at www.wsj.com. Note that additional GM models were affected.

3 Hilliard, R. (March 29, 2014).

4 Colby, L. (2015). Road to Power: How GM’s Mary Barra shattered the Glass Ceiling. Hoboken, NJ: John Wiley & Sons.

5 Committee Majority Staff (2014, March 30). Memorandum regarding hearing on “The GM Ignition Switch Recall: Why did it take so long?” Retrieved at http://energycommerce.house.gov/hearing/”-gm-ignition-switch-recall-why-did-it- take-so-long”

6 Ibid.

7 Colby, L. (2015).

8 Kurth, J. (2014, April 1). “GM’s recalls mount.” Detroit News. Retrieved at www.detroitnews.com.

9 Ibid, p. ix.

10 Ibid, p. 9.

11 Ibid.

12 Ibid, p. 12.

13 Ibid, p. 18.

14 Muller, J. (June 16, 2014). “Exclusive: Inside New CEO Mary Barra’s Urgent Mission to Fix GM.” Forbes. Retrieved at www.forbes.com.

15 Ibid, p. 72.

16 Bennett, J. (Jan 15, 2014). “Former GM CEO says successor faces tough job” Wall Street Journal. Retrieved at www.wsj.com.

17 Sloan, A. 1963. My Years with General Motors. New York, NY: Doubleday.

18 General Motors Company, 2013 Annual Report. Retrieved at: http://www.gm.com/

19 Valukas, A. (2014). p. 18.

20 Jenner & Block interview of Lori Queen, March 17, 2014. All Jenner & Block interview quotations in this case are cited from Valukas (2014).

21 Jenner & Block interview of David Trush, March 17, 2014.

22 GM Annual Reports, 2003 and 2007.

23 Valukas, A. (2014). p. 23.

24 GM Component Technical Specification, Oct. 2, 1997.

25 Valukas, A. (2014). P. 26.

26 Valukas, A. (2014). P. 37.

27 Ibid.

28 Ibid.

29 Ibid, p. 38.

30 Valukas, A. (2014). p. 35.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

24 Case Research Journal  Volume 37  Issue 4  Fall 2017

31 Jenner & Block interview of William Skelton, May 20, 2014.

32 Jenner & Block interview of Ray DeGiorgio, May 7-8, 2014.

33 Valukas, A. (2014). p. 42.

34 Ibid. p. 43.

35 E-mail from Ray DeGiorgio, Sept. 21, 2001.

36 Valukas, A. (2014). p. 45.

37 E-mail from Erik Mattson, Feb. 18, 2002.

38 Ibid.

39 E-mail from Ray DeGiorgio, Feb. 19, 2002.

40 Ibid.

41 GM PPAP report, May 3, 2002.

42 Valukas, A. (2014). p. 51.

43 Ibid. p. 52.

44 Ibid, p. 58.

45 Jenner & Block interview of Joseph Taylor, May 16, 2014.

46 Ibid.

47 Ibid.

48 Jenner & Block interview of Doug Parks, March 18, 2014.

49 Valukas, A. (2014). P. 62.

50 GM PRTS N172404. Nov. 19, 2004.

51 Jenner & Block interview of David Trush, March 17, 2014.

52 PRTS N172404. Nov. 19, 2004.

53 Jenner & Block interview of Gary Altman, March 14, 2014.

54 Valukas, A. (2014). p.68.

55 Ibid. p. 69.

56 Valukas, A. (2014). p. 78

57 Ibid.

58 E-mail from Coniff, J., June 14, 2005.

59 Valukas, A. (2014). p. 90.

60 E-mail from John Handler to various GM employees, September 28, 2005. Valukas, A. (2014). p.18, footnote 31.

61 Valukas, A. (2014). p. 97.

62 E-mail from Alcala, A., Jan. 6, 2006.

63 Delphi Change Request Form, Jan. 15, 2006.

64 Form 3660, April 26, 2006.

65 EWO 302726, Feb. 19, 2004.

66 Form 3660 Cover Memo, May 26, 2006.

67 Valukas, A. (2014). p. 101.

68 Valukas, A. (2014). p. 101.

69 Ibid. p. 102.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 25

70 GM, Settlement Roundtable, Oct. 3, 2006.

71 Lasby, G. (2014, March 22) “General Motors Recall Spurs Lawsuit over Fatal 2006 Wisconsin Accident.” Milwaukee Wisconsin Journal Sentinel.

72 Ibid.

73 Valukas, A. (2014). p. 114.

74 Young, K. Collision Analysis & Reconstruction Report, Feb. 27, 2007.

75 Jenner & Block Interview of Ming-Liang Chi, May 14, 2014.

76 Jenner & Block Interview of Brian Everest, March 27, 2014.

77 Jenner & Block Interview of Gay Kent, March 18, 2014.

78 Jenner & Block Interview of Brian Everest, March 27, 2014.

79 Ibid.

80 Jenner & Block Interview of Eric Buddrius, April 29, 2014.

81 Indiana University Transportation Research Center Investigation, April 25, 2007.

82 Valukas, A. (2014). p. 123.

83 Ibid. p. 131.

84 Ibid.

85 Vlasic, B. (2011). Once upon a car. NY, NY: Harper Collins.

86 Ibid. p. 366.

87 Ibid.

88 Ibid.

89 Meloni, R. (2014, April 1). “The GM Recall: The Pain of Grieving Relatives.” Click on Detroit.com.

90 Ibid.

91 Kurth, J. (2014, April 1). “GM Recalls Mount.” The Detroit News.

92 Ibid.

93 Jenner & Block interview of Lisa Stacey, May 8, 2014.

94 Valukas, A. (2014). p. 135.

95 Jenner & Block interview of John Sprague, May 27, 2014.

96 Jenner & Block interview of John Sprague, March 26, 2014.

97 Valukas, A. (2014). p. 140.

98 Jenner & Block Interview of Brian Everest, March 27, 2014.

99 Jenner & Block Interview of William Kemp, May 9, 2014.

100 Jenner & Block Interview of Douglas Wachtel, May 22, 2014.

101 Jenner & Block Interview of Deborah Nowak-Vanderhoef, May 1, 2014.

102 Jenner & Block Interview of Brian Everest, March 27, 2014.

103 Valukas, A. (2014). p. 153.

104 Valukas, A. (2014). p. 161.

105 Jenner & Block Interview of Mary Barra, April 14, 2014.

106 Jenner & Block Interview of Andrew Brenz, April 22, 2014.

107 Ibid.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

26 Case Research Journal  Volume 37  Issue 4  Fall 2017

108 Ibid, p. 163.

109 Jenner & Block interview of John Dolan, March 21, 2014.

110 Ibid. Note this is also provides a partial explanation for why some accidents recorded the ignition switch in the “Run” position. Cars that were jolted appeared to have their ignition switches jarred out of the run position only seconds before a crash. Because of the jolt that moves the switch, in some Delta platform models, the car’s onboard computer would record the switch in the last known position before the ignition switch cut the power, rather than in “Accessory” or another mode. See the Valukas Report for a detailed discussion of this point.

111 E-mail from Brian Stouffer, May 23, 2012.

112 Jenner & Block Interview of Brian Stouffer, May 9, 2014.

113 E-mail chain, Stouffer & DeGiorgio, Oct. 4-5, 2012.

114 Ibid.

115 Valukas, A. (2014). p. 190.

116 Jenner & Block Interview of Brian Stouffer, April 4, 2014.

117 E-Mail from Dan Davis, Sept. 17, 2012.

118 Jenner & Block Interview of Dan Davis, May 8, 2014.

119 Ibid.

120 Jenner & Block Interview of Brian Stouffer, May 9, 2014.

121 Jenner & Block Interview of Dan Davis, May 8, 2014

122 Valukas, A. (2014). p. 188.

123 Jenner & Block Interview of Brian Stouffer, May 9, 2014.

124 Jenner & Block Interview of Brian Stouffer, April 4, 2014.

125 Valukas, A. (2014). p. 196.

126 Ibid. p. 199.

127 Ibid, p. 204.

128 Jenner & Block Interview of Gay Kent, March 18, 2014.

129 Jenner & Block Interview of William Kent, May 9, 2014.

130 Jenner & Block Interview of Subbaiah Malladi, May 27, 2014. See Valukas, A. (2014) p. 206 footnotes for additional discussion on this point.

131 Jenner & Block Interview of Douglas Wachtel, March 18, 2014.

132 Brain Stouffer Presentation to ISR.

133 Jenner & Block Interview of Alicia Boler-Davis, March 21, 2014.

134 Jenner & Block Interview of John Murawa, April 23, 2014.

135 Colby (2015). p. 109.

136 Bennett, J. (Nov. 10, 2014). “GM ordered new switches long before recall.” Wall Street Journal. Retrieved at www.wsj.com.

137 Ibid.

138 Colby (2015). p. 109.

139 Colby (2015). p. 104.

140 Ibid.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.

General Motors and the Chevy Cobalt Ignition Switch Crisis 27

141 Valukas, A. (2014). p. 223. Note that Pontiac G5s with the same ignition switch were

included in the recall.

142 Ibid.

143 Ibid.

144 Ibid, p. 224.

145 Ibid.

146 Ibid.

147 Jenner & Block interview of Gay Kent, March 18, 2014.

148 Jensen, C. (Feb. 20, 2014). “GM recalls some cars, but not all, with ignition switch problem.” New York Times. Retrieved at www.nytimes.com.

149 Valukas, A. (2014). p. 225.

150 Ibid, p. 226.

151 Vlasic, B. & Jensen, C. (March 17, 2014). “Something went ‘very wrong’ at GM, Chief says.” New York Times. Retrieved at www.nytimes.com.

152 Ibid.

153 Goel, V. (March 23, 2014). “GM uses social media to manage customers and its reputation.” New York Times. Retrieved at www.nytimes.com.

154 Ibid.

155 Hilliard, R. (March 29, 2014).

156 Kuth, J. (April 1, 2014). “GM recalls mount.” Detroit News. Retrieved at www.detroitnews.com.

157 Spangler, T. (March 27, 2014). “Rep. Upton: Why didn’t rules catch GM problem?” Detroit Free Press. Retrieved at http://usat.ly/1mxnl3k.

158 Press Release, Congressman Murphy. (March 14, 2014). “Alert for GM car owners as Murphy launches major safety investigation.” Retrieved at https://murphy.house.gov.

159 PR Newswire. (March 29, 2014). “Lawyers leading consumer class action against General Motors.” Retrieved at http://search.proquest.com.

For the exclusive use of R. Belviy, 2025.

This document is authorized for use only by Ronald Belviy in Copy of Copy of Industry 4.0 GSBA 590_MBA taught by Carsten Zimmermann, University of San Diego from Jul 2025 to Jan 2026.