FINDINGS OF FACT AND CONCLUSIONS OF LAW

JUSTIN S. ANAND, United States Magistrate Judge.

Plaintiff, a workers compensation self-insurance fund, sues Defendant, its former third-party administrator, for breach of contract.1 In essence, Plaintiff alleges that Defendant mismanaged thousands of claims that Defendant handled for Plaintiff from approximately 2008 through 2010. This matter is before the Court following a bench trial from May 29, 2014 through June 4, 2014, and the filing of various posttrial briefs and other submissions. The Court has considered the testimony and other evidence, as well as the parties' arguments, and hereby renders its findings of fact and conclusions of law.

I. MATERIAL FACTUAL FINDINGS

A. The Parties and Their Contract

Plaintiff Georgia Operators Self-Insurers Fund ("Plaintiff' or "Georgia Fund") is the entity through which over 90% of the McDonald's restaurant franchise owners in Georgia self-insure for workers' compensation liability. (Tr.2 at 24-26, 36-37). Randall Squires, a commercial insurance broker at Arthur J. Gallagher in Tampa, Florida, has served as the administrator and principal of Georgia Fund since 1993. (Tr. at 23-25, 29). Squires and Arthur J. Gallagher administer the Georgia Fund and provide certain services including insurance, loss control and risk management. (Tr. at 27-28). Squires, however, lacks expertise in claims handling and adjusting and therefore relies on outside expert firms to handle claims handling and other duties. (Id.)

In January 2008, Georgia Fund contracted with Defendant PMA Management Corp. ("Defendant" or "PMA") for Third Party Claims Administrative ("TPA") Services. (Pl. Ex. 1). Specifically, PMA took over the handling of Georgia Fund's then-pending workers' compensation claims as well as newly-filed claims through December 31, 2010. (Id.). Defendant committed in the contract to provide "customary and appropriate workers' compensation claims handling services for all" claims to be handled. (Id. § 3(a)).

This agreement was extended for two one-year periods, ultimately through December of 2012. (Pl. Ex. 149, 150). The extensions included a 60-day termination period as opposed to the 90-day termination period that applied in the original contract. These extensions were only for one year each because, as Squires explained, "at this point we had all figured out ... that we had some serious claims issues ... I didn't want to commit beyond a year certainly, and I wanted a 60-day out in case things really continued to be done poorly." (Tr. at 50).

PMA contracted to handle both "lost time" claims, "where the injured employee will likely be entitled to medical benefits and has lost or may lose time from work or will submit or has submitted a claim for indemnity benefits," and "medical only claims," where an injured worker "is or may be entitled to medical benefits; but, is not likely to lose time from work or submit a claim for indemnity benefits." (Pl. Ex. 1 at 2). Claims filed before PMA's tenure as TPA, but which were still pending as of the beginning of TPA's tenure and which therefore were subject to PMA's ongoing handling responsibilities, were referred to as "takeover claims." (Id. at 3). PMA took over 137 "lost time" and 49 pending "medical only" claims on a "takeover" basis. Id.

As Squires explained, the role of a Third Party Administrator, such as PMA, is to "actually handle the claims as they come in, communicate with the employees, the injured workers, communicate with the insured, which would be the members of my group [i.e., restaurant franchise owners], manage the claims process and try to bring the claims to some type of resolution either through settlement or just claim closure because of inactivity." (Tr. at 29).

PMA withdrew money from an account funded by Georgia Fund to pay benefits and expenses relating to claims handling, including medical expenses, legal costs, and payments to private investigators. (See Pl. Ex. 1 at 8-11). PMA received a flat fee for each claim that it handled in a year, based on a schedule by which different categories of claims generated different fees. (See Pl. Ex. 1 at 11-12). The Agreement required PMA to "indemnify, defend and hold harmless [Georgia Fund], its officers, directors, employees, and agents, from all claims, losses, damages, costs, liability or expenses, including attorneys' fees, caused by or resulting from negligence or willful misconduct of [PMA]." (Id. at 14).

B. Georgia Fund Began to Notice Significantly Increased Claim Costs And Other Problems With PMA At Least Through 2010

From its inception in 1993 through 2007, the Georgia Fund used the Gallagher Bassett firm-an affiliated but independent company to Arthur J. Gallagher-as its TPA. (Tr. at 30-31). The Georgia Fund decided to hire PMA to be its TPA beginning in 2008, because of PMA's expertise and reputation in handling workers' compensation claims, and because Squires "felt very comfortable" with PMA supervisor Maggie Conatser. (Id. at 31-32).

Squires started to become concerned with PMA's performance during 2009:

(Tr. at 53).3

Contributing to Squires's concern, he also began observing an "escalation" in claims reserves, reflecting increased claims-related expenditures and anticipated expenditures. (Tr. at 53-54). This escalation was "beyond what was expected and what I had seen historically over the prior 15 years." (Id.). Squires noticed that PMA was paying claims expenses at a rate that was depleting the Georgia Fund's bank accounts, and he also observed from loss reports received from PMA, and actuarial reports prepared by the Georgia Fund's actuarial professionals, that losses and estimated losses were trending higher than he had expected. (Tr. at 54). Squires recalled one incident in 2010 or early 2011 in which reserves had to be increased by over $1 million for a prior year. (Id. at 55).

C. PMA's Internal Correspondence Reflects Admissions That It Was Not Properly Handling The Georgia Fund Account

In approximately late 2010, Squires contacted PMA's Conatser to express his concern that "the claims adjuster was overworked and overwhelmed and not working the files." (Tr. at 57). Squires requested and PMA agreed to provide a detailed file review, so that PMA could show for the Georgia Fund Board of Directors the work that had been performed on the account. (Tr. at 57-60). However, neither the adjuster (Mabry) nor her supervisor (Adolphus Drain) appeared for the file review, and Conatser was not personally prepared to provide a claim-by-claim discussion. (Tr. at 60). Conatser apologized, and acknowledged to Squires that "we have a problem, I'll get back with you." (Id.).

Various internal emails among PMA supervisors, including Conatser, reflect a belief that Mabry was too overworked to properly handle the Georgia Fund account. For example, on December 3, 2010, PMA Sales Vice President Michael MacAulay stated to fellow PMA Vice President James Walsh that "[a]t least 60-75 days ago we knew Connie[] [Mabry's] pending was unmanageable," and that another major client had informed PMA that Mabry "wasn't getting the job done." (Pl. Ex. 9 at 1). MacAulay further stated that "we knew their [sic] were issues with GA claims handling ... Kathy Reeds audit of the [Georgia Fund] claims I believe showed questionable claims handling." (Id.) On December 1, 2010, Quality Assurance Specialist Kathleen Reed stated "I know that Connie's pending has been high and an addition to staff should help to give her some relief." (Pl. Ex. 8, at 2). That same day, Walsh sent Reed, Conatser and others an email stating that "[Georgia Fund] has expressed concerns regarding claim service. The issues are primarily driven by Connie Mabry's excessive workload." (Pl. Ex. 7).

Generally, the acceptable workload for an adjuster in Georgia is between 130-150 claims. (See Tr. at 534-545). Mabry testified that during the time she was assigned the Georgia Fund account, she was also assigned another major account for PMA, and because of the combination of these clients her workload at least for a period of time increased to approximately 230 total claims. (Mabry Dep. [154] at 41, 50, 133-134; Tr. at 872).

In an internal email dated January 6, 2011, Walsh asked Regional Claims Manager Jeff Bonner "how much leakage occurred?" (Pl. Ex. 11 at 1). Walsh's response, from a review of the Georgia Fund claims, stated that instances of "blatant leakage," such as outright missing court deadlines, was "minimal." (Id.). However, Bonner stated that "the issue has been delays," that is, possible delays in settling a case as promptly as possible, paying "late medical bills," or otherwise suffering "additional benefits paid due to the lack of aggressive handling." (Id.). Bonner opined that the financial impacts of such problems would be "tougher to quantify," or, as to the possibility of missing opportunities to settle at more opportune times in a case, "I have no idea how to quantify that." (Id.).

In approximately March of 2011, Squires hired an outside claims auditor, Stacy Hosman, to conduct an audit on the Georgia Fund claims. (Tr. at 81-82). Upon Squires's request, PMA gave Hosman access to Georgia Fund's claim files and Hosman audited fifteen lost time claim files and seven medical only claim filed from 2009 to 2011 for compliance with generally accepted best practices. (Tr. at 234). Hosman issued a report dated May 19, 2011. (Pl. Ex. 36). Her report analyzed PMA's compliance with regard to thirteen specific areas of workers' compensation practice. (Pl. Ex. 36 at GAOSIF 0001328). The report was highly critical of PMA's performance. Hosman, in her "summary and conclusion" section, opined that:4

Critical identified weaknesses in file management were:

(Id. at GAOSIF00001336-1337).

Hosman provided more detail in the body of her report. (Pl. Ex. 36). For example, in her section entitled "Claims/Medical Management," she stated that:

(Pl. Ex. 36 at GAOSIF0001331) (emphasis added).

As another example, Hosman found that PMA's investigative activity, which should be a "critical foundation for the appropriate and cost-effective resolution of claims[,]... fell significantly short of expectations." (Id. at GAOSIF0001329-1330). Hosman identified four files where "the lack of appropriate investigative activity prejudiced claims handling and resolution," three files where "clearly missed opportunities to continue the investigation process can be found," and three files in which costly surveillance by a private detective was assigned, but with "little follow-up or documentation of the [surveillance] results," and "no evidence" that surveillance results were provided to the treating provider. (Id. at GAOSIF0001330-1331). Hosman made similarly highly-critical findings as to a number of other specific categories. (Pl. Ex. 36).

Hosman spoke on multiple occasions with PMA supervisors Terry Kunselman and Maggie Conatser. (Tr. at 275-277). On more than one occasion, both Kunselman and Conatser "indicated that they went through the review, went through their own internal processes, went through the claim file in great detail, and they could not argue my findings based on what I presented in my report." (Tr. at 277). Similarly, PMA representatives acknowledged to Squires that "everything in the report was valid, other than some small points, and that we needed to fix this going forward." (Tr. at 85).5

In an even more telling internal email exchange within PMA, Walsh directed Kunselman (head of the Quality Assurance Department), to review and determine the accuracy of Hosman's findings, and on June 26, 2011, Kunselman responded with almost complete agreement with Hosman's conclusions, including her findings of "file abandonment":

(Pl. Ex. 38 at 1) (emphasis added).

Kunselman testified at trial that although she agreed with Hosman's findings at the time she sent the June 26 email, she has since "gained an understanding of the contract" between PMA and the Georgia Fund, and that this "understanding" has changed her view of the accuracy of Hosman's findings. (See Tr. at 668-670). Specifically, Kunselman now states that she disagrees with Hosman's findings to the extent Hosman was "very critical of the way we handled medical-only" claims in terms of not contacting and interviewing the claimants. (Tr. at 670-71). According to Kunselman, she now understands that the contract itself provides that PMA would simply process medical-only claims without interviewing the claimants, and on this basis Kunselman testified that she no longer agrees with Hosman's report "in its entirety." (Id.).

The Court does not give substantial weight to Kunselman's testimony in this regard. The Court perceives Kunselman's change-of-heart to concern only a small part of the Hosman analysis. As explained above, most of Hosman's analysis involved audit of lost time claims, and the medical only claims were a smaller subset of the audit. And Hosman's report discusses many more patterns of problems than simply failing to contact medical only claimants, including wholesale "file abandonment" throughout the claims process. Kunselman's testimony, which simply stated that she no longer agrees with the Hosman report "in its entirety," did not retract her broader admission that in most respects, and in particular with the analysis of lost time claims handling, the findings were "accurate with very few exceptions." (Pl. Ex. 38 at 1).6

D. Testimony Of Georgia Fund's Claim Handling Expert

The Court accepted the expert testimony of Plaintiff's expert witness, Douglas McCoy, on the subject of workers' compensation claims handling practices in Georgia. (Tr. at 326-327 et seq.). Among other qualifications, McCoy has approximately 40 years of experience adjusting workers compensation claims, including in Georgia; he is licensed as a workers' compensation adjuster in Georgia; and he prepared the required training program for new licensed workers' compensation professionals in Georgia. (Tr. at 326-340).

As the parties agreed at trial, all exhibits including McCoy's expert report were admitted into evidence, (Tr. at 493-498)7, although McCoy also testified at length to explain and expound upon his opinions. McCoy reviewed 88 claims files handled by PMA and ultimately concluded that PMA "failed to follow practices and procedures normally followed in the management of workers compensation claims as well as their own Claims Handling Standards, and as a consequence damaged the [Georgia Fund]." (Pl. Ex. 135 at 11). In addition to McCoy's testimony and report, the parties also submitted arguments and relevant documents in support of their positions as to the 88 claims discussed by McCoy. (Joint Claim Summary [168]).8

McCoy opined as to patterns that he perceived in his review of the 88 claims reflecting PMA's failure to follow proper claims handling procedures, including the following issues, as examples.

One recurring problem, which McCoy referred to as "Change in Condition," was PMA's repeated failure to file certain documents limiting financial exposure in certain cases where a claimant's medical condition improved. (See Tr. at 374-376; Pl. Ex. 135 at 3-4). For example, where an employee is released by his physician to light duty work, but where the employer has no available light duty jobs to provide, the employer may be obliged to keep paying benefits for a lengthy period of time. (See Tr. at 374-376). This obligation can be limited, and the number of months of ongoing financial exposure can be reduced under Georgia law, by the filing of a document known as the WC-104. (Id.; Pl. Ex. 135 at 3-4). McCoy testified that he found "many cases" where this form should have been filed, but was not filed, including cases where "[PMA] even mentioned it should be done [in claim file notes], but it wasn't done. In some cases they did it, so they knew it should be done, but in others they didn't do it." (Tr. at 376). In McCoy's opinion, this caused financial damage because although most of these cases were ultimately settled, the greater ongoing financial liability associated with the failure to promptly file the appropriate form necessarily impacted negotiating leverage and settlement valuations. (See id.)

Similarly, Georgia law also protects an employer in the event an employee is released to work on light duty and the employer offers a light duty position, but the employee refuses to take it. (See Tr. at 383; Pl. Ex. 135 at 4). In that circumstance, which McCoy also referred to as a type of "Change in Condition," the employer can file a document known as WC-240. (Pl. Ex. 135 at 3-4). This document has the effect of terminating benefits altogether. (Id.). McCoy testified that he discovered incidents where PMA failed to file or promptly file this document where it was clearly available and that this resulted in ongoing benefits being paid unnecessarily. (See Tr. at 383-389; Pl. Ex. 135 at 5-6).

The Court's review of the underlying claims records corroborates McCoy's opinion that, on multiple occasions, PMA filed to timely file the appropriate forms to limit Georgia Fund's liability or otherwise failed to monitor the claimant's change in condition. See, e.g., Zakiesha L. Claim, # W890919725 (summarized in Joint Claim Summary [168] at 94) (adjuster received approved job description from treating physician on May 26, 2011 but did not file the WC-240 until July 26, 2011, resulting in several hundreds of dollars worth of unnecessary indemnity payments); Blanca L. Claim # W890906837 (summarized in Joint Claim Summary [168] at 114-15) (Defendant continued to pay claimant indemnity benefits for three months after claimant was released to work by her physician and had returned to work; no evidence Defendant adjuster engaged in any diligence to discover this change in condition); Maria R. Claim # W890736031 (summarized in Joint Claim Summary [168] at 182) (Defendant received WC-240 offering claimant a position, approved by claimant's physician, in early April, but still paid indemnity benefits through late June that were not required to be paid because of the furnishment of the WC-240; Defendant argues that surgery was scheduled for June and "there was no point" in returning claimant to light duty in the meantime, but the doctor notes indicate that claimant was still allowed to do light duty work from April through June despite the upcoming surgery, and therefore three months of indemnity payments were paid unnecessarily); Patricia M. Claim # W880914160 (summarized in Joint Claim Summary [168] at 147) (claimant was released to work in August 2009 and WC-240 was filed providing for return to work in October 2009, but benefits continued to be paid until April 2012 despite the fact that the WC-240 should have terminated the right to indemnity payments; Defendant does not directly refute this but rather argues "this case settled for just $4,000 over PMA's estimated exposure," which has no bearing on whether or not unnecessary indemnity payments continued to be paid for approximately 18 months).

McCoy also discussed PMA's compliance with its own claims handling standards with regard to use of expensive private investigation services. (Tr. at 393-400, Pl. Ex. 135 at 6-7). McCoy explained that, pursuant to PMA's written claims handling standards, an "activity check must be performed prior to any assignment of surveillance in order to justify the need for surveillance." (Tr. at 395).9 This requirement may be "waived" if the adjuster otherwise has reliable information from another source that suggests the need for surveillance. (Id.). These claims handling standards are in evidence, and the Court finds that McCoy accurately summarized them. (Pl. Ex. 2 at 18 ("Activity checks must be performed prior to any assignment of surveillance in order to justify need for surveillance.") (emphasis added)).

McCoy explained that surveillance was assigned in 39 of the 88 files that he reviewed, amounting to 44%. (Tr. at 396-7). Yet, despite the requirement that surveillance only be assigned pursuant to an activity check, no activity check had been done prior to any surveillance, and no file contained a notation documenting any other credible information upon which the adjuster relied in assigning surveillance. (See id.).

McCoy provided and the Court has reviewed examples of improper and wasteful use of surveillance that he found in the files. For example, in the case of Earleen W., Claim # W890938592, surveillance was assigned in late 2010 without any prior activities check, although there had been several uncompleted requests for one documented in the file. (Tr. at 397-398). The only documented reason that Defendant points to in the file for approving surveillance was that the owner/operator "does not want [the claimant] back," (See Log Note 24, submitted with the Joint Claim Summary [168] at 229), which is not a justifiable reason under Defendant's standards for assigning surveillance. Ultimately, when surveillance was assigned, two separate investigators were assigned duplicative surveillance responsibilities for the same time period, without any explanation or other apparent need for the overlap. (See Tr. at 397-98; see also Aubrey C., Claim # W890951020 (summarized in Joint Claim Summary [168] at 28) (Defendant paid $1,600 for surveillance with no indication of any prior activities check or based on any specific rationale, other than speculation that surveillance may "determine if [claimant] is doing any activity that may be delaying healing and progress to being released from care.")).

Another problem that McCoy identified "in many of the files," was that PMA made late payments of benefits including medical charges, which under Georgia law results in penalties. (See Tr. at 416-420). McCoy cited the example of Claudia R. (Claim # W890909526), in which PMA failed to pay certain medical bills on time in 2009 and 2010, incurring penalties in excess of $4,000, which were paid with the Plaintiff's funds. (See Tr. at 417-418).10

McCoy also discussed in detail PMA's failure to utilize outside counsel in a cost-effective manner, pursuant to their own claims handling standards as well as best practices. (Pl. Ex. 135 at 8-9; Tr. at 420-433). First, with regard to settlement negotiations, PMA's own standards provide that all negotiations are to be handled by the adjuster, not outside counsel, unless there is a specific reason documented in the file as to why this task is assigned to counsel. (Pl. Ex. 135 at 8; Pl. Ex. 2 at 22). McCoy testified that, in the workers' compensation field, the adjuster still takes an active role supporting the litigation even when outside attorneys are hired, including not just settlement but also in obtaining witness statements and medical records and other tasks. (Tr. at 428-29). This is to keep claim expenses down since an adjuster's time is much cheaper than that of an outside lawyer. (Id.). According to McCoy, however, PMA authorized outside counsel to conduct negotiations in almost all the cases he audited, with no explanation in the file as to why this adjuster task was being assigned to a more expensive outside lawyer. (Pl. Ex. 8).11

PMA, through delay or inaction, also likely lost potential chances to reach settlements at more favorable amounts in cases that otherwise resulted in extensive liability. For example, Plaintiff and McCoy pointed to the case of Kendra M., Claim # 890735756. In this case, the claimant's attorney made a settlement demand of $35,000 on June 3, 2009, and the adjuster already had requested and received authority to settle for this amount. (Joint Claim Summary [168] at 123). But with no explanation in the file, the demand was not accepted, two months went by, and on July 31 the claimant retained a new lawyer. (Id.) The claim eventually settled in 2011 for $43,000, which was $8,000 more than it appears PMA had the opportunity to and had decided it should settle for back in June of 2009. (Id.;12 see also Yuriana C., Claim # W890735779 (summarized in Joint Claim Summary [168] at 31) (claimant made demand of $15,000 in December 2008; file shows minimal efforts to negotiate settlement with several months going by before making requests for authority or communicating offers; case ultimately settled for $13,500 after passage of sixteen months during which nearly $13,000 in indemnity payments were paid that could have been avoided with a prompt settlement; also, $1,235 in surveillance was paid for with no documented rationale, and with little apparent benefit when the claim was ultimately settled for only $1,500 less than the initial demand made sixteen months earlier)).

The above issues are not the only breaches of claims handling standards that McCoy identified in his report and testimony, and the various individual claims discussed above are not the only claims as to which the parties submitted evidence. The Court generally finds McCoy's analysis as to PMA's failures to follow proper claims handling procedures to be credible, as corroborated by the Court's review of the underlying claims files, as well as by Hosman's finding (admitted as accurate by Defendant) of "complete file abandonment," as well as by admissions of other PMA supervisors suggesting that they knew Mabry's workload was unmanageable and causing claims handling deficiencies.

E. PMA's Actions To Address The Problems

In early 2011, in reaction to the issues raised by Squires and Hosman, PMA's Vice President for Claims James Walsh supervised implementation of an "action plan" to address the concerns. (See Tr. at 531-32, 640; Pl. Ex. 20). This plan included increasing adjuster staffing on the Georgia Fund account, so that any adjuster's pending case load would not exceed 130 claims, active supervision and claims reviews, a performance guarantee, and various audits including a monthly quality assurance audit conducted by corporate quality assurance specialists. (Pl. Ex. 20 at 2). Also, at Squires's request, Defendant agreed to pay for Hosman to remain actively involved, including by conducting annual audits. (See Tr. at 86-87).

Plaintiff argues that Defendant was dishonest in the furnishment of the monthly quality assurance audit reports, which Plaintiff argues were manipulated to show a falsely high level of performance.

At issue in particular was the first monthly audit report, sent to Hosman and Squires on August 10, 2011. (Pl. Ex. 53, 55). This report reviewed 10 selected claims for compliance with best practices, specifically with regard to 15 separate categories of claims handling tasks including conducting a timely investigation, proactively managing the physicians, making timely contacts with the claimant and employer, etc. (Pl. Ex. 20, 53). Based on their review, PMA's quality assurance personnel scored PMA's overall performance with regard to the audited files at 85%, which was considered a "passing" score. (See Tr. at 756).

Unbeknownst to Plaintiff, however, this audit report had been substantially changed from an original version that reported PMA's overall performance at a dismal 52%. (See Tr. At 50). On Friday, August 5, 2011, the QA specialist who conducted the audit, Daphyne Daniels, sent an email entitled "[Georgia Fund] August 2011 Internal Audit," announcing that "the cumulative score for the audit was 52%," to Kunselman and Reed. (Pl. Ex. 50 at 1). That same morning, Kunselman forwarded the email without comment to Walsh. (Pl. Ex. 51 at 1). The next business day, Monday, August 8, 2011, Kunselman sent an email to Daniels and Reed stating "[l]et me know if the two of you are available for conference call today at 12:30.... we can't release the attached audit results...." (Pl. Ex. 52 at 2).

The next day, apparently after the teleconference, Daniels sent another version of the audit to Kunselman and Reed that reported the score at 85%. (Pl. Ex. 53 at 1). In this version, several items had been changed from the original, in several cases noting a "Y" indicating that Defendant complied with the appropriate standard for a particular claim, whereas the original August 5 version of the audit results noted a "N" for that same category on that same claim. (Compare id. with Pl. Ex. 50). Daniels also replaced four of the ten claims altogether. (Id.). Each of these changes was favorable to the overall score. (Id.). Squires was not aware at the time of a prior, non-passing audit report that was replaced by the August 10 report. (See Tr. at 89-90).

Kunselman at trial could not recall receiving the August 5 audit results, could not recall looking at the document before she forwarded it to Walsh, could not recall whether there was any conference call discussing the August 5 audit results, and could not recall receiving the second version. (See Tr. at 658-61). Nevertheless, Kunselman had no trouble explaining what she meant in her August 5, 2011 email stating "we can't release the attached audit results." (See id. at 659-660). According to Kunselman, she believed that PMA could not release the audit results "because they were in draft form." (Id.).

Daniels testified in detail about the specific reasons why the scoring changes were made from the August 5 version to the August 10 version of the audit report. (Tr. at 797-98, et seq.). Daniels stated that one issue that was discussed on the call was that she applied more onerous Florida standards and not the Georgia standards that governed Defendant's claims handling work for Plaintiff. (Id.). For example, Daniels applied a stricter Florida standard for following up with medical providers. (Tr. at 726-27). Daniels also explained that she was instructed to specifically focus only on auditing the performance of the current adjuster on the claim at the time of the review, which apparently had not been explained beforehand. (Tr. at 798-99). This led to at least one change. In the August 5 report, Daniels gave PMA a negative score in a particular case (Claim No. W000223224), for failing to comply with the rule that the adjuster make a "three point" contact within 24 hours with the claimant, the employer, and the physician. (See Pl. Ex. 50 at 2). Daniels gave PMA a negative score because this contact was never made when the claim was opened as it should have been. (Tr. at 803). However, the claim was reassigned to another adjuster a few weeks before the audit; thus, Daniels re-scored this for purposes of the August 10 report as a "N.A.," since the compliance failure was not the fault of the current adjuster. (Id.). In another case (Claim No. W000249089), Daniels initially gave PMA a negative score because the initial adjuster failed to make the required contact, but then changed that score to a positive one, because the claim was reassigned shortly before the audit, and a new adjuster made the belated contact. (Tr. 806-807).

In instances that the Court cannot help but find curious, at least three of Daniels's changes were based on adjuster actions that occurred on August 5, 2011 itself, which was the very day Daniels transmitted her initial report showing a 52% overall score. (Tr. at 814-16, 825, 827). Because Daniels sent her August 5, 2011 report at 6:25 am (see Pl. Ex. 50), the Court infers that these changes to various claim files were made after Daniels sent this unsatisfactory report. In other words, in at least three instances, Daniels changed her audit results to improve PMA's score based on actions taken after she conducted her analysis.

In another curious action, Daniels in many cases improved certain scores from the August 5 report to the August 10 report, because she herself fixed the problem she discovered while auditing the claim. (Tr. at 804, 813, 818). One of the audit categories was called "data integrity," which referred to the claims professionals' obligation to enter certain data in a separate database apart from the claims file. (Id.). In many cases, Daniels discovered that this was not done and therefore assigned a negative score in her August 5 audit. (See Pl. Ex. 50). However, after Kunselman instructed that "we can't release the attached audit results," Daniels went back and changed those scores to positive ones, because she herself (i.e., Daniels) had fixed the data integrity problem. (Tr. at 804, 813, 818).

The Court does not credit Kunselman's testimony that she issued her instruction that "we can't release the attached audit results" for the reason that the August 5 results "were in draft form." First, in light of Kunselman's inability to recall nearly any of the relevant facts surrounding this document, or the subsequent "final" report, or any of the discussions about it, it is not credible that she would recall this particular detail. Second, neither the audit report nor the cover email (Pl. Ex. 50) is labeled as a "draft" in any way, and nothing about the contents of those documents suggest that the findings were tentative, subject to discussion, that Daniels had any questions about how to complete the project, or that there was anything at all left for her to do. The email is simply entitled "[Georgia Fund] August 2011 Internal Audit," it begins by stating "The cumulative score for the audit was 52%. Areas/categories of immediate necessary improvement at or below 50% on the audit are: [listing issues]," and it concludes by saying "Please let me know if you have any questions. Have a great day!!!!" (Pl. Ex. 50 at 1). This language does not suggest that the document was considered at the time to be a draft.13 Therefore, by definition, the Court concludes that there was another reason why Kunselman made the instruction not to release the August 5 results.

As explained above, Defendant offered reasons for having made the scoring changes from the August 5 to the August 10 version. In several of these cases, the Court is unable to find that Defendant's reasons are not valid. For example, in cases where Daniels specifically explained that her negative score on August 5 was based on an erroneous application of Florida standards, the Court accepts that explanation. In other cases, however, the Court questions the credibility or legitimacy of the reasons for the change. In particular, it strains credulity for an auditor to change a negative to a positive score based solely on work that the auditor herself did to fix a deficiency. This appears to contradict the stated purpose of the audit, which was to grade the performance of the adjuster currently-assigned to the file. The Court also considers that in several instances, grades were improved based on work accomplished later the same day as the unfavorable August 5 report. Based on all of these circumstances, the Court concludes that there was at least a mixed motive behind the revisions to the August 2011 audit report. Some changes may have been justifiable on grounds of accuracy, but the Court concludes that there were also choices made for the purpose of increasing the score and avoiding sending a report documenting a failing grade to the client.

Nevertheless, PMA's action plan to address the claims handling deficiencies was otherwise successful. Hosman saw significant improvement in Defendant's performance after implementation of Walsh's action plan, beginning in early 2011. (See Tr. at 278-79). And, at that time, the Georgia Fund's claim handling expenses began to settle back down to their pre-PMA averages or even lower, as detailed further below.

F. Total Estimated Increased Claim Costs Associated With The Period Of Defendant's Deficient Performance

1. The Spike In Claim Expenses For 2007-2010

The Georgia Fund introduced business records in the form of its actuarial loss data from 1994-2013, to show how its estimated losses increased significantly for claims filed in the 2007-2010 period as compared to claims filed in prior and subsequent years. (See Pl. Ex. 170 at 37). Both Squires and the Georgia Fund's actuary, J. Edward Costner, testified to explain these records. (See, e.g., Tr. 53-55, 143-76). The actuarial reports keep track of historical claims loss data and estimate losses going forward. (Tr. at 54, 143-144). Squires explained that these records must by law be provided to state regulators and also are "vital parts of our financial statements which we use to figure out how much money we have and how much money we have available to pay dividends back to our members." (Tr. at 54-55).

Costner explained that actuarial reports are necessary for the Georgia Fund to keep track of its financial position, because claims filed in a particular year may continue to incur ongoing liabilities and expenses in future years. (Id. at 146). For example, a claim for an injury filed in 2007 may not be completely settled or otherwise closed out by the end of 2007, and may require additional medical treatments, income indemnity, or other payments for several years after 2007. (See id.). Thus, to accurately assess the financial liability associated with that still-open claim at the end of 2007, it is necessary to calculate not just the expenses incurred and committed to date, but also include an estimate for the future expenses associated with that claim that may develop. (Id.). Costner, as actuary, receives the actual loss data from the TPA. (Id. at 145-46). He then applies a multiplier to that actual loss data, based on the amount that Georgia Fund's claims liabilities have historically increased over time, to estimate how much liability those 2007 claims will incur over their lifetime. (Id. at 145-53).14

As of the March 2014 actuarial report (Pl. Ex. 170), for claims filed in the years 2000-2006, Georgia Fund's estimated losses for every $100 of payroll range from $.82 to $1.34. (Id. at 36-37). Costner testified, and a simple review of the figures stated in the document allows the Court to find, that the simple arithmetic average of all estimated incurred losses during that seven-year period yields an overall average estimated incurred loss of $1.11 per $100 of payroll. (Pl. Ex. 170 at 36-37; Tr. at 177). By contrast, the estimated ultimate incurred losses for claims filed during the years 2007-2010 (per $100 in payroll) are $1.46, $1.40, $1.63, and $1.44, respectively. (Pl. Ex. 170 at 37). Thus, the claims filed in 2007-2010 have generated estimated losses that are more than 30% greater than the claims filed in 2000-2006. The dollar amount of this "spike" in claims expenses is $3,014,501.15

The 2007-2010 policy period is significant in this case, because it generally covers the claims handled during the period when Hosman and McCoy found deficient claims handling practices by Defendant. As noted above, PMA took over as TPA on January 1, 2008 and therefore TPA handled from the inception any claims filed in 2008-2010. As for claims filed in 2007, PMA acquired management responsibility for any such claims not completely closed out during 2007. PMA took over a large number of "takeover" claims. Indeed, the evidence suggests that nearly 200 claims filed in 2007 were acquired as takeover claims by Defendant. (Compare Pl. Ex. 170 at 53 (reflecting 570 total claims in 2007) with Pl. Ex. 138 at 52 (reflecting that 379 claims closed in 2007)). It is logical to infer that less complicated and less expensive medical only claims were more likely to be "closed out" on a quicker basis, and therefore that the nearly 200 claims PMA took over from 2007 were disproportionally weighted towards the more expensive lost time claims. Therefore, while Defendant was not Plaintiff's TPA in 2007, the Court infers that Defendant engaged in substantial claims handling work for lost-time claims filed during 2007.

For claims filed from 2011 through 2013, after Defendant began correcting the deficiencies in its claims handling performance, the estimated ultimate incurred losses fell back down to $1.08, $1.11 and $1.01, respectively, per $100 in payroll, which was actually slightly less than the $1.11 weighted average from 2000-2006. (Pl. Ex. 170 at 37).

2. Defendant's Evidence Of Other Cost-Driving Factors

This $3 million-approximately 30% — spike in estimated losses for claims filed during 2007-2010 is quite stunning. Nevertheless, Defendant argues that other factors outside its control contributed to or caused the spike. The Court accepts that in a complex business, economic and regulatory environment, many factors contribute to costs. However, as explained below, the Court does not find Defendant's specific arguments as to alternative causes to be persuasive, at least to the degree Defendant suggests these factors impacted costs during 2007-2010.

For example, Walsh testified that during the years 2001 through 2007, the Georgia Fund received reimbursements for certain claims, totaling over $1.5 million, from an entity known as the Second Injury Trust Fund. (Tr. at 549). This fund was no longer in operation as of 2008, when PMA took over claims handling duties for Plaintiff. (Id. at 551). Plaintiff received reimbursements of approximately $215,000 per year on average from this source prior to 2008, and none thereafter. (Id. at 549-51).

Plaintiff does not refute that this Trust Fund expired, and the Court accepts that the expiration of this alternative source of recovery may have impacted costs to some degree during the relevant time period. However, the Court doubts that it had the impact advanced by Defendant. First, according to Defendant this source of recovery expired in 2008, and therefore this fact cannot explain the very notable spike in claim expenses that Plaintiff experienced for claims filed in 2007. Estimated claim expenses for claims filed in 2007 were $1.46 per $100 in payroll, which still reflected nearly a one-third jump from the $1.10 average expense rate from 2000-2006. This was so despite the continued recoveries from the Trust Fund.

Second, despite the elimination of the Trust Fund recoveries, average estimated claims expenses in 2011-2013 fell back to the historical average of approximately $1.10. There is no evidence that the Trust Fund was available again during this period. Plaintiff's costs in 2007 and 2011-2013 thus corroborate that the existence of the Trust Fund, while perhaps a factor, was not a major factor driving costs.

Walsh also testified, based on his review of the claims data, that the lost time claims in 2008-2010 contained a significantly greater proportion of claims from older workers (ages 50-64) than during the 2001-2007 period. (Id. at 555-56; Pl. Ex. 26 at 13).16 It is logical that this shift would have increased claims costs to some extent, since, as Walsh explained (and common sense provides) "older workers take longer to recover on average, cost, expenses, durations of disabilities last longer." (Tr. at 556).

However, Defendant offers no suggestion or analysis as to how much this demographic factor impacted costs. Based on the evidence in the record, it is hard to find that this factor impacted costs to a substantial extent. Plaintiff's evidence does not show a gradual increase in costs over time, but rather a steep jump from claims filed in 2006 to those filed in 2007. Costs jumped by over 50% between those two years and by 33% compared to the historical average. Defendant does not suggest-and without evidence the Court finds it incredible to infer-that the workforce of McDonald's restaurants in Georgia suddenly aged so dramatically overnight such that age can explain this "spike." Notably, also, average costs retreated in 2011 to historical norms, and Defendant does not suggest-nor does the Court find it reasonable to infer on this record-that the workforce suddenly rejuvenated to some great extent between 2010 and 2011. Thus, while the Court accepts that age shifts were a factor, the Court does not find that this was a major issue explaining the 33% cost spike experienced by Plaintiff.

Walsh also testified that Georgia required an 11% increase in income indemnity benefits, which obviously impacted and increased claim expenses for beneficiaries owed such benefits. (Tr. at 552).17 However, this increase only became effective in July 2007. Thus, if it were a major factor one would expect to see Plaintiff's costs for claims filed in 2007 to rise to a much lesser extent than the costs for claims filed in 2008 and afterwards. And one would expect to see costs continue to be higher for claims filed in 2011-2013. Again, this is not what the data shows. Thus, again, the Court considers this increase to be a factor, but not a major one.

Walsh also points to "annual WC inflation of over 4%," as well as the fact that additional locations and employees came under the Georgia Fund's coverage during the time of the spike. (See Tr. at 551; Pl. Ex. 26 at 18). However, the spike in claim losses from 2007/2008 through 2010, as documented in the Georgia Fund's actuarial reports, was expressed as a function of the overall payroll each year. (See Pl. Ex. 170 at 37). Normalizing the loss figures by payroll each year would seem to largely account for increases or decreases in the employee population, and at least partially account for inflation, because salaries are subject to inflation to some degree as well.18 Moreover, these factors would seem to apply equally during the 2011-2013 timeframe, during which costs returned to their pre-2008 average. Without substantially more explanation than was presented to the Court, therefore, the Court cannot find that these factors presented a substantial role in affecting costs during the time period of the spike.

The Court also places some weight, but not substantial weight, on the increase in lost time claims during 2008-2010. (Tr. at 546-47). Walsh explained that lost time claims increased on average during 2008-2010-compared to the prior seven year period-at a greater rate than the increase in less expensive medical only claims. (See Tr. at 545-46). Thus, the claims filed in 2008-2010 included a higher percentage of more expensive lost time claims than the claims filed in the prior period. (Id.) At face value, this could help explain why claims filed in 2008-2010 tended to be more expensive, at least to some degree.

However, the impact of this factor is limited because Defendant yet again has not explained how this increase in claims is not accounted for at least in part by the fact that payroll also increased during this same period. While the data Walsh points to shows that, on average, the absolute number of lost time claims filed each year during 2008-2010 was 21% greater than during the prior period, (see Pl. Ex. 26 at 7), it is also apparent that the amount of payroll increased significantly as well during the 2008-2010 time period, (see Pl. Ex. 107 at 37). It appears that when one normalizes for changes in payroll there may have been even fewer lost time claims per $100 in payroll during 2008-2010 than there were during the prior years.

For example, Walsh's data shows 61 lost time claims in 2005, compared with 75 lost time claims in 2008, a 23% increase. (See Pl. Ex. 26 at 7). But payroll from 2005 to 2008 also increased from $1.58 million to $1.97 million, almost a 25% increase. Thus, per $100 in payroll, there were actually slightly fewer lost time claims filed in 2008-and significantly fewer claims filed overall-than in 2005. The higher proportion of lost time claims versus medical only claims in 2008 versus 2005 may have affected costs, but it is difficult to conclude that this could have had a great impact when so many fewer claims overall were filed per $100 in payroll in 2008.

In sum, the Court finds that Defendant has pointed to at least some factors that could have impacted the cost increase from 2007/2008 through 2010 to some degree. Nevertheless, the Court finds that Walsh's testimony and Defendant's arguments are not credible to the extent they suggest these factors substantially impacted costs. The Court finds that the more logical and credible explanation for the substantial and sudden spike claim cost spike was that Defendant took over as TPA and, in Hosman's words, which Defendant admitted were accurate, performed "significantly below industry standards." (Pl. Ex. 36 at GAOSIF00001336-1337).

II. CONCLUSIONS OF LAW

A. Rulings On Certain Evidentiary Issues

1. The Testimony Of Plaintiff's Actuary, Edward Costner

As explained above, Plaintiff offered into evidence its annual actuarial reports. (See, e.g., Pl. Ex. 170). Plaintiff also introduced testimony explaining and discussing those reports and the data within those reports from its actuary, Costner, and its administrator, Squires. (See Tr. 53-55, 143-176). Costner and Squires explained how and why these financial records are used in the ordinary course of Plaintiff's business. (Id.). Costner, as the actuarial professional who prepared these reports, also testified to explain the meaning of the estimated ultimate loss figures, which is the data generally relied upon by Plaintiff in asserting its theory of damages. (See Pl. Ex. 170 at 37).

Defendant did not object to the admission of these financial records themselves into evidence. Rather, Defendant objects to the testimony of Costner, who was not permitted to testify as an expert, to the extent that he offered testimony beyond the reports themselves and specifically by performing calculations using the data in the reports. (See Def. Br. [166] at 42-43).

The Court overrules this objection. At the threshold, regardless of Costner's testimony, the actuarial records themselves were introduced into evidence without objection. The actual business records themselves establish Plaintiff's estimated ultimate losses for each of the years from 1993 through 2013. (See Pl. Ex. 170 at 37). These records state on their face that Plaintiff experienced a more than 33% increase in costs (per $100 in payroll) between 2006 and 2007, and that this spike in costs lasted until 2011, at which point costs retreated to their general historical norm. (Id.). This is the basic point on which Plaintiff relies to show damages, and it is supported by the data in the business records alone.

The Court takes Costner's testimony as having two purposes. First, Costner, along with Squires, explained the meaning of these business records, and how they are created and used in Plaintiff's business. This foundational testimony was helpful to the Court, although technically not necessary for admissibility since no objection was lodged to the records. In any event, this testimony was not expert testimony within the meaning of FED. R.EVID. 702.

A company can call its regular accountant or actuary as a fact witness to explain financial records that the witness prepared or otherwise has personal knowledge of from his or her ordinary employ. See, e.g., First Annapolis Bancorp., Inc. v. U.S., 72 Fed.Cl. 204, 207-208 (Fed.Cl.2006) ("a lay witness accountant may testify on the basis of facts or data perceived in his role as an accountant based on his personal knowledge of the company."). Indeed, a company's officer or regular accountant can even offer lay opinions about lost profits or other questions about damages, based on his or her personal knowledge of the company's records. See, e.g., Ryan Development Co., L.C. v. Indiana Lumbermens Mut. Ins. Co., 711 F.3d 1165, 1170 (10th Cir.2013) ("Though accountants often testify as expert witnesses, the trial court reasonably concluded that Ms. Williams and Mr. Rump offered lay testimony given their involvement in preparing Agriboard's proofs of loss and the like."); Lightning Lube, Inc. v. Witco Corp., 4 F.3d 1153, 1175-76 (3d Cir.1993) (admitting accountant's lay opinion testimony on lost profits based on particularized knowledge of and participation in business); In re Merritt Logan, Inc. 901 F.2d 349, 360 (3d Cir.1990) (admitting lay opinion testimony of principal shareholder of company concerning lost profits).

Here, Costner simply explained what the figures in his reports mean, how they are derived, and for what business purposes these data are used. While professional knowledge obviously contributed to his creation of these records, Costner's simple explanation of the documents and how they were created and used did not require expert witness treatment.

Second, while Costner testified to calculations he engaged in for purposes of this case, that testimony also did not cross the line into being expert testimony. Costner presented simple, grade-school level arithmetic based on the data contained in the actuarial reports. Costner engaged in nothing more than clear addition and division exercises applied to the cost and payroll data in the reports, to come up with results such as the average costs for different periods of time.

While Costner may have advanced degrees, training and specialized experience, nothing about this basic math required expert testimony treatment. See Ryan Development, 711 F.3d at 1170-71 (trial court properly declined to treat accountant's testimony as expert testimony, where "the trial court observed that the accountants used basic arithmetic, personal experience, and no outside expert reports in calculating lost income and other claims for coverage"). Indeed, as the Court noted during the trial, (see Tr. at 907-908), this arithmetic was more akin to ordinary summary evidence than expert testimony. Counsel in argument could have asserted these calculations based on the data in evidence, and the Court itself could have reached these same basic calculations with an ordinary calculator, with or without Costner.

Defendant argues that "[i]n calculating [Plaintiff's] losses, Mr. Costner presumably employed various methods based on claim data, loss data, and numerous assumptions, including loss development, severity trends, frequency trends, and expected loss ratios." (Def. Br. [166] at 43-44). Defendant cites nothing to support that assertion. To the contrary, Costner simply explained financial business records that were independently admitted into evidence without objection, and testified about the results of simple calculations using that data.19 None of this was expert testimony.

2. The Testimony Of Plaintiff's Claims Handling Expert, Douglas McCoy

Defendant moved to exclude McCoy as an expert witness in October of 2013, arguing among other things that McCoy offered common sense conclusions, and not expert opinions. (See Def. Mot. to Exclude [97]). Prior to the reassignment of this case to the undersigned, Judge Evans denied this motion, finding McCoy to possess sufficient expert qualifications, and that his expert report disclosed opinions that drew on specialized knowledge or expertise. (See Order dated November 23, 2013[107] at 24-25). In doing so, Judge Evans interpreted McCoy's anticipated testimony as not just constituting a "restatement of PMA's log notes." (Id. at 25). Rather, Judge Evans stated that "the Court sees potential value in McCoy's testimony in identifying the sections of the Georgia Workers' Compensation Code which prescribe what was supposed to be done and the statutory penalty for non-compliance on the one hand, and the actions PMA's adjusters took with respect to handling Georgia Fund's workers' compensation claims on the other hand." (Id.).

Defendant moved again on the eve of trial, in Limine, to restrict McCoy's testimony to only "identifying specific provisions of the Georgia Workers' Compensation Code and whether PMA violated those provisions." (Def. Mot. in Limine [134] at 2-3). Defendant sought to prevent McCoy from otherwise opining as to whether Defendant properly managed and adjusted claims, in light of Defendant's own standard practices and procedures and the best practices of claims handlers generally in Georgia. (Id). According to Defendant, the latter subject matter went beyond what Judge Evans's order permitted. The undersigned denied the Motion in Limine, but without prejudice to the renewal of any objections in light of specific testimony during trial. (See Minute Order [144]). Defendant generally renewed this objection at several points during McCoy's testimony. (See, e.g., Tr. at 317).

The Court overrules Defendant's objection. (See Tr. at 320-23). Defendant's effort to limit McCoy's testimony to just "identifying specific provisions of the Georgia Workers' Compensation Code and whether PMA violated those provisions" reads Judge Evans's order much too narrowly. While it cites McCoy's knowledge of Georgia workers compensation law as a clear example of his specialized knowledge and expertise, the Order does not expressly state that other topics were excluded. Nor is such a ruling implicit. Indeed, Defendant's Motion to Exclude [97] only sought the relief of exclusion, and not any restriction on the topics of McCoy's testimony. Thus, Defendant's position is that Judge Evans went beyond adjudicating the requested relief, and also made sua sponte rulings governing the scope of McCoy's testimony, all without expressly saying so. The Court cannot read the Order in this manner.

Nor after the hearing the testimony in full does the Court find that McCoy's testimony should be limited to just "identifying specific provisions of the Georgia Workers' Compensation Code and whether PMA violated those provisions." As set forth in detail above, much of McCoy's testimony and opinions were in fact premised on particular legal requirements under Georgia statutes. McCoy also testified about industry standards and best practices more generally, however, along with Defendant's compliance with its own standards. The Court finds these to be permissible topics of expert testimony. The ordinary factfinder does not have personal experience adjusting workers compensation claims, and could benefit from the testimony of one with such experience. As an experienced and licensed adjusting expert, McCoy helped define terms, explain relevant concepts, and identify standards of performance. Workers' compensation is a creature of state law, and it was evident that McCoy's opinions on proper claims handling practices were generally intertwined with his knowledge of state law and practice. But he did not have to specifically cite to the Georgia Code to support every point he made.

To be sure, McCoy's testimony included some conclusions that involved more specialized knowledge than others. Many points involved detailed specialized knowledge of claims handling practices and legal requirements, but some points were more ordinary. As an example of the later, as discussed above, McCoy testified about the case of Kendra M. (Claim # 890735756), in which it appears that Defendant neglected to respond to a favorable settlement demand. (See Joint Claim Summary [168] at 123). McCoy's conclusion was based primarily on the correspondence history expressly shown in the case file itself, which showed that Defendant failed to accept or even respond to a particular demand despite having received authority to settle at the amount demanded. (Id.). Defendant's objection to this conclusion and similar ones is that any ordinary factfinder could understand the argument and interpret the correspondence history in the file for themselves. One need not be an expert, in other words, to discern whether Defendant neglected to answer a letter.

That McCoy made some points edging closer to ordinary logic, however, does not generally detract from his overall status as an expert. Such conclusions did not predominate his testimony, most of which drew on special knowledge of legal requirements or claims adjustment practice. Indeed, even statements such as those relating to the claim of Kendra M. were made in the context of McCoy's extensive personal experience adjusting claims and resulting knowledge of best practices and legal requirements. An ordinary factfinder would not necessarily understand what happens in settlement negotiations, or what it means to have settlement authority or to be faced with a demand, or appreciate the urgency and fluidity of settlement negotiations, all to the same extent as would a lawyer or claims adjuster personally experienced in settlement negotiations. Thus, while some conclusions by McCoy were more "expert" in nature than others, the Court finds his testimony as a whole to be properly based on his specialized knowledge as an experienced and licensed Georgia claims adjuster.20

B. Rulings On The Merits

1. Defendant Breached Its Obligation To Provide "Customary And Appropriate Workers' Compensation Claims Handling Services"

Georgia courts have explained that:

UWork.com, Inc. v. Paragon Technologies, Inc., 321 Ga.App. 584, 740 S.E.2d 887, 893 (2013) (internal quotes omitted) (quoting Norton v. Budget Rent A Car System, 307 Ga.App. 501, 705 S.E.2d 305, 306 (2010)); see also Bd. of Regents of the Univ. System of Ga. v. Doe, 278 Ga.App. 878, 630 S.E.2d 85, 93 (2006).

It is undisputed that Defendant bore a contractual obligation to Plaintiff to, among other things, "provide customary and appropriate workers' compensation claims handling services," (Pl. Ex. 1, § 3(a)), and that Plaintiff is entitled to complain about a breach of this obligation. As described at length above, Plaintiff, who bears the burden of proof, has shown several individual examples of mismanaged claims as well as a broader pattern of claims mismanagement by Defendant at least until 2011. The Court therefore finds in Plaintiff's favor on this point and specifically finds that Defendant is liable for breach of contract.

The Court's conclusion in this regard is based on various independent sources of evidence. First, a variety of correspondence among senior PMA officers and supervisors in late 2010 and early 2011 reflect admissions that PMA was mismanaging the Georgia Fund account. The Court affords great weight to these emails and other statements, as they were made in the ordinary course of PMA's business, and were in the context of internal and confidential discussions.

For example, on December 3, 2010, PMA Sales Vice President MacAulay stated to fellow PMA Vice President James Walsh that "[a]t least 60-75 days ago we knew Connie[] [Mabry's] pending was unmanageable," and that another major client had informed PMA that Mabry "wasn't getting the job done." (Pl. Ex. 9 at 1). MacAulay further stated that "we knew their [sic] were issues with GA claims handling ... Kathy Reeds audit of the [Georgia Fund] claims I believe showed questionable claims handling." (Id.)

More broadly, PMA's Quality Assurance supervisor Kunselman opined to Vice President Walsh that Stacy Hosman's outside audit in 2010 was "accurate with very few exceptions." (Pl. Ex. 38). What Kunselman in this candid and confidential analysis agreed was "accurate" was a sweeping condemnation of Defendant's claims handling practices on the Georgia Fund account. Hosman minced no words in opining that she observed:

(Pl. Ex. 36 at GAOSIF00001336-1337). As explained above, Hosman went on to detail specific categories and types of claims mishandling problems that she observed along with citations to specific claim files as examples. Kunselman's analysis of this report was not a casual one. She was the PMA supervisor tasked with reviewing the accuracy of Hosman's report, and Kunselman's email went into great detail as to the reasons why she agreed with Hosman's conclusions. Indeed, Kunselman not only stated that Hosman's findings were "accurate with very few exceptions," but that "I think she was actually looking for reasons to be favorable," and that Hosman "clearly saw the file abandonment which we knew would be brought out in the audit." (Pl. Ex. 38 at 1).21

Second, the Court also places significant weight on McCoy's expert testimony, both in terms of the patterns of claims mishandling problems that he observed in his review of 88 Georgia Fund claims, and as to his analysis of various specific files. McCoy's general findings and conclusions are consistent with those expressed in internal PMA emails, which also acknowledge widespread claims mishandling problems including file abandonment. For example, the Court finds that McCoy's testimony regarding recurrent problems with mishandling changes in condition, litigation management (including settlement), and misuse of surveillance expenses is credible and well supported by specific examples in the record.

Third, that Defendant was providing less than "customary and appropriate" claims handling services was corroborated by the results of the monthly audits of the Georgia Fund account that Defendant itself began conducting in August 2011. As noted in detail above, PMA's quality assurance staff initially gave its claims handling efforts a failing score of 52% in the August 2011 audit. While the Court accepts that some changes were made to that report for accuracy purposes, the Court also finds that some changes were made for purposes of simply increasing the score. Other changes were made to eliminate otherwise accurate negative scoring related to historical claims mishandling by prior adjusters, because that deficiency was deemed irrelevant for the purposes of the audit. In any event, the Court finds that the original 52% failing score likely more accurately reflected the state of the claims in the sample than the subsequent 85% passing score that was provided to Plaintiff.

Fourth, the Court considers that the Georgia Fund's claims expenses rose so notably for claims filed during the 2007-2010 period, which generally corresponds to the time period when Hosman and McCoy identified patterns of claims mishandling problems. This "spike" for 2007-2010 claims does not itself prove that Defendant was in breach of its duties during that time period. But the pattern strongly corroborates the other evidence discussed above showing systemic claims mishandling problems largely resulting from the unmanageable caseload of the primary adjuster.

In sum, the Court finds that Defendant breached its obligation to "provide customary and appropriate workers' compensation claims handling services" (Pl. Ex. 1, § 3(a)) to the Georgia Fund, as alleged in Count I. The Court will therefore proceed to consider causation and damages.22

2. Causation

The Court has little difficulty finding that the Defendant's "file abandonment" problems resulted in higher expenses and liabilities for Plaintiff. Indeed, Kunselman admitted as much. (See Pl. Ex. 38 at 1 ("[D]ue to the adjustor's lack of managing the disability, settlement values and the ultimate cost of the claims have increased.") (emphasis added)).

That neglecting file handling responsibilities across many claims will increase costs is well supported by the testimony of McCoy, the findings of Hosman (as adopted as "accurate with very few exceptions" by Kunselman), as well as by simple common sense. The Court also considers the Georgia Funds' actuarial reports, which show that estimated claim expenses associated with claims filed in 2007-2010 are significantly higher, on average, than such expenses for claims filed both before and after. That claim expenses so dramatically increased when Defendant began handling Plaintiff's account, strongly suggests that the file abandonment problems acknowledged by PMA during this period had at least some detrimental financial impact. The more difficult question, which the Court will discuss below, is the precise calculation of the measure of damages.

3. Calculation of Damages

a. The Court Does Not Calculate Damages Individually With Respect To Every One Of The Thousands Of Claims Handled During The Relevant Period

Once liability is otherwise established, "[t]he ability to estimate damages to a reasonable degree of certainty is all that is required ... and mere difficulty in fixing the exact amount will not be an obstacle to an award." Shepherd v. Aaron Rents, Inc., 208 Ga.App. 139, 430 S.E.2d 67, 70 (1993) (quoting Pottinger v. Cross, 170 Ga.App. 647, 317 S.E.2d 850, 852 (1984) (internal quotes omitted)). "The rule against the recovery of vague, speculative, or uncertain damages relates more especially to the uncertainty as to cause, rather than uncertainty as to the measure and extent of the damages." Id. (quoting Kuhlke Constr. Co. v. Mobley Inc., 159 Ga.App. 777, 285 S.E.2d 236, 239 (1981) (internal quotes omitted)).

The parties significantly disagree as to how the Court must approach the question of damages. According to Defendant, the Court must analyze every underlying workers compensation claim individually to determine whether Plaintiff has proved an amount of damages specifically associated with that claim. In other words, Defendant's position is that the Court must convene a mini-trial over, at a minimum, each of the 88 individual claims that McCoy discussed, and must make separate factual findings as to breach, causation and damages as to each one.

According to Plaintiff, having found that Defendant breached the contract by engaging in claims mishandling over a period of time, the Court has discretion to consider evidence of damages in the aggregate, and not just on a claim-by-claim basis. Indeed, Plaintiff argues that its damages claim is not limited to the 88 specific claims discussed by McCoy, and that the Court can infer that the widespread claims mishandling practices demonstrated in that sample also infected the thousands of other claims Defendant was handling for Plaintiff during that same time.

The Court agrees with Plaintiff that a claim-by-claim analysis is not legally necessary and, in this case, is not the most accurate measure of damages.

The Complaint does not allege 88 or more separate breaches specific to every individual workers compensation claim. The Complaint simply alleges a single claim that, in the aggregate, Defendant did not "provide customary and appropriate workers' compensation claims handling services for all Qualified Claims." (Pl. Ex. § 3(a)) (emphasis added). The Court finds Defendant to have breached this provision, because Plaintiff proved that Defendant did not provide customary and appropriate services for all claims.

Although Plaintiff introduced claim-specific evidence by way of examples, the Court infers that major claims mishandling problems occurred more generally throughout the Georgia Fund account. This inference flows from the basic nature of the problem-that, as PMA employees acknowledged, the adjustor's overall caseload was "unmanageable," resulting in "complete file abandonment" in many cases. It is logical that this deficiency did not anecdotally impact just certain individual claims but rather more broadly impacted the overall handling of the account. Indeed, the findings that Hosman made and that PMA admitted were accurate were not limited to individual mistakes but rather related to "the overall performance of PMA Management Corporation."

Given that the deficiencies related not just to individual specifically-identifiable claims, but rather to "the overall performance of PMA Management Corporation," it is logical to consider how the Georgia Fund's claims expenses were impacted overall during this period. And as noted above, there is an obvious and very large increase for claims filed in 2007-2010 as compared to claims filed in earlier years. The Court agrees that this significant "spike" in expenses is the more logical starting point for considering damages than attempting to calculate the specific damages associated with every single workers compensation claim.

Indeed, Defendant's own employees in candid, internal correspondence have acknowledged the difficulty in attempting the quantify the impact of file abandonment on individual claims. As noted above, Kunselman admitted the general point that, due to the adjustor's "lack of managing the disability, settlement values and the ultimate cost of the claims have increased." (Pl. Ex. 38 at 1) (emphasis added). But Regional Claims Manager Jeff Bonner also acknowledged that the financial impact of "the lack of aggressive handling" is "tougher to quantify," and, as to the possibility of missing opportunities to settle at more opportune times in a case, "I have no idea how to quantify that." (Pl. Ex. 11 at 1). That this impact is tougher to quantify in any individual case is obvious. An adjuster's failure to properly work a file may have prejudiced Georgia Fund's position as to that claim in many ways and may have thereby reduced settlement leverage, but it would be exceedingly difficult in many individual cases to find with any confidence that a particular claim could or would have settled for any particular amount at any particular time. There may have been any number of anecdotal reasons why that particular claim might not have settled.

Similarly, it is exceedingly difficult to litigate in retrospect, and make accurate findings, with regard to the costs of overuse of surveillance. McCoy has demonstrated in many instances that Defendant did not follow its procedures and engage in sufficient diligence before resorting to the expensive step of authorizing surveillance. But it is simply guesswork, and close to impossible, to say in retrospect that the surveillance ordered in a given instance had no utility at all.

Thus, while PMA employees admit that "the ultimate cost of the claims have increased" as a whole, the more one attempts to measure damages claim-by-claim, the more guesswork is required. For this reason, the best and most accurate way to calculate damages at least in this case is to look at what actually happened to Plaintiff's overall costs during the time period impacted by Defendant's deficient performance.23

b. The Court's Calculation

All of the above discussion finally leads to the Court's explanation of its calculation. For all of the reasons discussed at length above, the Court generally finds that the "spike" of $3,014,501 for claims filed in 2007-2010 was caused in major part by Defendant's deficient performance. The question remains whether it is possible and appropriate to reduce this figure to reach a more precise amount. The Court believes that a reduction can be applied that results in a more conservative award of damages.

Specifically, the Court believes that it is most appropriate to take the 2007 claims out of the analysis for purposes of calculating damages. For all of the reasons discussed at length above, the Court agrees with Plaintiff that the high costs for claims filed during this year is at least partly attributable to Defendant's claims management performance beginning in 2008. However, the evidence shows that 379 claims filed in 2007 were closed during that same year. (See Pl. Ex. 138 at 52). The Court infers that these claims tended to be less complex and included a disproportionate number of medical only claims. But there were obviously many 2007 claims that Defendant did not handle, and it likely handled very few of the 2007 claims exclusively.

Taking the 2007 claims out of the damages calculation entirely would likely understate Plaintiff's damages. This is so, because of the evidence of the large number of pending 2007 claims taken over by Defendant, and the Court's inference that those "take over" claims tended to be more complicated and expensive. Therefore, excluding these claims for purposes of considering damages likely excludes a large number of high value claims that Defendant handled deficiently.

Nevertheless, the Court finds that excluding 2007 claims from the damages calculation results in a more conservative and more reliable analysis. Among other things, this overbroad exclusion of claims likely more than accounts for the impact that the other factors discussed by Walsh may have had on costs during this time. As discussed above, the Court accepts that some of these factors impacted the increase in costs in 2008-2010, although to an extent that cannot be quantified, and not to the extent suggested by Walsh. Excluding the 2007 "take over" claims thus helps offset whatever unquantifiable influence these factors may have played.24

As noted above, Plaintiff calculates the average claim cost per $100 in payroll from 2000-2006 at approximately $1.11. Applying simple arithmetic to the data in evidence, Plaintiff's payroll in 2008-2010 totaled $6,161,136.25 Simply multiplying this total payroll figure by the average historical $1.11 cost, results in the figure of $6,838,861. This figure is what Plaintiff's claims costs would have been in 2008-2010 had it experienced costs at the same rate as it did in 2000-2006. However, more simple arithmetic from Plaintiff's business records shows that, as of March 2014, Plaintiff's estimated ultimate incurred costs for claims filed in 2008-2010 were actually $9,190,692.26 The difference between the estimated costs for 2008-2010 under PMA's claim handling ($9,190,692) and the costs that would have been experienced at the historical average rates ($6,838,861) is $2,351,831. The Court settles on this figure as its damages award in this case.

4. The Court Declines To Award Plaintiff's Attorney's Fees And Costs

a. The Contract Does Not Provide For Fee Shifting

Plaintiff seeks an award of its attorney's fees in prosecuting this action pursuant to the "Indemnification and Hold Harmless" section of the contract, which provides, among other things, that:

(Pl. Ex. 1 at 14). Defendant disputes that this indemnity provision includes attorney's fees expended in enforcing the contract itself.

The first rule of contract construction is to ascertain the intention of the parties, which is to be discerned from the language of the contract. Service Merchandise Co. v. Hunter Fan Co., 274 Ga.App. 290, 617 S.E.2d 235, 237 (2005). "Additionally, where the contract provides for indemnification, that provision must be construed strictly against the indemnitee...." Inland Atlantic Old Nat. Phase I, LLC v. 6425 Old Nat., LLC, 329 Ga.App. 671, 766 S.E.2d 86, 93 (2014).

The indemnification clause of the contract does not apply here for the simple reason that this is not an action for "negligence or willful misconduct." Indeed, Plaintiff's negligence claims were dismissed because it could not show that Defendant owed Plaintiff any extra-contractual duties. (See Order dated November 23, 2013[107] at 21). Thus, this case is one for breach of contract, which is a category of claims not included in the scope of the indemnity. To be sure, there is overlap between Plaintiff's breach of contract theory and the elements of a professional negligence case. But the claims are different. And these sophisticated parties were clearly capable of including expenses resulting or caused by a breach of contract in the indemnity clause if they intended such expenses to be so included. The parties instead deliberately chose to only include expenses resulting from negligence and "willful misconduct," which the Court interprets as exclusive of all other harms, violations or potential causes of action.

Also, the reference only to "negligence" and "willful misconduct" appears to corroborate Defendant's more fundamental point-that the indemnity clause was not intended to cover costs incurred in a lawsuit between the parties to enforce the contract. Generally speaking, "the purpose of an indemnity clause in a contract is not to protect the parties to the contract from legal action by each other to enforce the contract." SRG Consulting, Inc. v. Eagle Hospital Physicians, LLC, 282 Ga.App. 842, 640 S.E.2d 306, 309 (2006). Rather, typically, "in a contract of indemnity the indemnitor ... promises to indemnify and hold harmless the indemnitee against liability of the indemnitee to a third person, or against loss resulting from such liability." Thomasson v. Pineco, Inc., 173 Ga.App. 794, 328 S.E.2d 410, 411 (1985) (quoting National Bank of Monroe v. Wright, 77 Ga.App. 272, 48 S.E.2d 306 (1948)) (internal quotes omitted). Defendant cites authority interpreting a similar indemnity provision, and finding that it "does not provide for recovery of attorney fees and expenses incurred to enforce the indemnity agreement." Citadel Corp. v. All-South Subcontractors, Inc., 217 Ga.App. 736, 458 S.E.2d 711, 713 (1995).27

Here, not only does the contract omit any agreement to shift attorney's fees and costs in the context of a dispute about the contract itself, but the contract expressly limits the scope of the indemnity to "negligence" and "willful misconduct." Parties to a contract generally cannot sue each other for torts-such as negligence or fraud-relating to the performance of the contract. See Gen. Elec. Co. v. Lowe's Home Centers, 279 Ga. 77, 608 S.E.2d 636, 637 (2005) ("The `economic loss rule' generally provides that a contracting party who suffers purely economic losses must seek his remedy in contract and not in tort."); see also City of Cairo v. Hightower Consulting Engineers, 278 Ga.App. 721, 629 S.E.2d 518, 525 (2006). This suggests that the parties intended to fashion an ordinary indemnity clause focused on liability to third parties, not to each other under the contract.

Thus, the Court finds that the indemnity provision of the contract does not provide for an award to Plaintiff of attorney's fees or any other costs or expenses relating to the prosecution of this action.

b. The Court Declines To Shift Fees On The Basis Of Bad Faith Contractual Performance

Alternatively, Plaintiff sees an award of attorney's fees pursuant to O.C.G.A. § 13-6-11, which provides that:

Fee shifting under this statute is permissive, not mandatory, and therefore can be denied by the factfinder even in cases where some bad faith or other predicate conduct is found. See Multimedia Technologies, Inc. v. Wilding, 262 Ga.App. 576, 586 S.E.2d 74, 80 (2003).

Plaintiff argues that "Georgia law would allow the Court to award fees based solely on PMA's bad faith in mishandling Georgia Fund's workers compensation claims." Pl. Br. [160] at 48. The Court rejects this argument. While the Court finds that Defendant's performance of the contract was deficient, the Court does not generally find that Defendant acted in bad faith in failing to earlier recognize or address the unmanageable workload of the adjuster or other factors contributing to the deficiencies.

Plaintiff's principal argument, however, focuses specifically on the circumstances surrounding the August 2011 audit score supplied to Plaintiff. (Pl. Ex. 53). As discussed above, this report, which indicated a passing 85% score, had been altered significantly from an original version that would have reported a failing 52% score. (Pl. Ex. 50). As also discussed above, the Court has found that at least some of the changes that were made were done for the purpose of being able to report a better score to Plaintiff.

The Court accepts for purposes of this discussion that there was some bad faith involved in the creation and transmission of the report. But the Court finds that, regardless, any bad faith that contributed to the changes to the original draft does not warrant an award of attorney's fees in light of the entire context of this case. Overall, Defendant appeared to act responsibly and diligently in response to the allegations of deficient performance. Defendant cooperated with the outside auditor (Hosman), bore the costs of her services, engaged in a corrective plan including assigning additional adjusters and engaging in more proactive review, all of which succeeded in bringing costs back down to their historical norms. In light of this broader picture, the alleged bad faith of some employees with regard to some aspects of the revisions of this document does not, in the Court's judgment, warrant a wholesale award of fees.

III. CONCLUSION

For the reasons discussed above, the Court finds in Plaintiff's favor as to liability on Count I and awards damages in the amount of $2,351,831.00. The Clerk is DIRECTED to enter judgment.