Summary
In Kleinsasser, the Court denied the plaintiff's motion to remand because “the average claim value [was] $4,198.25, ” which established that the amount in controversy exceeded the jurisdictional minimum of $5,000,000.
Summary of this case from Assaf v. Progressive Direct Ins. Co.Opinion
CASE NO. C17-5499 BHS
07-19-2018
MARK D. KLEINSASSER, Plaintiff, v. PROGRESSIVE DIRECT INSURANCE COMPANY, et al., Defendants.
ORDER DENYING PLAINTIFF'S MOTION TO REMAND AND MOTION TO STRIKE
This matter comes before the Court on Plaintiff Mark Kleinsasser's ("Plaintiff") motion to remand (Dkt. 14) and motion to strike declaration of Michael Silver (Dkt. 74). The Court has considered the pleadings filed in support of and in opposition to the motions and the remainder of the file and hereby denies the motions for the reasons stated herein.
I. PROCEDURAL HISTORY
On April 1, 2016, Plaintiff filed a class action complaint against Defendants Progressive Direct Insurance Company and Progressive Max Insurance Company (collectively, "Progressive") in Pierce County Superior Court for the State of Washington. Dkt. 1-2 ("Comp."). Plaintiff seeks to recover diminished value on a class-wide basis and individual loss of use damages under the Underinsured Motorists Property Damage ("UIMPD" or "UMPD") provision of his insurance contract with Progressive. Id. Plaintiff alleged that the total amount of compensatory damages would be "approximately $3,010,903." Id.
On June 28, 2017, Progressive removed the matter to this Court. Dkt. 1.
On July 28, 2017, Plaintiff filed a motion to remand challenging both the timing of the removal and Progressive's calculation of damages. Dkt. 14. On August 28, 2017, Progressive responded. Dkt. 24. On September 1, 2017, Plaintiff replied. Dkt. 29. On October 10, 2017, the Court set the matter for an evidentiary hearing and requested additional responses. Dkt. 34. On October 17, 2017, Plaintiff filed a motion for reconsideration. Dkt. 36. On October 19, 2017, the Court denied the motion for reconsideration. Dkt. 37. On October 30, 2017, the Court set the matter for hearing on January 9, 2018. Dkt. 40. On November 10, 2018, Plaintiff filed a motion for reconsideration. Dkt. 41. On November 13, 2017, the Court denied the motion for reconsideration. Dkt. 42. On December 14, 2017, the Court granted the parties' stipulated motion to reset the hearing to February 6, 2018. Dkt. 48.
On January 12, 2018, the parties filed supplemental briefs. Dkts. 50, 58. On January 26, 2018, the parties filed supplemental replies. Dkts. 61, 63.
On January 24, 2018, the Court denied the plaintiffs' motion for class certification in Jenkins v. State Farm Mut. Auto. Ins. Co., C15-5508 BHS, 2018 WL 526993 (W.D. Wash. Jan. 24, 2018). Relevant to the instant matter, the Court concluded the expert report of Dr. Bernard Siskin ("Siskin") was outdated and an improper fit to the facts of the case. Id. at *4.
On February 6 and 7, 2018, the Court held an evidentiary hearing. Dkts. 66, 67. Progressive called Christopher Andreoli ("Andreoli"), Michael Silver ("Silver"), Dr. Michael Salve ("Salve"), and Jonathan Berg ("Berg"), and Plaintiff called Siskin and Darrell Harber ("Harber"). Id.
On March 6, 2018, the parties submitted supplemental briefs and proposed findings of fact and conclusions of law. Dkts. 69, 72, 73.
On March 8, 2018, Plaintiff moved to strike the supplemental declaration of Silver. Dkt. 74. On March 14, 2018, Progressive responded. Dkt. 78. On March 16, 2018, Plaintiff replied. Dkt. 81.
On March 16, 2018, the parties filed supplemental replies. Dkts. 80, 83.
II. FACTUAL BACKGROUND
On September 18, 2015, an uninsured driver hit Plaintiff's vehicle causing significant damage. Comp., ¶ 1.8. The vehicle was towed to a repair shop, and Plaintiff submitted a claim to Progressive. Id. ¶ 6.7. Plaintiff was without the use of his vehicle until November 24, 2015, and, on two separate occasions, he returned the vehicle to the repair shop for additional repairs. Id. ¶ 1.9. On an individual basis, Plaintiff alleges that Progressive failed to provide him with a rental car or otherwise reimburse him for the loss of use of his vehicle. Id. ¶¶ 6.7, 6.11.
Regarding the class claim, Plaintiff alleges that Progressive's failure to compensate its insured for diminished value has been "systematic and continuous" and has affected a large number of insureds over time. Id. ¶ 5.1. As such, Plaintiff seeks certification of a class as follows:
All PROGRESSIVE insureds with Washington policies issued in Washington State, where the insured's vehicle damages were covered under Underinsured Motorist coverage, andId., ¶ 5.3. Plaintiff alleged that "the total amount sought in compensatory damages in this action will be approximately $3,010,903," based on "approximately 2107 claims" and "$1429 per claim on average as the average damages recoverable." Id. ¶ 2.5. In the request for relief section of the complaint, Plaintiff alleges that the class is entitled to "[p]ayment of the difference between the insured vehicles' pre loss fair market values and their projected fair market values as repaired vehicles immediately after the accident." Id., ¶ 7.1(a).
1. the repair estimates on the vehicle (including any supplements) totaled at least $1,000; and
2. the vehicle was no more than six years old (model year plus five years) and had less than 90,000 miles on it at the time of the accident; and
3. the vehicle suffered structural (frame) damage and/or deformed sheet metal and/or required body or paint work.
Excluded from the Class are (a) claims involving leased vehicles or total losses, and (b) the assigned Judge, the Judge's staff and family.
On June 28, 2017, Progressive removed the matter to federal court asserting that the Court has jurisdiction under the Class Action Fairness Act of 2005 ("CAFA"), 28 U.S.C. § 1332(d). Dkt. 1. Progressive conducted an independent search of its database and concluded that the total amount of damages exceeds $5,000,000, which is the jurisdictional minimum under CAFA jurisdiction. Id. ¶¶ 19-31. Progressive submitted two declarations in support of this calculation. Frist, Silver, a master data analyst for Progressive, conducted a review of Progressive's records. Dkt. 2, Declaration of Michael Silver ("Silver Decl."), ¶¶ 1, 3. Silver constructed search criteria that matched his understanding of Plaintiff's class definition and "identified 3,814 claims that fit the search criteria." Id. ¶ 8. Silver, however, "was unable to eliminate claims involving leased vehicles from the proposed class through electronic searches because Progressive does not regularly record this information as part of its claims data." Id. ¶ 9.
Second, Kevin Rehmke ("Rehmke"), a claims supervisor for Progressive, conducted a manual inspection of Progressive's claim files in order to determine the proportion of leased vehicles within the claims identified by Silver. Dkt. 3, Declaration of Kevin Rehmke, ¶¶ 1, 4. Rehmke declares that he did not review a sample of claims within the proposed class because "insureds do not always inform Progressive when they purchase a vehicle they had previously been leasing." Id. ¶ 4. Instead, Rehmke reviewed a sample of claims wherein the vehicle was a total loss. Id. ¶ 5. In total-loss claims, he "could identify whether the vehicle was leased or owned by identifying total loss payments to lienholders and then reviewing the claims notes." Id. ¶ 7. Upon review of 172 total-loss claims, he found that ten claims involved leased vehicles, or 5.8%. Id. ¶ 10. Rehmke concluded that it is reasonable to assume that 5.8% of the claims identified by Silver should be excluded from the class list. Id. ¶ 11.
On July 28, 2017, Plaintiff moved to remand, challenging Progressive's assertion that CAFA jurisdiction existed and submitted evidence in support of his position. Plaintiff's expert, Siskin, reviewed Silver and Rehmke's declarations and claims that they made two specific errors as follows:
a. [They failed to remove] all vehicles that have only non-Class related damage (such as to lights, chrome bumpers, etc., i.e., damage that is not to frame/structure and/or body/paint);Dkt. 14-1 at 7-14, Declaration of Bernard Siskin ("Siskin Decl."), ¶¶ 4-5. After explaining why Rehmke's assumptions and analysis were flawed, Siskin concludes as follows:
b. [They failed to accurately] determine the percentage of leased vehicles because they have used as a starting point a biased and improper sample which dramatically understates the percentage of leased vehicles.
If I apply the expected percentage (9.5%) based upon lease rates in Washington, the Class Size would be 3,452 claims [3,814 claims x 90.5%] which would be $4,932,908 in damages at $1,429/claim.Id. ¶¶ 13-15.
If I apply the percentage I have derived from Mr. Rehmke's flawed sample (by adjusting the population to account for his under sampling of leased vehicles) of 10.3%, the Class size would be 3,421 claims [3,814 claims x 89.7%] which would be $4,888,609.
If I apply the percentage inflation in the lease rate in the class compared to that among non class damaged cars (3.36*5.8% or 19.5%), the Class size would be 3,070 claims [3,814 claims x .80.5%] which would be $4,387,030.
Siskin then discussed flaws in Silver's analysis. Silver assumed that "vehicles with repair estimates of at least $1,000 would have sustained either structural damage, deformed sheet metal, and/or required body or paint work." Silver Decl., ¶ 6. Siskin contends that this assumption is flawed because, based on data from a previous diminished value case, "accidents with over $1,000 in damages can result from damages to only replaceable parts that do not require painting (such as lights, grills, chrome bumpers, etc.), and these type of accidents are not uncommon." Siskin Decl., ¶ 16. For example, a headlight assembly for a 2016 Honda Accord has a list price of $1,140.87. Id. n.5. Accounting for these claims, Siskin further reduced Progressive proposed class size by concluding as follows:
[T]he actual class size would be 3,372 (using the 9.5% figure for leased vehicles) or 3,372 (using the 10.3% figure I have derived from Mr. Rehmke's under-sampled sample) or 2,999 (using the 19.5% figure based on Allstate data on the relationship between the lease rate of class vehicles and non-class damaged vehicles). This would be respectively $4,818,588 (using 9.5% leased vehicles) or $4,77,718 [sic] (using 10.3% leased vehicles) in damages or $4,285,571 (using the 19.5% leased rate).Id. ¶ 18.
In response, Progressive's expert, Salve, pointed out flaws in Siskin's work and provided other conclusions. Dkt. 27, Declaration of Dr. Michael Salve. First, Salve declares that
[a]side from the obvious fact that [Dr. Siskin's] estimates vary by a factor of more than 100% and that Dr. Siskin has not stated which estimate he thinks is correct, I find that the calculation methodologies that underlie all three of these estimates are deficient and misleading and that the estimates themselves are unreliable.Id. ¶ 8. The errors are based on a misinterpretation of a leading car publication, a deficient and misleading attempt to correct for bias in Rehmke's work, and an unsupported comparison of vehicle lease rates from Siskin's work in an earlier Allstate case. Id. ¶ 9-15.
Second, Salve concluded that the average claim would be $1,854.08. Id. ¶ 20. Salve based this number on Siskin's average claim amount in the older Allstate matter, which was $1,386.87, and increased it proportional to the average increase in retail selling prices for cars during the time period between the older case and this case. Id. ¶¶ 16-19.
In reply, Plaintiff submitted additional evidence from Siskin and Harber, a professional in the auto and auto repair industry. Siskin declares that Salve made numerous errors in his critique and in his conclusions. Dkt. 30 at ¶¶ 5-13. Siskin also declares that, in a similar diminished value case, he applied his statistical model for damages to an actual sample and determined that damages were "$590.96 per claim." Id. ¶ 16.
Harber declares that he provides actual diminished value estimates for customers. Dkt. 31, Declaration of Darrell Harber, ¶¶ 4-5. Harber claims that most insurers reject smaller diminished value claims outright, but will pay claims for higher amounts. Id. ¶ 7. Thus, Rehmke's sample based on diminished value claims that were paid is "highly misleading" because the proposed class consists of insureds with smaller, rejected claims. Id. ¶ 8.
To resolve the conflicts in the evidence, the Court held an evidentiary hearing. Progressive first called Andreoli, a claims director for Progressive. Dkts. 75-76, Transcript of Evidentiary Hearing ("Tr.") 25-26. Andreoli reviewed 100 random estimates from the class list. Tr. 32-33. He used the National Automobile Dealers Association's ("NADA") guide to determine the average claim amount of the 100 repair estimates. Tr. 35-36. The NADA guide provides three values for the trade in price of used vehicles, which are a rough value, an average value, and a clean value. Id. If the vehicle suffered frame damage, Andreoli obtained a damage estimate by subtracting the rough value from the clean value. Id. For example, if the vehicle's clean value was $39,050, the vehicle's rough value was $31,675, and the vehicle suffered reportable frame damages, then the diminished value would be $7,375. Tr. 37. Andreoli determined that the average claim value for the 100 claims in the random sample was $4,198.25. Id.
Andreoli also obtained an estimated claim amount based on Siskin's regression model. Tr. 39-41. Andreoli testified that each of the 100 vehicles suffered qualifying damages and that the average claim amount was $1,576. Id.
Progressive then called Silver, Progressive's data analyst. Silver described how he generated a claims list before removal and after removal. He discovered that the original claims list did not include Plaintiff's claim because Progressive paid the claim under collision coverage with a UMPD feature opened in the claim data. Tr. 79-81. On December 4, 2017, Silver determined that there were 4,484 claims that met the class definition as of November 30, 2017. Tr. 78-79.
Progressive then called Salve, an expert in statistical sampling. Tr. 103-5. Salve testified that a sample size of 100 is sufficient to provide a statistically accurate sample across a population of 4,484 claims. Tr. 105-8.
Progressive's last witness was Berg, a claim's manager for Progressive. Tr. 111. Berg's task was to determine the number of leased vehicles in the random sample of 100 claims. Tr. 112. He also explained why some UMPD claims were coded as collision claims with an example as follows:
Mr. Kleinsasser has $10,000 of limits for his UMPD coverage. His damage to his truck exceeded the policy limits. We afforded coverage under collision in order to provide the maximum benefit to Mr. Kleinsasser. We also applied the $100 deductible for UMPD, meaning the lower deductible.Tr. 122.
Plaintiff then called two witnesses, Siskin and Harber. Siskin provided background information on his statistical regression model that was developed for Moeller v. Farmers Ins. Co. of Washington, 173 Wn.2d 264 (2011). Siskin also testified that Progressive's sample of 100 claims is inappropriate because it contains vehicles that are older than vehicles in the class definition. Tr. 152. Siskin explained Plaintiff's ambiguous class definition of "the vehicle was no more than six years old (model year plus five years)" as follows:
The reason is if you read it carefully, the opening statement says the vehicle will be no more than six years old. That statement is confusing. What do you mean by old? What is the age of a car? No birth date for a car. What do you use? You have in parentheses the definition, which explains it. That is commonly what statisticians are used to seeing. We write Greek and explain it in parentheses so somebody might understand it better. It says, "Model year plus five years." That clearly defines what they are talking about. That is, for practical purposes, almost exclusively.Tr. 152:23-153:7. After removing vehicles that were older than the class definition and leased vehicles, Siskin implemented his regression analysis and determined that the average claim amount was $1,158. Tr. 155-157. Siskin also opined that this estimate would be high because the sample included collision claims, which "tend to be higher than the uninsured motorist accidents." Tr. 157. Siskin then testified that the amount should be reduced by 2.31% for overlapping damage. Tr. 162-163. This number was based on larger samples from previous cases. Id.
Regarding lease rates, Siskin provided numerous estimates as to the rate. In the end, he opined that 10.3% is the best estimate. Tr. 172. He also stated that Progressive's lease rate of 5.3% from its sample of 100 random claims was low because the sample size was too small and it didn't account for the fact that the majority of leases are three years or less. Tr. 169-171.
Progressive then questioned Siskin on cross-examination. Siskin admitted that he did not have a copy of the original data upon which he built his regression model. Tr. 192-196. He also stated that he had no idea how counsel obtained the average damage amount of $1,429 in the complaint. Tr. 201.
After questions from the Court, Siskin testified that his model did not account for other types of damages. Specifically, Siskin stated that the model does not include the drop in value of a vehicle when a potential buyer is aware that the vehicle had previously been in an accident. Tr. 223-4. The parties and the Court refer to these damages as stigma damages, which have been described as damages that "occur when the vehicle has been fully restored to its preloss condition, but it carries an intangible taint due to its having been involved in an accident." Moeller, 173 Wn.2d at 271. Siskin also testified that his model did not account for secondary damages, which were described as loose parts or other things wrong with the vehicle as a result of an accident. Tr. 225.
Plaintiff then called Harber as the final witness. Harber testified that he determines "a pre-loss value and a post-loss value" and that "the difference of those two numbers is the loss." Tr. 228:22-23. Harber stated that the value of a vehicle does not necessarily drop to the rough trade-in value if it has been structurally damaged. For example, Harber would assess the first of the 100 random claims as follows:
What I would do is I would consider those areas that are affected, then looking at the NADA guide, I would be closer to an average value, and using the average trade value, that would show a loss in value of $3,600, and in this particular case, then what I would do is I would take a look at what was the cost of repair, which was $1,100. I certainly wouldn't say that the loss in value in this particular vehicle was three times the amount of the cost of repair. I would reduce it even more. I would have been closer to about a $2,400 loss in this particular vehicle as opposed to almost $4,500 if I had used rough. I wouldn't use rough.Tr. 231. Harber also testified that if the vehicle didn't incur structural damage, then the loss in value would only be a "couple hundred dollars." Tr. 231-2.
On cross-examination, Progressive inquired about Harber's report for Plaintiff's vehicle. Based on the severity of the damage, Harber used the rough trade-in value for the post-accident value of Plaintiff's vehicle. Using this number, Harber reached the same estimate for Plaintiff's loss as Andreoli, which was $7,375. Tr. 251-2.
III. DISCUSSION
A. Motion to Strike
Although Plaintiff moved to strike the supplemental declaration of Silver for numerous reasons, his reply narrows the issues to timeliness. See Dkt. 81. Plaintiff argues that because Progressive submitted the declaration after the evidentiary hearing it is untimely and prejudicial. Plaintiff contends that the "potential for prejudice to Plaintiff should be self evident." Id. at 2. If the prejudice is so obvious, Plaintiff should be able to articulate it in his brief. He does not. When considering a contested motion to remand, the "parties may submit evidence outside the complaint, including affidavits or declarations, or other 'summary-judgment-type evidence relevant to the amount in controversy at the time of removal.'" Ibarra v. Manheim Investments, Inc., 775 F.3d 1193, 1197 (9th Cir. 2015) (quoting Singer v. State Farm Mut. Auto. Ins. Co., 116 F.3d 373, 377 (9th Cir. 1997)).
In this case, Progressive has submitted a declaration that is summary-judgment-type evidence with an opening supplemental brief. Plaintiff had three opportunities to contest the substance of the declaration in this motion, the reply, and response to the supplemental brief. While the Court did not expect the submission of additional evidence, Plaintiff's due process rights have been satisfied with notice and an opportunity to contest the additional evidence. Having shown no actual prejudice, Plaintiff has failed to articulate any reason to strike the declaration. Therefore, the Court denies Plaintiff's motion to strike.
B. Construction of the Pleadings
"A defendant generally may remove a civil action if a federal district court would have original jurisdiction over the action." Allen v. Boeing Co.,784 F.3d 625, 628 (9th Cir. 2015). CAFA vests federal district courts with original jurisdiction over class actions involving more than 100 class members, minimal diversity, and at least $5,000,000 in controversy, exclusive of interests and costs. Dart Cherokee Basin Operating Co. v. Owens, 135 S. Ct. 547, 552 (2014) (citing 28 U.S.C. § 1332(d)). A defendant seeking removal under CAFA must file a notice of removal "containing a short and plain statement of the grounds for removal." 28 U.S.C. § 1446(a); see also Dart Cherokee, 135 S. Ct. at 551. The removing party must set "forth, in the removal petition itself, the underlying facts supporting its assertion that the amount in controversy exceeds" the jurisdictional minimum. Gaus v. Miles, Inc., 980 F.2d 564, 567 (9th Cir. 1992). The burden of establishing removal jurisdiction remains on the party seeking removal. Abrego Abrego v. The Dow Chem. Co., 443 F.3d 676, 685 (9th Cir. 2006).
In this case, the parties dispute the scope of the pleadings. First, when Progressive bears the burden of establishing that damages exceed a certain amount, it is placed in the unusual position of arguing that the complaint should be construed in the broadest manner possible. Thus, Progressive argues that the ambiguities in Plaintiff's complaint should be resolved in favor of a larger class size and higher potential damages. For example, Progressive contends that relevant claims are those that it paid under UIMPD coverage and claims paid under "comprehensive/collision claims where an underinsured motorist coverage feature was opened." Tr. 79-80. Plaintiff defined the class as all insureds "where the insured damages were covered under Underinsured Motorist coverage . . . ." Comp., ¶ 5.3. Progressive contends that this phrase is ambiguous when applied to the facts of this case, and the Court agrees. Plaintiff's claim exemplifies the ambiguity. Plaintiff was hit by an underinsured/uninsured motorist and suffered damages that exceeded his UIMPD coverage. In order to fully compensate Plaintiff, Progressive paid full coverage under UIMPD and the remainder under comprehensive/collision, but only required Plaintiff to pay the UIMPD deductible. Plaintiff did not specifically exclude dual coverage claims in his class definition and resolving the ambiguity in favor of Plaintiff being in the class, the Court adopts Progressive's position that dual coverage claims are included in the class.
Regarding vehicle age, the class definition provides in relevant part that "the vehicle was no more than six years old (model year plus five years) . . . ." Id. Although Siskin testified that the phrase "no more than six years old" is "confusing," Tr. 152:25, he claims that "model year plus five . . . clearly defines what" Plaintiff meant, Tr. 153:6-7. The Court agrees to the extent that "model year plus five" is not ambiguous whereas "no more than six years old" is subject to multiple interpretations. For example, a 2010 model year vehicle would qualify for the class for any accident in 2010 through 2015 if the class is restricted to model year plus five. On the other hand, Plaintiff fails to define the starting point for the age description of a vehicle if the class is defined as vehicles no more than six years old. See Tr. 152 ("What do you mean by old? What is the age of a car? No birth date for a car. What do you use?"). Is the starting point the day the vehicle is manufactured or the day the vehicle is sold? If the latter, then a 2010 model vehicle sold in 2012 would qualify for any accident in 2012 through 2018. Therefore, the Court adopts the interpretation of the complaint that defines an ascertainable class and concludes that the class is limited to all vehicles that were model year plus five when the accident occurred.
Second, Plaintiff strenuously and repeatedly objects to Progressive's "arguments and claims [that were] not made in the notice of removal." Dkt. 58 at 3. Plaintiff, however, fails to submit any authority for the proposition that Progressive is bound by the evidence it submitted in support of its notice of removal. Instead, Plaintiff's position is based on the Ninth Circuit's statement that "the circuits have unanimously and repeatedly held that whether remand is proper must be ascertained on the basis of the pleadings at the time of removal." Broadway Grill, Inc. v. Visa Inc., 856 F.3d 1274, 1277 (9th Cir. 2017). Plaintiff argues that the Court should only consider Progressive's pleading, or notice of removal, and evidence in support thereof submitted at the time of removal. Tr. 19-21. Contrary to this argument, the court in Broadway Grill addressed the question of whether post-removal amendments to the pleadings were proper. The court held that "plaintiffs' attempts to amend a complaint after removal to eliminate federal jurisdiction are doomed to failure." Broadway, 856 F.3d at 1277. In this case, the Court declines Plaintiff's invitation to extend the law preventing post-removal amendments to govern evidence obtained and submitted post-removal.
"When, as here, 'a defendant's assertion of the amount in controversy is challenged . . . both sides submit proof and the court decides, by a preponderance of the evidence, whether the amount-in-controversy requirement has been satisfied.'" LaCross v. Knight Transp. Inc., 775 F.3d 1200, 1202 (9th Cir. 2015) (quoting Dart Cherokee, 135 S. Ct. at 554)). "The parties may submit evidence outside the complaint, including affidavits or declarations, or other 'summary-judgment-type evidence relevant to the amount in controversy at the time of removal.'" Ibarra, 775 F.3d at 1197 (quoting Singer, 116 F.3d at 377).
As stated in Ibarra, the issue is the "amount in controversy at the time of removal." Id. As long as due process is satisfied, Plaintiff fails to explain how evidence created or obtained pre-removal is acceptable, but evidence created or obtained post-removal should be precluded. In other words, when the Court considers the amount in controversy at a set time, it makes no difference when the evidence in support of that amount is generated as long as the opposing party has an opportunity to respond to the evidence. For example, Progressive removed on the basis of a class list that did not include Plaintiff. Progressive realized its mistake post-removal and generated a new list that included Plaintiff. Progressive produced the list to Plaintiff and gave Plaintiff ample opportunity to contest the evidence. Plaintiff offers no logical reason why the Court should preclude this post-removal list from consideration of the amount in controversy at the time of removal.
Furthermore, "[i]f the court determines at any time that it lacks subject-matter jurisdiction, the court must dismiss the action." Fed. R. Civ. P. Rule 12(h)(3). The phrase "at any time" has been construed to encompass the trial court proceedings as well as when the matter is on appeal, even if the issue of subject matter jurisdiction is raised for the first time on appeal. Summers v. Interstate Tractor & Equip. Co., 466 F.2d 42, 49 (9th Cir. 1972) ("Although [appellant's] argument is being raised for the first time on appeal, this court will consider the matter because it is claimed that it deals with the subject matter jurisdiction of the federal court and such an issue can be raised at any time.") Under this interpretation of the rule, if there is no temporal limitation on when the issue may be raised, it would seem that there is no temporal limitation on when a party must obtain or generate its evidence in support of jurisdiction. Therefore, the Court declines to impose an arbitrary deadline for evidence created in support of the amount in controversy at the time of removal. Putting aside Plaintiff's objection, the Court will turn to the evidence submitted by the parties.
C. Amount in Controversy
When determining the amount in controversy at the time of removal, "the district court must make findings of jurisdictional fact to which the preponderance standard applies." Dart Cherokee, 135 S. Ct. at 554 (citing H.R.Rep. No. 112-10, p. 16 (2011)). "[W]hen the defendant relies on a chain of reasoning that includes assumptions to satisfy its burden of proof, the chain of reasoning and its underlying assumptions must be reasonable ones." LaCross, 775 F.3d at 1202 (citing Ibarra, 775 F.3d. at 1199). "Under the preponderance of the evidence standard, if the evidence submitted by both sides is balanced, in equipoise, the scales tip against federal-court jurisdiction." Id. at 1199.
In this case, the parties contest the potential size of the class, the average amount of each potential claim, and the amount of awardable attorney's fees.
1. Class Size
Plaintiff asserts that the class size is 3,052, Dkt. 73-1, ¶ 6, and Progressive asserts that the class size is 4,217, Dkt. 69 at 7. Neither of these numbers withstand scrutiny. First, and most obviously erroneous, is Progressive's estimate because it is based on the number of claims "[a]s of November 30, 2017." Dkt. 69 at 3. The only significance of this date is that it is the date Progressive's employee created the list. The Court, however, is tasked with determining the class size as of the date of removal, which was June 28, 2017. Ibarra, 775 F.3d at 1197 ("'evidence relevant to the amount in controversy at the time of removal.'") (quoting Singer, 116 F.3d at 377). Progressive fails to assert any grounds for an extension of law such that a court should consider the number of claims approximately five months after removal. As such, the Court agrees with Plaintiff that Progressive's arguments on this issue fall below the requirements of Fed. R. Civ. P. 11. See Dkt. 83 at 3. In a footnote, Progressive states that, "[o]f the 4,484 claims on the Class List, 4,131 have a date of loss prior to Progressive's removal of this case on June 28, 2017." Dkt. 69 at 3 n.4. Even though Progressive labels these 4,131 claims the "Pre-Removal Claims," id., it submits proposed findings of fact asserting that "the class consists of at least 4,217 members." Dkt. 72, ¶ 34. This is a failure to follow binding law without any explanation or argument for an extension of that law. More importantly, the Court is unable to sua sponte apply the applicable exclusions to the Pre-Removal Claims. Progressive should have started with the Pre-Removal Claims and then excluded leased vehicles, non-owned vehicles, or alleged duplicate claims. Therefore, the Court finds that Progressive has failed to submit a reliable estimate of the class size "at the time of removal." Ibarra, 775 F.3d at 1197.
Second, Plaintiff offers a starting list of 4,217 claims. Dkt. 59, ¶ 5. Siskin declares that this is the number of claims that appeared on a spreadsheet that Progressive produced either on December 7, 2017 or sometime thereafter. Id. Thus, it appears that Plaintiff's starting point (4,217) is Progressive's end point (4,217). If Plaintiff concedes this starting point, the Court will accept it because Progressive's admitted inclusion of post-removal claims is facially erroneous.
Upon review of the list, Siskin excluded numerous vehicles. Siskin's first exclusion was based on the age of the vehicle. As stated above, the Court construes the class to include vehicles that were "model year plus five" on the date of the accident. Siskin discovered that 370 claims needed to be removed from the list because the vehicles exceeded the "model year plus five" requirement. Tr. 159:17. Although Progressive contests Plaintiff's and the Court's restrictive interpretation of the complaint, Progressive did not argue in the alternative or contest Siskin's conclusion that exclusion of 370 vehicles conforms to the "model year plus five" restriction. Thus, the Court accepts this number and reduces the class size to 3,847. Dkt. 59, ¶ 6.
Siskin then accounted for "claims knowable to Progressive" on the date of removal. Siskin declares that the class should be further reduced to 3,529 claims. Id. ¶ 7. Progressive does not contest this reduction based on claims at the time of removal. Thus, the Court accepts this exclusion and reduces the class size to 3,529.
Progressive contends that the Court should reduce the class size by 1% to account for "non-owned" vehicles, which are vehicles that were not owned by the insured at the time of the accident. Dkt. 72, ¶¶ 32-34. Plaintiff does not contest this reduction. Therefore, the Court accepts Progressive's exclusion of 1%.
The next two exclusions are contested by Progressive. Plaintiff excluded from the class "claims involving leased vehicles." Comp., ¶ 5.3. Progressive concedes that the "Class List could not be filtered to exclude claims involving leased vehicles." Dkt. 69 at 3 (citing Dkt. 53, ¶ 8). In support of Plaintiff's original motion to remand, Siskin declared that the lease rate was 9.5%, 10.3%, or 19.5%. Dkt. 14-1 at 12, ¶¶ 13-15. None of these estimates were based on Progressive's claims data. After removal and in preparation for the hearing, Progressive selected "a random sample of 100 class claims to determine how many claims involved vehicles leased at the time of the loss." Dkt. 80 at 6. Progressive determined that 5% of the vehicles in the sample were leased vehicles and contends that the Court should reduce the class size by that amount to account for the exclusion of leased vehicles. Dkt. 72, ¶ 31. At the hearing, Siskin contested this percentage because (1) the sample included six vehicles that were older than model year plus five and (2) Progressive failed to account for the fact that, in his opinion, leases mostly occur in the first three years of ownership. Tr. 166:9-168:10. The Court accepts the former adjustment because the class is limited to model year plus five.
Regarding the latter adjustment, Siskin was not qualified as an expert on the yearly lease rates of vehicles and his adjustment is based on vehicles that are six years old, not model year plus five. See Tr. 168:6-8 ("if we assume [the leases] are evenly distributed, half the [leased] cars should have occurred in the first three, half the [leased] cars in the second."). Therefore, the Court finds that 5.43% is the best evidence of the lease rate across the class because it is based on actual class data and accounts for the class limitations.
Progressive also reviewed the "actual class data to identify the percentage of vehicles in the proposed class that suffered Qualifying Damage." Dkt. 60 at 6. Plaintiff limited the class to vehicles that "suffered structural (frame) damage and/or deformed sheet metal and/or required body or paint work." Comp., ¶ 5.3. Similar to the lease rate issue, Plaintiff relies on data from other cases whereas Progressive relied on a sample of data from this class. The Court finds that Progressive's evidence is the best evidence to determine the qualifying damage rate for this class. Moreover, Salve testified that the sample size was sufficient to accurately extrapolate across the class. Tr. 108:4-11. Therefore, the Court finds that no reduction is necessary for the exclusion of qualifying damage. Dkt. 72, ¶¶ 35-37.
In sum, the Court finds that the class size at the time of removal was 3,529 and must be reduced by 5.3% to account for leased vehicles and 1% to account for non-owned vehicles. This results in a potential class size of 3,307.
2. Claim Amount
The parties have submitted numerous estimates of the average claim value. The Court rejects all of Siskin's estimates as well as any estimate based on his regression model. As set forth in Jenkins, Siskin's model is outdated and a poor fit for current cases. Jenkins, C15-5508 BHS, 2018 WL 526993 at *3-4. The most fatal flaw, however, is that it does not track the contours of Plaintiff's complaint. Plaintiff alleges damages as "[p]ayment of the difference between the insured vehicles' pre loss fair market values and their projected fair market values as repaired vehicles immediately after the accident." Comp., ¶ 7.1(a). This allegation clearly encompasses any loss in value due to an accident. As such, it encompasses both diminished value damages and stigma damages. "[D]iminished value damages may be available when the vehicle cannot be fully restored to its preloss condition, whereas stigma damages may be available when the vehicle can be fully restored to its preloss physical condition, but is perceived as being less valuable due to the accident." Ibrahim v. AIU Ins. Co., 177 Wn. App. 504, 511 (2013). Siskin testified that his model does not account for stigma damages. Tr. 223-4. Thus, any model that fails to account for loss in value regardless of whether the vehicle is restored to its pre-loss condition is irrelevant to the issue of potential damages of this class.
Plaintiff previously argued that he was not seeking stigma damages on behalf of the class. Dkt. 13-1. In doing so, Plaintiff cited his complaint as follows:
Plaintiff claims that, when certain automobiles, i.e. those within the proposed Class, sustain damage to their structural systems and bodies, they cannot be fully repaired to their pre-accident condition and as a result are tangibly different than they were pre-accident....Id. at 11 (citing Comp., ¶ 1.5). Plaintiff, however, fails to include this specific exclusion in his class definition. Plaintiff excludes vehicles based on age, mileage, repair estimates, and types of damage, but fails to exclude vehicles that were fully restored to their pre-loss condition. See Comp., ¶ 5.3. Thus, construing Plaintiff's complaint in the broadest terms possible, Plaintiff seeks to recover any loss in value to any vehicle that suffered damage as a result of an underinsured or uninsured driver. This includes stigma damages as the potential damages in controversy.
Since the areas of repaired damage would be detectable in any later inspection, and [Plaintiff's] vehicle, like those in the proposed Class, could not be fully restored to its pre-loss condition, the vehicle was worth less (it had "diminished value") as a result of the accident.
Furthermore, stigma damages are not precluded by Moeller. Instead, the Washington State Supreme Court has left the door open for trial courts to explore whether stigma damages are recoverable in any particular case. Moeller, 173 Wn.2d at 271 ("Stigma damages are generally disfavored . . . . Undoubtedly, the nature of the damages Moeller claims and how they can be proved will be explored by the trial court should this case proceed to trial."). In the absence of a conclusion that Progressive's policy specifically excludes stigma damages, these damages are potentially in controversy in this case. Therefore, the Court rejects all of Siskin's estimated average claim values and Harber's estimates based on Siskin's regression model.
The other estimated average claim values is from Andreoli. Plaintiff claims Andreoli's method is unreliable because he assumed that any vehicle with structural damage would only be worth the rough, or lowest, trade-in value. Dkt. 73 at 11. The Court agrees that Andreoli may have overestimated some claims, which results in a higher average claim amount. However, "when the defendant relies on a chain of reasoning that includes assumptions to satisfy its burden of proof, the chain of reasoning and its underlying assumptions must be reasonable ones." LaCross, 775 F.3d at 1202 (citing Ibarra, 775 F.3d. at 1199). The Court concludes that, even if Andreoli's method overestimates some individual damages, overall the chain of reasoning and underlying assumptions are reasonable. Moreover, Plaintiff fails to submit any evidence or estimate to correct Andreoli's alleged overestimates. For example, Harber could have submitted his own estimates for some or all of the random 100 claims, and then the Court could have compared the final results. Plaintiff did not do this or anything similar. Instead, the Court is provided one estimate that Andreoli and Harber agree upon, which is Plaintiff's claim for $7,365, and Plaintiff's argument that Harber is more adept at determining which vehicle could be sold for the rough trade-in value versus the average trade-in value. On this record, the Court finds that Progressive has submitted the best and most reasonable evidence because Plaintiff's "[a]rgument without evidence is hollow rhetoric . . . ." Teamsters Local Union No. 117 v. Washington Dep't of Corr., 789 F.3d 979, 994 (9th Cir. 2015). Therefore, the Court finds that the average claim value is $4,198.25.
3. Summary for Amount in Controversy
In sum, Progressive has met its burden to show that the amount in controversy exceeds the jurisdictional minimum of $5,000,000. The Court concludes that the most reasonable number of claims at the time of removal was 3,307. The Court finds that, based on the requested relief in the complaint, the most reasonable number for the average claim amount is $4,198.25. In combination, the average number of claims times the average claim amount equals the potential amount in controversy of $13,883,612.75. This number easily exceeds the jurisdictional minimum, and the Court will not address the question of first impression regarding class-wide attorney's fees under RCW 4.84.015. Therefore, Plaintiff's motion to remand based on lack of jurisdiction is denied.
D. Timing
Plaintiff argues that Progressive's removal was untimely. Dkt. 14 at 13. It is black letter Ninth Circuit law that a defendant's subjective knowledge regarding the potential amount in controversy does not trigger the second thirty-day removal period. Roth v. CHA Hollywood Med. Ctr., L.P., 720 F.3d 1121, 1126 (9th Cir. 2013) ("we are inclined to think that the sentence should be understood in context to mean only that a 'defendant's subjective knowledge cannot convert a non-removable action into a removable one'"); Stone v. GEICO, 18-35502, 2018 WL 3454484, at *2 (9th Cir. July 18, 2018) ("That GEICO did in fact investigate—and may have realized five months before removal that damages could exceed $5 million—is irrelevant."). Giving defendant step-by-step instructions on how to search its database to determine the potential amount in controversy seems irrelevant whether a defendant acts or fails to act because, at most, acting will result in defendant's subjective knowledge of the potential class size. Applied to this case, Plaintiff's argument that Progressive's "counsel had plenty of occasion to inquire reasonably of Progressive how its claims data reflects on the accuracy of class size estimate alleged by Plaintiff in his complaint without the benefit of discovery" is irrelevant. Absent any evidence of objective knowledge, Plaintiff has failed to show Progressive's removal was untimely. In fact, the majority of Plaintiff's newer arguments are based on the position that Progressive didn't generate the evidence to support its removal until after removal. Regardless, the Court denies Plaintiff's motion to remand based on an untimely removal.
IV. ORDER
Therefore, it is hereby ORDERED that Plaintiff's motion to remand (Dkt. 14) and motion strike the declaration of Michael Silver (Dkt. 74) are DENIED. The parties shall submit a joint proposed class certification briefing schedule no later than August 1, 2018.
Dated this 19th day of July, 2018.
/s/_________
BENJAMIN H. SETTLE
United States District Judge