Crash Course 2020 Is ‘IT’S TIME,’ Intended to Draw Attention to What • Insights We Deliver Must Be Worthy of Trust, Predictable, and Actionable

2020 CRASH COURSE

25 2020 COURSE CRASH

Years

CCC Information Services Inc. 222 Merchandise Mart Plaza, Suite 900 Chicago, IL 60654

(800) 621.8070

| IT’S TIME CRASH COURSE

IT’S2020 TIME Prepared by Susanna Gotsch Prepared by Susanna Gotsch

Facebook.com/cccis

@cccinfoservices

Youtube.com/CCCInfoServices cccis.com/LinkedIn

© 2020 CCC Information Services Inc. All Rights Reserved. The information and opinions in this publication are for general information only, are subject to change and are not intended to provide speciﬁ c recommendations for any individual or entity. Although information contained herein has been obtained from sources believed to be reliable, CCC does not guarantee its accuracy and it may be incomplete or condensed. CCC is not liable for any typographical errors, incorrect data and/or any actions taken in reliance on the information and opinions contained in this publication. Note: Where CCC Information Services Inc. is cited as source, the data provided is an aggregation of industry data related to electronic appraisals communicated via CCC’s electronic network or from total loss valuations processed by CCC. Where Auto Injury Solutions (AIS), a CCC Company is cited as a 1

source, the data provided is an aggregation of industry data collected from claims data communicated via Auto Injury Solutions, Inc.’s electronic network. TOWELCOME THE 25TH ANNIVERSARY OF CRASH COURSE.

Anniversaries, in my view, are not a measure of time. Anniversaries measure commitment. I’m proud of the 25-year commitment CCC has made to the industry through the publication of Crash Course. Githesh Ramamurthy is Chairman & Chief The 2020 IT’S TIME edition provides insights about where we’ve been and about the key drivers shaping Executive Officer of CCC. our future — AI, ADAS, driverless vehicles, data access, digital consumers, quality repairs, and more. Without question, the data clearly shows we are entering the most exciting time in our history.

I think back 25 years to 1995. Netscape, Windows 95 and Internet Explorer were being introduced, setting the foundation for the future of digital connections and communications. At the same time, CCC was launching a new digital platform to connect our industry in new and exciting ways. A byproduct of this digital transformation was the creation of roughly 10 times more data, providing more granularity and the foundation of deeper insight. At the same time, the possibilities of AI are increasing. Over the last five years, we have worked extensively to establish hundreds of models. AI at its core requires a massive This transformation marked one of several key milestones in our company’s history. At that time, CCC amount of information to inform models, and its power can only be unleashed once decided to harness the power of, and invest in, data with the explicit purpose of advancing the industry integrated into real-time processes. Today, after years of investment, AI guides the and our customers. We organized around establishing clear practices and procedures to provide front-line people who manage claims, conduct repairs, or interact with customers. By trustworthy and actionable information to customers, helping to improve operations and maintain reducing time required to manage routine operations, people now have more time to differentiation in ways that had not been possible before. deliver high-touch experiences that exceed customer expectations. This effort has been ongoing for 25 years, resulting in three core principles: Our theme for Crash Course 2020 is ‘IT’S TIME,’ intended to draw attention to what • Insights we deliver must be worthy of trust, predictable, and actionable. is today uniquely possible by connecting various parts of the process and applying data-driven insights from upstream and downstream steps. It is also a testament to • Great care is taken to be as free of confirmation bias as possible. the people at CCC and our ongoing commitment to serve and connect our customers • Feedback from these insights is used to continuously improve the products and services we deliver — insurers, repairers, OEMs, parts providers, fleets, rental companies, and more. to our customers. From our team of data scientists to our customer-care teams, to individuals like Over the years, the depth and breadth of information has expanded, enabling the delivery of sufficiently Susanna Gotsch — who has been at the helm of sharing insights through Crash Course high-confidence predictions, including driving data, mobile usage patterns, estimatics, parts, for 25 years — CCC remains steadfast in our commitment to thoughtfully connect the demographic data, total loss valuations, estimate audits and reviews, medical, and supply chain. Today, industry and to empower what comes next. tens of millions of claims are routed based on a predictive view of the likely outcome in real time at first We are proud to move forward with you and deeply appreciate the tremendous trust notice of loss and countless repairs are expedited through real-time connections and data flow. The time you place in us. it takes to make decisions is shrinking.

TABLE OF CONTENTS

07 SEC 1: TIME IS MONEY Reconsidering Personal Mobility

55 SEC 2: TIME TO INVEST Modernizing Repair Capabilities

97 SEC 3: TIME FOR SERVICE Humanizing the Experience

109 SEC 4: TIME FOR EFFICIENCY Simplifying the Complex

121 SEC 5: TIME FOR DATA Informing and Inciting Action

155 SEC 6: TIME TO PREPARE Embracing the Next Wave of Tech

5 CRASH COURSE 2020 COURSE CRASH

| TIME IS MONEY IS TIME TIME IS MONEY RECONSIDERING PERSONAL MOBILITY

Talk of the end of the auto industry as we know 01 it continues, even as automakers produce products of unprecedented diversity in terms of body styles, engine propulsion, safety and comfort features.

We no longer talk just about automobiles; it’s all about personal mobility — whether that is by walking, biking, using a scooter, ride-share service, public transportation, or flying car. Change has come to how consumers purchase and use vehicles, and more change is coming.

7 CRASH COURSE 2020 COURSE CRASH

With more consumers opting for alternative modes of mobility, the auto industry is facing a need U.S. New and Used Vehicle Sales CY1998 to CY2019 (in millions) (FIGURE 1) to reinvent itself or face shrinking revenues. The cost to own and operate a vehicle, as well as the SOURCE: WWW.AUTONEWS.COM, COX AUTOMOTIVE, WWW.EDMUNDS.COM. cost of wasted time, are some of the reasons why consumers are looking for alternatives. And today, more than ever, time is money. Time spent looking for a parking space, driving in bumper- New Sales Used Sales to-bumper traffic, or driving a truck thousands of miles to transport goods is time that could be 70.00 repurposed in the future. In this section of Crash Course, “Time is Money,” we will explore where we 60.00 are on this continuum. | TIME IS MONEY IS TIME 50.00

40.00 U.S. Auto Sales Slow, but Growth in Vehicles in Operation Continues 30.00

20.00 CY 2019 was the fifth consecutive year that U.S. new vehicle sales surpassed 17 million. Used vehicle sales were also strong, closing out the year with an estimated 40 million units sold (see Figure 1). 10.00 CY 2019 new sales were down over one percent for the year, coming in at 17.1 million versus 17.3 million in CY 2018, and 17.5 million in CY 2016.1 0.00

CY1998CY1999 CY2001 CY2010CY2011CY2012CY2013CY2014CY2015CY2016CY2017CY2018 Most analysts predict slower auto sales in the coming years, as slower job growth and consumer CY2000 CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009 CY2019E spending, pent-up demand and higher new vehicle prices are forcing consumers to forgo a vehicle purchase or purchase a used vehicle versus new. The growing popularity of larger, more expensive vehicles and new vehicle technology drove up the average new vehicle MSRP to over $35,000 in CY 2019, and many consumers opted for longer loan term lengths to afford a new vehicle. Average New Vehicle MSRP and Average Used Vehicle Transaction Price CY2005- CY2019 (FIGURE 2) | SOURCES: NADA DATA AND EDMUND.COM USED VEHICLE MARKET REPORTS 2014, 2018, 2019.

Higher used-vehicle demand and larger volumes of clean late-model used vehicles also lifted the NADA Average Selling Price of New Vehicles Sold Edmunds.com Avg Used Tranx Price average used vehicle transaction price (see Figure 2). Through the first nine months of CY 2019, prices on vehicles aged eight years and newer were up 1.7 percent versus the same period in 2018.2 $35.0 Analysts forecast used vehicle prices for vehicles aged eight years and newer will rise between 1 $35.0 and 1.5 percent in 2020.3 $30.0 Fewer passenger vehicle miles driven, as well as better-built vehicles, mean vehicles last longer $25.0 than ever before; data from IHS Automotive reveals the average age of vehicles on the road in the U.S. was 11.8 years in CY 2019 (see Figure 3).4 The aging vehicle fleet over the last decade was $20.0 driven largely by the addition of new vehicles that did not displace old vehicles that remain in the fleet despite being driven less (seeFigure 4).5 The normal displacement of older cars by newer $15.0 cars was slowed further during the Great Recession, leading to a larger number of older cars in use $10.0 today. $5.0

$0.0

CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2005 CY2006 CY2007 CY2008 CY2009 CY2019E

With auto sales trending about 17 million annually, the overall vehicle fleet continues to grow at IHS Automotive Passenger Cars and Light Trucks in Operation in the U.S., Annual about 2 percent annually (see Figure 5). As overall vehicle quality improved, vehicle scrappage Growth Rate and Percent Scrappage CY2002-CY2019 (FIGURE 5) | SOURCE: IHS AUTOMOTIVE rates remained low, leading to a slow turnover of the overall U.S. vehicle fleet and older fleet. Growth in overall vehicles in operation is a key driver of steady growth in the collision repair Total Passenger Car and Light Trucks (GVW 1-3 only) in Use as of Jan 1 of CY (in millions) industry, countering more recent slowdown in auto claim frequency. Percent increase in vehicles in use from prior year

IHS Automotive Average Age of Vehicles in Operation in the U.S. CY2002-CY2019 Scrappage % of Passenger Vehicles in Use Starts of CY |

(FIGURE 3 ) | SOURCE: IHS AUTOMOTIVE MONEY IS TIME 290 7.0% Average Age Car Average Age Light Truck Average Age Total 280 6.0% 14.00 5.0% 270

12.00 4.0% 260

10.00 3.0% 250

8.00 2.0% 240 6.00 1.0% 230 0.0% 4.00 Total Passenger Car and Light Trucks and Light Car Passenger Total 220 -1.0% 2.00 (GVW 1-3 only) in Use as of Jan 1 of CY (in millions) in Use as of Jan 1 CY only) 1-3 (GVW 210 -2.0% of VIO Percent Change in VIO and Scrappage Percent 0.00

CY2003 CY2004 CY2005 CY2006 CY2007 CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY20117 CY2018 CY2019

CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2017 CY2018 CY2005CY2006CY2007CY2008CY2009 CY2019E CY2019E

Average Annual VMT for 1-Year-Old, 10-Year-Old, and 20-Year-Old Vehicles in CY2017 (FIGURE 4) | SOURCE: SAFE ANALYSIS BASED ON DATA FROM 2017 NHTS. HTTP://ENERGYFUSE.ORG/ARE-AMERICANS-DRIVING-OLDER-CARS- Headwinds for Industry "ACES" in 2020 and Beyond OR-JUST-LEAVING-THEM-IN-THE-DRIVEWAY/ 14,000 Looking to CY 2020 and beyond, the auto industry is facing some headwinds. Much uncertainty remains in place regarding tariffs and regulations — and uncertainty about the outcome of the 2020 12,000 presidential election certainly doesn’t help. Shifting demographics and consumer expectations require companies to be nimbler than ever, not only in product development, but also in the concepts brought to market and how they are marketed. As vehicles become more expensive but last longer than ever, new vehicle sales will slow, and automakers will explore new revenue streams in order to recreate 8,000 themselves as personal mobility companies as opposed to simply car companies.

In their efforts to do so, automakers continue to invest huge amounts of money in the four megatrends "ACES" — autonomy, connected car, electrification and shared. Growing acceptance of just how hard it is to develop a fully autonomous vehicle and the economic feasibility of operating a shared vehicle fleet have resulted in many automakers increasing their focus on electrification and connected car.

1-Year Old 10-Year Old 20-Year Old

This is an industry that has seen a lot. A couple “ of world wars. The Great Depression. The Great China 2.8% Recession. Labor strife. Oil embargoes. But now it |

faces technological changes that threaten to obsolete MONEY IS TIME

all the infrastructure and intellectual property they’ve U.S. 1.4% invested in."

John McElroy

EU 0.8% SOURCE: “WE ARE EMBARKING ON A DANGEROUS DECADE.” WWW.WARDSAUTO.COM , JAN 2, 2020.

Electric Vehicles Charge Ahead Connected Vehicle Counts Grow Demand for electric vehicles is driven largely by regulations to curb emissions in China and the EU. LMC Automotive projects sales of battery-powered vehicles will quadruple by CY 2025 and Demand for greater digital capability and wireless connectivity in the vehicle itself has led account for 4.8 percent of the U.S. auto market versus just 1.3 percent today.6 And while electric many automakers to introduce connected vehicle technology into their fleets. Through vehicle vehicle sales have grown, they still represent a small percent of sales (see Figure 6). It wasn’t until connectivity, or telematics, vehicle performance data is shared over the air to assess vehicle mid-2018 that the U.S. market achieved the 1 million EV-mark, where each sale included at least health; navigation and driving data can be used to assess vehicle acceleration, braking patterns, $7,500 in tax incentives.7 The cost of electric vehicles remains a challenge for automakers: Battery location and route history; data is shared for vehicle-to-vehicle and vehicle-to-infrastructure costs remain high, and the build-out of a widespread charging infrastructure must occur. A recent communication; emergency alerts can be generated in the case of an accident or emergency; and report from the MIT Energy Initiative warned that electric vehicles’ reliance on lithium-ion batteries numerous services are streamed directly into the vehicle.13 Ford uses connected vehicle technology might prevent them from ever achieving the same sticker price as conventional gasoline-powered to help identify a vehicle’s features that passengers do not interact with, and subsequently removes vehicles. While the cost of lithium-ion batteries has declined steadily, that decline is expected those features in later models to help improve overall vehicle affordability.14 Telematics can be to slow over the next several years as they approach limits set by the cost of the raw materials.8 either a factory/OEM embedded system, a consumer electronics device such as a smartphone, With battery costs accounting for about one-third of their total cost, an electric vehicle’s sticker an aftermarket telematics control unit (dongle that plugs into the on-board diagnostics port), or price is expected to remain thousands of dollars more than a comparable gas-fueled vehicle a hybrid device that includes an embedded telematics control unit and a consumer electronics — a tough proposition for consumers despite potential for lower lifetime cost of ownership.9 device.15 Today, connectivity via a smartphone accounts for approximately 50 percent of the Despite these challenges, automakers are forging ahead with new models at various price points. market, but embedded systems are rapidly being brought to market. According to March CY 2019 LMC Automotive predicts 110 models available for sale in the U.S. market alone by CY 2025.10 data from Statista, there were 49.7 million connected cars with embedded systems in the U.S. in CY AlixPartners LP projects automakers will spend $225 billion to develop new electric-vehicle models 2019, up from 29.3 million in CY 2017 (see Figure 7). over the next several years.11 Many are creating closer ties with battery companies to lower costs and secure future battery supplies — for example, GM with LG Chem; and Toyota with Panasonic.12

Battery-electric vehicles as a share of region’s CY2018 light vehicle sales volume

Connected Car U.S. Stock and Percent of Households with Connected Car (FIGURE 7) other companies in the oil and gas industries may look to make future investments in car-sharing SOURCE: STATISTA, MARCH 2019. (HTTPS://WWW.STATISTA.COM/OUTLOOK/320/109/CONNECTED-CAR/UNITED- services, particularly with the emergence of electric vehicle fleets.24 Expect car-sharing to develop STATES?CURRENCY=USD#MARKET-ARPU) further in the coming years.

120 73.6% 80%

64.9% 70% “Not in My Lifetime” 100

56.2% 60% |

80 47.7% MONEY IS TIME 50% In CY 2019, many of the technology and car companies working on the development of 39.3% autonomous vehicles (AV) acknowledged publicly just how difficult the development of a fully 60 40% 31.1% autonomous (Level 5) vehicle truly is. 95.7 23.4% 30% 40 83.8 72.1 60.8 20% 49.7 20 39.1 I stepped way back [on] this idea of Level 5. I’ve 29.3 10%

Stock of Connected Cars (in millions) Cars of Connected Stock 0 0% really given up. I don’t even know if that will happen Penetration Rate per Total # of Households per Total Rate Penetration CY2017 CY2018 CY2019 CY2020 CY2021 CY2022 CY2023 in my lifetime." Car-Sharing Services Reset “ Steve Wozniak Co-founder Apple Inc.

Car-sharing services also saw a great deal of activity in CY 2019. BMW Group merged its urban SOURCE: BOND, VINCE, JR. “APPLE CO-FOUNDER: ‘I’VE REALLY GIVEN UP’ ON LEVEL 5.” WWW.AUTONEWS.COM, mobility services with Daimler AG’s Car2Go, grouping the numerous services offered to their 60 OCTOBER 28, 2019. million users under five new ventures: Reach Now, Charge Now, Park Now, Free Now, and Share Now.16 Additional OEs — including VW, Toyota and Hyundai — expanded investment in mobility Daimler chairman Ola Kaellenius, Waymo chief executive John Krafcik, Honda CEO Takahiro services in CY 2019. Others scaled back or discontinued services, such as GM’s scale-back on its Hachigo, and Steve Wozniak, the co-founder of Apple Inc., have made public statements about car-sharing service Maven, Car2Go’s discontinuation of its services in the U.S., and Ford’s lost their tempered expectations. While many prognosticators had predicted the arrival of Level 4 investment in the van-based shuttle service Chariot.17 The automakers’ goal is to better understand and Level 5 autonomy by CY 2025, most have shifted their timelines out to the CY 2030’s and CY the potential of tech-enabled mobility services to ultimately convince consumers to give up their 2040’s. personal cars en masse.18 The challenge is that car-sharing requires a critical mass of population density to work; large distances between a user and an available shared car makes the service undesirable,19 and the cost of setting up and running a mobility fleet is expensive. The reality is that driving is a social process that involves numerous complex interactions with other drivers, pedestrians, cyclists, and all types of changing road conditions.25 Common sense and generalized knowledge are relied upon in many everyday traffic situations — things very difficult A survey of 10,000 respondents from the U.S., China, France, Germany and the U.K. conducted by for a software program or even AI to mimic. Today it’s estimated that the companies developing OC&C Strategy Consultants found car-sharing, short-term renting, and taxis are largely unappealing autonomous vehicle technology have developed 80 percent of needed technology to put self- to consumers in most Western countries because they want a vehicle “when and where they need driving vehicles into routine use; but the remaining 20 percent (referred to as ‘corner cases’) it” and “it’s too much of a hassle to pick up a vehicle.”20 Results from this survey also showed car that could reliably anticipate what other drivers, pedestrians and cyclists are going to do is much ownership remains critical to Americans — 84 percent say having their own car is “essential to more difficult.26 Other situations such as knowing not to follow a vehicle too closely if it looks like getting around,” — highest among the five countries, and 64 percent of American respondents also it’s looking for a parking spot (an example of a ‘micro maneuver’) are also very hard to program said they expect to own their own vehicle in the future, ahead of all countries except China.21 around.27 Add the challenges inherent in having AVs driven alongside manually driven vehicles for many years to come, and the timeframes potentially extend further. There are nearly 300 million In CY 2016, Frost & Sullivan predicted the number of car-sharing users globally would grow from vehicles on the road in the U.S. today alone with an average age of 11.8 years, so roads will be 22 10 million in CY 2017 to 36 million by CY 2025. While that represents impressive growth, it is still shared for quite some time. small compared to the overall global population which the U.N. anticipates will reach 8.1 billion by CY 2025.23 Essentially the 36 million car-sharing users would only be 0.4 percent of the global population. Poland’s state-owned public power company now is an investor in the car sharing services company4Mobility, and there is speculation now that power companies and

Ultimately, one of the challenges of self-driving is that Yet more money and effort continue to be invested in AVs. More companies announced partnerships around autonomous vehicle development, in part to offset the billions of dollars being you’re trying to predict human behavior, and human spent on AV research and development, but also to pool resources: VW and Ford invested jointly behavior tends to not fall into rational agent models in Argo AI; Honda invested in Cruise, GM’s autonomous vehicle operations; Mercedes and BMW agreed to work together; FCA and JLR are working with Waymo; and Toyota and Volvo are working we have for game players … One of the reasons I think 34

with Uber. Autonomous logistics focus on the ‘last-mile’ has seen increased investment, with “ it’s going to be challenging to deploy [AVs] is because FedEx, Ford and Domino’s, Kroger and others looking to solve the challenge of delivering goods at | you’re going to have to get these predictions right when the last mile, which is estimated to be nearly a third of an item’s total delivery cost.35 MONEY IS TIME

traveling at high speeds in dense urban areas.” While the timeline has been pushed out, the advances being made in sensor, camera, lidar, radar, AI, software, infrared cameras and sensors, cybersecurity and other technologies continue, Matthew Johnson-Roberson suggesting AVs are inevitable — it’s just a question of when. Assistant Professor of Engineering at University of Michigan and co-founder of Refraction AI And while fully autonomous vehicles are presented as the best way to get to zero road fatalities SOURCE: HTTPS://WWW.WIRED.COM/STORY/TEACHING-SELF-DRIVING-CARS-WATCH-UNPREDICTABLE-HUMANS and accidents, they’re unlikely to be the only way to get us there. Research from the Casualty Actuarial Society of the National Motor Vehicle Crash Causation Survey concluded that driverless Real-world data already shows consumers don’t clearly understand how the various advanced vehicles would potentially address only 78 percent (not 93 percent) of accidents if they could not driver assistance systems (ADAS) work — essentially the building blocks to full vehicle autonomy. overcome weather, vehicle errors and inoperable traffic control devices.36 So far, the best ADAS The range in capabilities, implementation, and feature names used across the OEs for the various system developed for hail is the garage. ADAS technology certainly doesn’t help. The fact that most of these systems are considered Level 1 or Level 2 per the SAE autonomous vehicle definitions speaks to increasing challenges ahead. A study conducted by the Insurance Institute for Highway Safety (IIHS) surveyed more than 2000 Despite New Mobility Alternatives People drivers on five Level 2 systems available at that time. The study found the feature names alone can lead to misunderstandings about the technology’s capability.28 In its ongoing investigations of Still Drive vehicle crashes that included ADAS, the National Transportation Safety Board (NTSB) examined and identified the role of automation complacency — where overreliance on driver-assistance Among the numerous indicators that prognosticators pointed to when arguing that the era of 29 systems led to driver inattention. Finally a study from the AAA Foundation for Traffic Safety personal vehicle ownership was coming to an end was a decline in the share of individuals getting assessed people’s understanding of AV technology, their expectations and concerns about AVs and their license, buying a vehicle, and a decline in overall miles driven. All of these declines did happen, their rationales behind their distrust and discomfort toward AVs. The study found people perceived but the reasons why may be more nuanced than originally thought. higher levels of vehicle automation as more effective in preventing crashes related to specific driving behaviors and situations, but concern increased as levels of automation increased due to Licensed drivers as a percent of the total population fell during the Great Recession and several unfamiliarity with the technology and perceived unreliability of it.30 years after, but the two latest years of data suggest a reversal of that trend. Data from the U.S. DOT Federal Highway Administrations’ 2017 Highway Statistics reveals growth in drivers and vehicles has Many challenges remain before automakers can address every potential accident scenario, at outpaced growth in population since CY 2015 (see Figure 8).37 The ratio of drivers to population all traveling speeds and in a consistent manner. And this is just the technology; how each driver as of CY 2017 was 69 percent up from the recent low of 66.9 percent in CY 2013. And while there responds to the technology will vary as well, with three key challenges of mode confusion, role was an increase in the percent of licensed drivers to overall population for all age groups between 31 confusion, and misplaced trust likely to further complicate the path towards full vehicle autonomy. CY 2014 and CY 2016, in the next two years drivers ages 16-18 and those ages 25-29 experienced a slight decline of less than 1 percentage point. Drivers ages 65-69, 55-59, 30-34, and 60-64 saw the To date, there are no clear regulations at the state or federal level governing the safety testing and largest increases (see Figure 9). Recent research from the auto analyst firm Benchmark reveals deployment of autonomous vehicles. In October 2019 41 states had enacted legislation or signed millennials are getting licensed to drive at the highest rate in 40 years and predicts licensed drivers executive orders regulating the testing and use of autonomous vehicles, and in September NTHSA will grow another 12.5 million in the U.S. over the next five years.38 released new federal guidelines, but they are only voluntary.32 After their investigation into the Uber autonomous test car that killed a pedestrian in Tempe, Arizona, in March CY 2018, the NTSB called out NTHSA for not providing enough direction to autonomous vehicle development.33

Annual Percent Change in U.S. Population, Drivers, and Registered Vehicles Maybe They’re Not So Different After All CY1960 to CY2018 (FIGURE 8)

SOURCE: U.S. DOT FHWA POLICY & GOVERNMENT AFFAIRS HIGHWAY POLICY INFORMATION HIGHWAY STATISTICS, 1995 TO 2018. One of the key reasons people were predicting the end of personal vehicle ownership and a shift

7.0% to alternate modes was the large decline in licensed driver data among the younger age groups. 6.0% This was a clear sign that the millennial generation had much less interest in vehicle ownership than

5.0% prior generations. Subsequent analysis suggests that there are additional or alternative reasons for | 4.0% lower licensing levels among younger generations. Several of these studies are discussed below. TIME IS MONEY IS TIME 3.0% 2.0% 1.0% One study comes from the Board of Governors of the Federal Reserve System. This study 0.0% compared the socioeconomic and demographic characteristics of Millennials versus prior Population Drivers Vehicles -1.0% generations. Accounting for their age and other factors, such as differences in income, saving, -2.0% and consumption, this study found Millennials don’t appear to have preferences for consumption -3.0% that differ significantly from those of earlier generations.41 A second study by Christopher R.

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 Knittel and Elizabeth Murphy, “Generational Trends in Vehicle Ownership and Use: Are Millennials 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Any Different?” found similar results. When differences in age, income and other demographics 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 were controlled for versus prior generations, their research revealed that Millennials don’t own fewer vehicles, and actually drive more miles per year than Baby Boomers (see Figure 10).42 It is worth noting that Millennials accounted for nearly half of U.S. home-mortgage originations as of December 2019.43

U.S. Licensed Drivers as Percentage of Their Age-Group Population (FIGURE 9) SOURCES: SIVAK AND SCHOETTLE, UMTRI; U.S.DOT FWHA HIGHWAY STATISTICS 2016-2017 Millennials Have Similar VMT and Vehicle Ownership After All (FIGURE 10) SOURCE: CHRISTOPHER R. KNITTEL AND ELIZABETH MURPHY. GENERATIONAL TRENDS IN VEHICLE OWNERSHIP AND USE: ARE MILLENNIALS ANY DIFFERENT? NBER WORKING PAPER NO. 25674, MARCH 2019. CY1983 CY2008 CY2011 CY2014 CY2016 CY2017 100.0% 90.0%

80.0% Conclusions NBER study of Millennials’ vehicle decisions 70.0% and whether they dier from those of previous 1. Results conﬁrm Millennials’ observed decrease 60.0% generations. in vehicle ownership & VMT arises from dierences in demographics. 50.0% Focus on two main facets of personal mobility: 40.0% 2. When Millennials & Baby Boomers with 1. Vehicle ownership, measured by how many similar demographics are compared, Millennials 30.0% vehicles a given household owns. have higher ownership rates & VMT. 20.0% 2. Vehicle usage, measured by annual vehicle 3. Millennials are altering life-choices that aect 10.0% miles traveled (VMT). vehicle ownership. 0.0% Control for demographic and macroeconomic Age 16 Age 17 Age 18 Age 19 Ages Ages Age Age Age Age Age Age Age Age Age >70 4. However, the net eect of these endogenous 20 - 24 25 -29 30 - 34 35 - 39 40 - 44 45 - 49 50 - 54 55 - 59 60 - 64 65 - 69 variables such as income, household size, choices is to reduce vehicle ownership by less location, education, sex, race, marital status, than one percent (with eects larger than two number of children, etc. percent statistically ruled out).

At the same time, the percentage of households living car-free continues to hover just below 10 percent. Data for CY 2018 from the American Community Survey’s “Household Size by Number of Vehicles Available” reveals only 8.5 percent of households in the U.S. were car-free, versus 14.2 Finally, a third study further combats the theory that generational differences are driving percent for one-car families; 18.3 percent for one-car one singles; and 58.9 percent for more than change in vehicle miles traveled, showing instead that changes in miles driven are impacted two cars.39 This last category includes all households with two or more vehicles — including single most by economic factors. The Insurance Information Institute analysis of a study in the Energy households, where a surprisingly large percentage own more than one vehicle — specifically 16 Journal titled “Explaining the Evolution of Passenger Vehicle Miles Traveled in the United percent of singles and 4.5 percent of all households in CY 2018.40

States” highlighted the study’s conclusion that it was changes in demographics and economic Percent Change in Miles Driven CY2019 through November versus Same Period Prior characteristics, rather than changes in household driving habits that largely explain changes in Year by State (FIGURE 12) SOURCE: USDOT OHPI FHWA VMT [vehicle miles traveled] between CY 1995 and CY 2015.44 The study also predicts the average annual VMT growth rate from CY 2015 to CY 2025 would be 0.9 percent, based on the opposing factors of rising income (which increases VMT) and aging population (which decreases VMT).

Overall miles driven in the U.S. continued to grow in CY 2019. For the rolling 12-month period | ending October CY 2019, miles driven are up 1.0 percent versus the prior period, versus only 0.4 MONEY IS TIME percent for full year CY 2018 (see Figure 11). Fewer than a third of individual states saw a decline in miles driven between CY 2018 and CY 2019 (see Figure 12), and most of those states saw lower rates of population growth over the last decade (see Figure 13).

Annual Vehicle-Distance Traveled — Moving 12-Month Total on All Roads (Million

Miles) and Percent Change from Prior Year Jan ‘91 — Nov ‘19 (FIGURE 11) >-1% SOURCE: FRED® MOVING 12-MONTH TOTAL VEHICLE MILES TRAVELED, WWW.RESEARCH.STLOUISFED.ORG -1% to -0.5% -0.4% to 0%

3,500 5.0% -0.1% to 0.9%

3,000 4.0% 1% to 2% 3.0% 2% to 4% 2,500 2.0% 2,000 1.0% 1,500 Projected Growth: Census 2010-Census 2020 (FIGURE 13) 0.0% SOURCE: WILLIAM H FREY ANALYSIS OF CENSUS BUREAU POPULATION ESTIMATES, RELEASED DEC 30, 2019 | NOTE: APRIL 1 2020 POPULATION 1,000 IS CALCUATED BY ASSUMING THAT THE GROWTH RATE FROM JULY 1, 2018 TO JULY 1, 2019, REPORTED IN CENSUS ESTIMATES, CONTINUES -1.0% PERIOD PRIOR YR THROUGH APRIL 1, 2020 500 -2.9% MILES DRIVEN, IN BILLIONS

0 -3.0% %CHANGE IN MILES DRIVEN VS SAME Mar-11 Mar-16 Mar-01 Jul-19 Jul-14 May-15 Sep-13 Sep-18 May-10 Mar-96 Nov-17 Nov-12 Mar-06 Jan-17 Jan-12 Jul-99 Jul-94 May-95 Jul-09 Jul-04 Sep-93 May-05 Sep-98 Sep-03 Nov-97 Sep-08 Nov-92 May-00 Nov-07 Nov-02 Jan-97 Jan-92 Jan-07 Jan-02 WA, 14.3% Miles Driven (in billions) %Chg Same Period Prior Yr ME,... MT, 8.6% ND, 13.7% OR, 10.8% MN, 6.8% V.. ID, 15.7% N.. SD, 9.2% WI, 2.6% NY, 0.1% WY, 2.8% MI, 1.1% Yet a key proxy for vehicle accident exposure is how many miles an individual vehicle drives per IA, 3.7% PA, 0.8% NE, 6.3% NV, 15.5% IL, -1.5% OH, 1.4% year. Analysis of claims data shows the average annual miles per loss vehicle (as measured by the UT, 17.4% IN, 4.3% odometer of the loss vehicle divided by the vehicle age) continues to trend downward, except CA, 6.2% CO, 15.5% WV,... KS, 2.2% MO, 2.7% VA, 7.0% for the newest model year vehicles, after experiencing a big jump between CY 2013 and CY 2014 KY, 3.1% NC, 10.8% (see Figure 14). A higher share of individuals who work at home also is likely a culprit — the TN, 8.3% AZ, 15.3% OK, 5.8% number of workers aged 16 years or older whose primary means of transportation to work was NM, 2.0% AR, 3.7% SC, 12.4% AL, 2.8% % Change ‘worked at home’ grew from 3.6 percent in CY 2005 to 5.1 percent in CY 2017.45 Nearly 4 million GA, 10.4% MS, 0.2% 17.8% U.S. employees worked from home at least half the time in 2015 — roughly 3 percent of the U.S. TX, 16.4% workforce, and a number that grew 115 percent between CY 2005 and CY 2015, and grew nearly 10 LA, 2.4% AK, 2.6% FL,... times faster than the rest of the workforce according to a study by Global Workforce Analytics and -3.8% Flexjobs (see Figure 15).46

Annual Mileage by Vehicle Age Group (All Vehicle Conditions — All Loss Categories Is Growth in E-Commerce Changing the Makeup of CY02-CY19) (FIGURE 14) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY Miles Driven? CY2002 CY2003 CY2004 CY2005 CY2006 CY2007 CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 People still buy cars but drive them less. Access to alternate methods of mobility are only part of

the reason. People are spending more time than ever at home, streaming movies, ordering dinner

20,000 in, telecommuting, and shopping online (see Figure 16). The explosive growth of online shopping |

18,000 has created an experience of significant convenience for consumers. Even consumers shopping for MONEY IS TIME autos now do much of their preliminary shopping online. 61 percent of car shopping is now spent 16,000 online, with 21 percent of that ultimately spent on the website of the dealer the vehicle is purchased 14,000 from, and 60 percent of dealership website traffic now coming from mobile devices.47 12,000

10,000

8,000

6,000 4,000 The curb today in most American cities is 2,000 overwhelmingly devoted to single-occupancy personal- 0 vehicle parking, which is simply unsustainable." Current Yr Vehicles 1-3 Yr Vehicles 4-6 Yrs Vehicles 7 Yrs Plus All Vehicles Ages “ A United Parcel Service Inc. Spokesman SOURCE: CALVERT, SCOTT. “NEW GRIDLOCK IN AMERICA: THE FIGHT FOR CURB SPACE.” WWW.WSJ.COM, DEC. 13, 2019.

Telecommuting Growth Since CY2005 (FIGURE 15) SOURCE: SPECIAL ANALYSIS OF U.S. CENSUS DATA CONDUCTED FOR FLEXJOBS BY GLOBAL WORKPLACE ANALYTICS. According to SJ Consulting Group, U.S. consumers will get 8.6 billion packages delivered Non-telecommuters Telecommuters domestically in 2019, up 150 percent from 2010.48 While some of this means fewer trips and miles 140% overall, some of this results in a transfer of the mileage elsewhere — i.e., delivery services and ride- 120% hailing services.

100% According to the ATA, U.S. trucks moved 10.8 billion tons of freight in CY 2017 — a number that 80% equates to roughly 30 pounds of goods for every person in the U.S.49 In the 2019 Annual Energy 60% Outlook, the U.S. Energy Information Administration projected truck miles traveled (the dominant mode of freight movement in the U.S.) will grow by 52 percent from 397 billion miles in CY 2018 to 40% 601 billion miles in CY 2050.50 The U.S. EIA projects growth in vehicle miles traveled between CY 20% 2017 and CY 2050 will be highest for commercial light trucks, followed by freight trucks, and finally light-duty vehicles (see Figure 17).51 0% CY2006 CY2007 CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 One thing is certain — there are more vehicles in operation in the U.S. today than ever before, operating alongside a growing number of trucks of all sizes in operation to deliver all the goods people are now ordering online.

Online Shopping Grows in Popularity (FIGURE 16) Share of U.S. Cumulative Miles Driven by Road Type CY2004-CY2019 (FIGURE 18) SOURCE: “FHWA NTSA BRIEF: CHANGES IN ONLINE SHOPPING TRENDS, 2017 NATIONAL HOUSEHOLD TRAVEL SURVEY.” AUGUST 2018. HTTPS:// SOURCE: HTTPS://WWW.FHWA.DOT.GOV/POLICYINFORMATION/TRAVEL_MONITORING/TVT.CFM NHTS.ORNL.GOV. Total Rural Total Urban 2009 NHTS 2017 NHTS CY2019 30.4% 69.6%

CY2018 30.1% 69.9% Made at least one online purchase in the last 30 days 43% 55% CY2017 30.2% 69.8% 43% 56% | Urban respondents CY2016 30.1% 69.9% TIME IS MONEY IS TIME Rural respondents 42% 51% CY2015 30.5% 69.5% CY2014 31.7% 68.3% Average number of online purchases made in last 30 days 3 5 CY2013 33.1% 66.9% CY2012 33.1% 66.9%

CY2011 33.1% 66.9%

CY2010 33.1% 66.9% U.S. Vehicle Travel Projections (Reference case) — in billions of vehicle miles traveled CY2009 33.6% 66.4% (FIGURE 17) SOURCE: U.S. ENERGY INFORMATION ADMINISTRATION ANNUAL ENERGY OUTLOOK 2019 — TRANSPORTATION SECTOR KEY 33.9% 66.1% INDICATORS CY2008 CY2007 34.2% 65.8% Light-Duty Vehicles<=8,500 Ibs Commercial Light Trucks Freight Trucks >10,000 Ibs. CY2006 34.4% 65.6%

4000 CY2005 34.7% 65.3%

Growth (2018-2050) = 0.6% CY2004 37.7% 62.3% 3500

3000 Percent Change in Miles Driven per Road Type Versus Prior Year CY2004-CY2019 (FIGURE 19) | SOURCE: HTTPS://WWW.FHWA.DOT.GOV/POLICYINFORMATION/TRAVEL_MONITORING/TVT.CFM

2500 CY2004 CY2005 CY2006 CY2007 CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 2000 0.1

1500 0.08

1000 0.06 Growth (2018-2050) = 1.3% 500 Growth (2018-2050) = 1.4% 0.04 0 0.02 CY2017 CY2021 CY2031 CY2018 CY2019 CY2041 CY2037 CY2027 CY2022 CY2032 CY2023 CY2033 CY2035 CY2025 CY2028 CY2038 CY2047 CY2024 CY2026 CY2039 CY2029 CY2036 CY2034 CY2042 CY2043 CY2045 CY2048 CY2020 CY2030 CY2049 CY2046 CY2050 CY2044 CY2040 0

Growing Congestion is Costly -0.02

Congestion is a major problem in the U.S., especially in urban areas, where share of overall miles -0.04 driven continues to grow (see Figures 18 and 19). Nearly 90 percent of individuals aged 16 years and older in the U.S. use a private passenger vehicle to get to work. Today the average private car in -0.06 the U.S. carries only 1.6 people, and 95 percent of the time the car is parked and not being driven at all.52 When it is driven, increased congestion often means driving is a frustrating experience — in CY -0.08 Rural Interstate Rural Other Other Rural Urban Interstate Urban Other Other Urban All Systems 53 2018 nearly every major U.S. city recorded a downtown last-mile travel speed below 20 mph. Arterial Arterial

A growing number of hours and dollars for fuel are wasted, with the average commuter in CY 2017 U.S. Primary Means of Transportation to Work by Selected Characteristics (FIGURE 21) spending an extra 54 hours (more than a week of vacation) traveling, wasting over 21 gallons of fuel SOURCE: U.S. CENSUS BUREAU, 2017 AMERICAN COMMUNITY SURVEY 1-YEAR ESTIMATES — a week’s worth of fuel for the average driver (see Figure 20).54 The American Trucking Research CY2005 CY2006 CY2007 CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 Institute estimates congestion costs for the trucking industry alone are $74.5 billion annually from 1.2 billion lost hours of productivity, a number equivalent to 425,533 truck drivers sitting idle for an Car, truck, or van total 87.7% 87.2% 87.0% 86.7% 86.7% 86.8% 86.6% 86.5% 86.3% 86.2% 86.1% 85.9% 85.8% 55 entire year. Car, truck, or van 77.0% 76.5% 76.6% 76.1% 76.6% 77.1% 76.9% 76.8% 77.0% 77.0% 77.0% 76.8% 76.9%

- drove alone |

National Congestion Measures CY1982-CY2017 (FIGURE 20) Car, truck, or van 10.7% 10.7% 10.4% 10.7% 10.0% 9.7% 9.7% 9.7% 9.4% 9.3% 9.0% 9.0% 8.9% MONEY IS TIME SOURCE: TEXAS A&M TRANSPORTATION INSTITUTE AND ATRI. URBAN MOBILITY REPORT 2019. AUGUST 2019. P. 2 - carpooled

4.7% 4.8% 4.9% 5.0% 5.0% 4.9% 5.0% 5.0% 5.1% 5.2% 5.2% 5.1% 5.0% Delay per Commuter (Hours per Yr) Total Cost (Billions of 2017 Dollars) Public transportation . (excluding taxicab) 60 $180 Walked 2.5% 2.5% 2.5% 2.5% 2.5% 2.4% 2.4% 2.5% 2.4% 2.4% 2.4% 2.3% 2.3% $160 50 $140 Taxicab, motorcycle, 1.6% 1.7% 1.7% 1.8% 1.7% 1.7% 1.7% 1.8% 1.9% 1.8% 1.8% 1.8% 1.8% bicycle, or other means 40 $120 $100 Worked at home 3.6% 3.8% 3.9% 3.9% 4.1% 4.2% 4.2% 4.2% 4.2% 4.3% 4.5% 5.0% 5.1% 30 $80 20 $60 $40 10 $20 Total Cost (Billions of 2017 Dollars) (Billions of 2017 Cost Total Delay per Communter (Hours per Yr) (Hours per Communter Delay 0 $- Workers by Commute Time and Vehicles Available, CY2005, CY2009, CY2013, CY2017 (FIGURE 22) SOURCE: AMERICAN COMMUNITY SURVEY TABLE S0802 (HTTPS://FACTFINDER.CENSUS.GOV/FACES/TABLESERVICES/JSF/ CY1991 CY2011 CY1987 CY1997 CY1982 CY1983 CY1992 CY1985 CY1993 CY2017 CY1995 CY2012 CY1988 CY2013 CY2015 CY1989 CY1986 CY1998 CY1984 CY1996 CY1999 CY2016 CY1994 CY2014

CY1990 PAGES/PRODUCTVIEW.XHTML?PID=ACS_17_1YR_S0802&PRODTYPE=TABLE) CY2001 CY2010 CY2007 CY2002 CY2003 CY2005 CY2008 CY2006 CY2009 CY2004 CY2000 Travel Time to Work CY2005 CY2009 CY2013 CY2017

According to data from the U.S. Census Bureau’s American Community Survey, 86 percent of people in the U.S. use a personal vehicle to commute to work each day (see Figure 21), a number Less than 10 minutes 14.7% 14.0% 13.1% 12.2% that has fluctuated only moderately since CY 2005. The time many spend commuting to work has 10 to 14 minutes 14.3% 14.6% 14.0% 13.3% also grown (see Figures 22 and 23), and among the fifty largest cities in the U.S., fewer than 15 to 19 minutes 15.5% 15.7% 15.5% 15.1% 12 percent of jobs are reachable within 30 minutes (see Figure 24).56 Compare to data showing 20 to 24 minutes 14.5% 14.7% 14.8% 14.4% where population growth is highest, and you can see the congestion is likely to get worse (see 25 to 29 minutes 6.1% 6.1% 6.3% 6.5% Figure 13). The United Nations Department of Economic and Social Affairs estimates that by 30 to 34 minutes 13.2% 13.5% 13.6% 13.8% CY 2050, two of every three people are likely to be living in cities and other urban centers, an 35 to 44 minutes 6.4% 6.3% 6.6% 7.0% additional 2.5 billion people.57 45 to 59 minutes 7.5% 7.4% 7.8% 8.3% 60 or more minutes 7.9% 7.8% 8.4% 9.3% Factor in U.S. Department of Transportation estimates that 58 percent of U.S. roads are in less than good condition and 23 percent of bridges need significant repair or can’t handle today’s traffic,58 Mean travel time 25.1 25.1 25.8 26.9 and the cost of driving gets even worse. to work (minutes)

According to research from the Insurance Information Institute, PCIAA and CAS, collision frequency Vehicles Available CY2005 CY2009 CY2013 CY2017 in the United States is strongly and positively related to various measures of congestion. To distinguish among the variables and find the ones which best predict collision frequency, a random No Vehicle Available 4.1% 4.3% 4.5% 4.2% forest was constructed to compare the importance of each variable to the model. Variables with 1 vehicle available 20.9% 21.4% 21.6% 20.3% more importance have the best predictive ability. As Figures 25 and 26 illustrate, five variables 2 vehicles available 43.2% 42.1% 41.6% 40.5% stand out: Drivers per Lane Mile (Licensed drivers/lane miles total), Urban Average Commute Time, 3 or more vehicles available 31.8% 32.2% 32.3% 34.9% Rural Average Commute Time, Tort System, and Urban VMT.59

Average Travel Time to Work in the U.S. by Metro Area, CY2017 (FIGURE 23) Frequency vs. Driver per Lane Mile (Q4 CY2015) (FIGURE 25) SOURCE: AMERICAN COMMUNITY SURVEY TABLE S0802 (HTTPS://FACTFINDER.CENSUS.GOV/FACES/TABLESERVICES/JSF/PAGES/ SOURCE: AUTO LOSS COST TREND REPORT, JANUARY 2018. | NOTE: RANKED BY INCNODEPURITY (CHANGE IN PREDICTIONS WHEN A VARIABLE PRODUCTVIEW.XHTML?PID=ACS_17_1YR_S0802&PRODTYPE=TABLE) IS RANDOMLY PERMUTED THROUGH RANDOM DECISION FOREST). *ADJUSTED FOR MILES DRIVEN.

Vehicle per Lane Mile

140

R2 = 0.7509 | TIME IS MONEY IS TIME 120

100

Drivers per Lane Mile Drivers 40

Travel Time to Work 20 32.0 minutes or greater 28.0 to 31.9 minutes 24.0 to 27.9 minutes - 20.0 to 23.9 minutes 0.00 0.05 0.10 0.15 0.20 Less than 20.0 minutes

U.S. Average is 27.1 minutes On-Level Frequency*

Variable Importance to Collision Frequency (FIGURE 26) SOURCE: AUTO LOSS COST TREND REPORT, JANUARY 2018. | NOTE: RANKED BY INCNODEPURITY (CHANGE IN PREDICTIONS WHEN A VARIABLE IS RANDOMLY PERMUTED THROUGH RANDOM DECISION FOREST). *ADJUSTED FOR MILES DRIVEN. Percent of Jobs Reachable by Number of Minutes, CY2016 (FIGURE 24) SOURCE: ANDREW OWEN, BRENDAN MURPHY, ACCESSABILITY OBSERVATORY, CENTER FOR TRANSPORTATION STUDIES, UNIVERSITY OF MINNESOTA. DAVID LEVINSON, SCHOOL OF CIVIL ENGINEERING, UNIVERSITY OF SYDNEY. ACCESS ACROSS AMERICA: TRANSIT 2016. CTS 17-07. Drivers/Lane Mile

Urban Avg Commute 12.0% Rural Avg Commute 10.0% Tort System

8.0% % Urban VMT

Rainy Area 6.0% % Rural VMT

4.0% Lane Miles

Lawyers per Million 2.0% Avg Miles/Driver

0.0% Rural VMT/Lane

DUI’s Miami Dallas Austin Tampa Detroit Seattle Boston Denver Bualo Atlanta Raleigh Phoenix Orlando Chicago Houston Hartford St. Louis St. Portland San Jose Nashville Riverside Louisville Charlotte New York New Baltimore Cleveland Cincinnati Columbus Las Vegas Richmond San Diego Pittsburgh Milwaukee Urban VMT/Lane Providence Kansas City Kansas Minneapolis Sacramento Washington Jacksonville Indianapolis Los Angeles Los Philadelphia Birmingham San Antonio New Orleans New Salt Lake City Salt Lake San Francisco Virginia Beach Virginia Oklahoma City PIP

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 29 CRASH COURSE 2020 COURSE CRASH Is Ride-Hailing Adding to Congestion? impact of ride-hailing on urban mobility found 70 percent of Uber and Lyft trips were made in nine large, densely-populated metros: New York City, Boston, Chicago, Los Angeles, Miami, Philadelphia, San Francisco, Seattle, and Washington D.C.62 The data also found private ride TNC services put Ride-hailing’s share of overall miles driven in the U.S. remains small but has seen significant growth. 2.8 new TNC miles on the road for each individual mile of personal driving removed, resulting According to data from the National Household Travel Survey, for-hire vehicle use doubled between in a 180 percent increase in driving on city streets.63 Several other reports released in CY 2019 60 CY 2009 and CY 2017, a trend believed to be primarily driven by the advent of ride-hailing. The concluded many consumers are using ride-hailing trips instead of walking or public transportation CY 2018 study from Conway, Salon, & King reveals 0.19 percent (±0.04) of person-trips utilized

at peak commute times. In April of CY 2019, the city of Chicago published anonymized trip data | for-hire vehicles in the CY 2009; by CY 2017, that number had increased to 0.50 percent (±0.10).61 from ride-hailing companies in operation in Chicago. This data showed the most common ride- MONEY IS TIME NHTS shows total household vehicle trips (or vehicle miles traveled “VMT”) accounted for 60 hailing trip made during the two-month period of November and December CY 2018 was a 1.1 mile percent of all person miles traveled (PMT) in CY 2009 but fell to 53 percent by CY 2017 (see trip from River North West to the Loop, with 80 percent of them occurring between 7:00 am and Figure 27). Applying for-hire vehicle use person-trip growth to PMT share of VMT in 2009 and 7:59 pm, and 84 percent occurring during weekdays.64 In Seattle, nearly half of all ride-hailing trips 2017 suggests ride-hailing grew from 0.1 percent of all vehicle miles traveled in CY 2009 to 0.3 originated in four neighborhoods that are some of the city’s most walkable and transit-friendly percent by CY 2017. But while the share of overall miles traveled remains small, the concentration of areas.65 And, a study from the Booth School at the University of Chicago found cities with high ride-hailing geographically has resulted in outsized impact elsewhere. adoption of Uber and Lyft had 3 percent more total miles driven daily on average than cities with low adoption, and had higher new vehicle registrations.66 U.S. National Household Travel Survey Person Trips and Household Vehicle Trips CY1969-CY2017 (FIGURE 27) A common theme among most of these studies is that where these services are available, SOURCE: 2017 NATIONAL HOUSEHOLD TRAVEL SURVEY, P.8 AND P.13, TABLES 1A AND 3B. HTTPS://NHTS.ORNL.GOV/ASSETS/2017_NHTS consumers are using them, and using them a lot. They are willing to pay to do so, even when SUMMARY_TRAVEL_TRENDS.PDF. walking or public transport would be a cheaper alternative. In fact, a working paper from the National Bureau of Economic Research found a consumer surplus of about $6.8 billion for all CY Daily Person Average Person Household Vehicle Person Miles of Travel NHTS Survey 2015 in the U.S. for Uber. This was based on data from nearly 50 million UberX consumer sessions Year Trips per Person Trip Length (miles) Trips (VMT in millions) (PMT in millions) from the first 24 weeks in CY 2015 from Uber’s four biggest markets that showed Uber customers may have been willing to pay more than they actually did (consumer surplus is defined as the 2.02 9.67 775,940 1,404,137 1969 difference between the amount consumers are willing and able to pay for a service versus what 1977 2.92 8.87 907,603 1,879,215 they actually pay).67 1983 2.89 8.68 1,002,139 1,946,662 TNCs may also lead to a potential decline in public transit. A CY 2019 study by the University of 1990 (adj) 3.76 9.47 1,695,290 2,829,936 Kentucky’s Department of Civil Engineering of 22 large U.S. cities revealed that for each year after 1995 4.3 9.13 2,068,368 3,411,122 TNCs enter a market, heavy rail ridership can be expected to decrease by 1.3 percent and bus 68 2001 4.09 10.04 2,274,769 3,783,979 ridership can be expected to decrease by 1.7 percent. Finally, a study from three economists at the University of Louisville and George State University found an increase in self-reported alcohol 2009 3.79 9.75 2,245,111 3,732,791 consumption compared with the year before Uber launches in a city. This study compared data 2017 3.37 10.7 2,105,882 3,970,287 on Uber availability with health surveys from America’s Centers for Disease Control and found on 2017 (adj) n/a 11.57 2,321,820 4,291,150 average alcohol consumption rose by 3 percent, binge drinking increased by 8 percent, and heavy drinking surged by 9 percent within a couple of years of a ride-hailing company coming to town.69 The bright side, of course, is at least these people are not driving drunk.

Not surprisingly as lawmakers try to adapt to the changes being brought by ride-hailing services, many are looking to impose new fees and taxes, and re-classify drivers as employees rather As usage of ride-hailing or transportation network companies (TNCs) — such as Uber and Lyft — than independent contractors. A study by New York University’s Rudin Center for Transportation have grown, there is growing debate about whether they add to problems of growing congestion in Policy recently looked at 13 large major markets including London, San Francisco, New York City many cities. Numerous studies published in the last several years look at how TNCs have impacted and others to understand how cities are rewriting the rules for ride-hailing and found a range of consumer travel behavior. Schaller Consulting published two studies — its second report on the responses, including new regulation around provision of trip data, congestion taxes, strict emission standards and restrictions in congestion zones, and driver pay.70

It is unknown whether cars will remain the transportation mode of choice. As more alternatives Many scooter rental operators struggle to make money due to significant expenses such as come online and become viable, fewer consumers will see vehicle ownership as critical. In their 2018 depreciation of the e-scooters, the cost to run warehouses that repair and help to position the Evolution of Mobility Study, Cox Automotive found the percent of individuals that felt owning a scooters, and development costs related to the development more durable hardware.72 Lime, the vehicle was necessary fell with each generation (see Figure 28). world’s biggest scooter operator had 120,000 scooters on the road globally during the summer of CY 2019, but scooters put on the road in the preceding spring lasted only a median of five months.73 Vandalism is also a problem, thanks perhaps to Instagram sites like Bird Graveyard

Cox Automotive 2018 Evolution of Mobility Study: “Having Transportation is

which celebrate e-scooter stunts and destruction; the growing number of injuries is another major | Necessary, But Owning a Vehicle is Not (% Agree)” (FIGURE 28)

SOURCE: HTTPS://WWW.COXAUTOINC.COM, FEBURARY 5, 2019. concern. MONEY IS TIME

Gen Z 55% There are numerous other challenges for e-scooter companies, including regulation and permitting, parking, age requirements, safety equipment requirements, maintenance, and liability when someone is injured.74 Late in CY 2019, the National Conference of Insurance Legislators Millennials 45% (NCOIL) met to review a draft of legislation setting insurance requirements for electric scooters. The legislation is modeled after the group’s model for transportation network companies (TNCs) to ensure there is liability insurance coverage in place during any rental period for the scooter Gen X 34% company, the rider, and any independent contractors servicing the scooters.75

Boomers 28% Trip Data Grows Ever More Valuable

One of the greatest potential benefits of ride-hailing and micro-mobility solutions is the ability Emergence of Micro-Mobility to capture trip-level data and use it to understand traffic patterns and infrastructure needs. Uber recently introduced dockless bike and scooter services, added third party services to its app As cities struggle to manage congestion related to cars, many are forging ahead with the expansion including Lime, and launched new capabilities with the City of Denver to add transit directions and of dedicated bicycle lanes and the introduction of micro-mobility solutions, including shared bikes, eventually support the ability to buy transit passes through the Uber app. While Uber gets a cut e-bikes, and e-scooters. In CY 2018, a total of 84 million trips were taken in the U.S. (45.5 million on of each ride taken through its app, it also expands the amount of data it now collects on where shared bikes and 38.5 million on e-scooters), double the number taken in CY 2017.71 These types of people are going and how they’re getting there.76 solutions are believed to be best suited for a large proportion of trips taken in most metro areas, where nearly half of the trips are less than three miles, based on data collected and analyzed by Eventually, highways will have embedded sensors that monitor wear-and-tear and traffic patterns. Inrix of more than 50 million car trips taken during the month of October CY 2018 in the top 25 The state of Colorado is testing this smart road technology. Over the next five years, a half-mile most congested U.S. metro areas (see Figure 29). stretch of U.S. Highway 285, southwest of Denver, will be covered with concrete slabs that include Wi-Fi connectivity and embedded fiberoptic cables acting as pressure sensors. The technology Inrix U.S. Cities by Micromobility Potential — Percent of All Trips by Car Trip Distance will enable connection to motorists’ cellphones to provide real-time alerts about traffic and (FIGURE 29) | SOURCE: INRIX 2018 GLOBAL TRAFFIC SCORECARD, “MICROMOBILITY POTENTIAL IN THE U.S., UK AND GERMANY” SEPTEMBER 2019. road hazards.77 Another electric road system project began in November 2019, with the island of Gotland in Sweden installing a proof of concept wireless charging road for trucks/buses in an 0-1 Mile 1-2 Miles 2-3 Miles Other effort to decarbonize the transport sector.78 100% 90% 80% As more metro areas struggle with growing congestion, understanding where, when, why and how 52% 44% 49% 49% 49% 49% 50% 49% 49% 50% 51% 51% 52% 52% 52% 54% 54% 54% 54% 54% 56% 55% 56% 57% 57% 58% 70% people are traveling is critical. It is no surprise that automakers are also looking to capitalize on 60% data that is now available with the connected vehicle. 50% 40% 30% Access to the data sets generated by the numerous electronic control modules throughout the 20% 10% vehicle helps automakers understand how their vehicles are performing. It also gives them new 0% access to how, when and where their customers are driving — creating an opportunity to develop new products and services.79 March CY 2019 data from Statista indicates overall revenue from the Seattle MiamiTampa Dallas Austin Chicago Boston Denver Atlanta OrlandoPhoenix Honolulu Nashville Charlotte Portland Houston New York Pittsburgh Baltimore connected car segment is expected to grow, but average revenue per connected car has fallen New Orleans Los Angeles PhiladelphiaWashington Minneapolis San Francisco 80 U.S. City Averages (see Figure 30).

Connected Car U.S. Average Revenue per Connected Car, and Projected Segment …the B2C elements [of connected car technology] Growth (FIGURE 30) | SOURCE: STATISTA, MARCH 2019. (HTTPS://WWW.STATISTA.COM/OUTLOOK/320/109/CONNECTED-CAR/UNITED- STATES?CURRENCY=USD#MARKET-ARPU) are a Trojan horse for the more powerful B2B

Connected Car Connected Car Connected Car proposition of service scheduling, insurance claims NHTS Survey Average Revenue Per Segment Total segment - Vehicle segment - Infotainment Year Connected Car revenue growth Services revenue growth Services revenue growth management, and related fleet management

“ opportunities including cybersecurity and over-the- | 2018 $163.00 58.0% 35.3% 1390.0% air software update propositions." MONEY IS TIME 2019 $140.07 39.6% 26.1% 7.4% 2020 $122.69 29.8% 20.0% 5.3% SOURCE: ROGER LANCTOT, STRATEGY ANALYTICS, JULY 8, 2019. HTTPS://WWW.STRATEGYANALYTICS. COM/STRATEGY-ANALYTICS/BLOGS/AUTOMOTIVE/CONNECTED-MOBILITY/AUTOMOTIVE-CONNECTED- 2021 $109.67 23.9% 16.0% 4.5% MOBILITY/2019/07/08/ONSTAR-ROLLS-BACK-FREE-TRIALS 2022 $99.93 19.9% 13.3% 4.5% 2023 $90.79 16.2% 9.2% 2.6% Trip Data Also Tells Us How People Are Driving and Why Crashes Occur

Trip data is an important source of information on how a vehicle is driven. On-board telematics Connectivity, when billed as a subscription service by the automaker, continues to see only limited devices capture information — such as speed, hard braking, cornering — and other behaviors that uptake by consumers.81 Because most new vehicles have limited maintenance issues during the numerous studies show correlate to higher accident rates. For example, a study from Canada’s first several years of ownership, the perceived value of a connected car subscription that provides University of Waterloo compared four bad driving behaviors — speeding, hard braking, hard vehicle system monitoring is much less.82 The connected car data is more valuable to automakers acceleration, and hard cornering, to the likelihood of crashes. The study concluded speeding is the and their dealers: Careful monitoring of vehicle diagnostic data in aggregate can help avoid the riskiest kind of aggressive driving, and a strong predictor of crashes.87 potential of hundreds of millions of dollars in warranty exposure.83 Connected car technology also helps automakers deliver less expensive and disruptive defect remedies.84 In CY 2018, nearly 8 Unfortunately, a CY 2018 national survey of drivers conducted by AAA Foundation for Traffic million vehicles in the U.S. were affected by software-based defects — a larger number than the Safety found that 49 percent of motorists reported exceeding the speed limit by 15 mph on a previous five years combined.85 Dealers can use that data to bring in service and preventive freeway and 40.1 percent reported driving 10 mph over the speed limit on a residential street maintenance work. (see Figure 31). Additionally, data from the Governors Highway Safety Association disproves the argument that bumping speed limits higher simply reflect how motorists actually drive. It is thought that the value of the data alone will encourage automakers to explore alternative Research has shown raising speed limits to match the 85th percentile speed increases the average business models. Audi’s new eTron launch strategy offers an a la carte menu of options where operating speed of the roadway, which then increases the 85th percentile speed.88 For example, drivers can buy connected car features on demand.86 Other automakers offer certain types when Texas raised its speed limits to 70 mph on urban freeways in the 1990s, the share of drivers of service — such as roadside assistance and accident detection — at no additional fee. These going faster than 70 mph jumped from 15 percent to 50 percent, while the share driving faster types of services are among the most valued connected car services. For example, in the case than 75 mph went up too, from 4 percent to 17 percent. When California raised its urban freeway of an accident, a connected vehicle can share crash data and initiate first notice of loss with the limits from 55 to 65 mph, drivers traveling faster than 70 mph increased from 29 percent to 41 automaker through its connected vehicle program. The OE can be the first to engage with that percent.89 Conversely, in states and metro areas such as Washington and Boston, where residential customer, provide emergency services, and provide information on their certified repair network. or business district speed limits were lowered, IIHS found lowering speeds do act as an effective They also can share crash data with the insurer and repairer, helping drivers move seamlessly into countermeasure to reduce speeds and improve road safety.90 the entire claims process.

The National Motor Vehicle Crash Causation Survey (NMVCCS) conducted from CY 2005 to CY 2007 looked at pre-crash movement, critical pre-crash event, critical reason, and associated factors. The critical reason, or last event in the crash causal chain, was assigned to the driver in 94 percent of the crashes. It’s important to note the critical reason is not intended to be interpreted

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 35 CRASH COURSE 2020 COURSE CRASH as the cause of the crash nor as assignment of fault, instead it is often the last failure in the causal Critical Reasons for Crashes Investigated in the National Motor Vehicle Crash chain of events leading to the crash.91 Also worth noting is the additional detail on critical reason Causation Survey, U.S., CY2005-CY2017 (FIGURE 32) assigned to drivers shows recognition error (includes driver’s inattention, internal and external SOURCE: NATIONAL SAFETY COUNCIL. INJURY FACTS® 2017 EDITION, P.119. CITED SOURCE NHTSA (DOT HS 812 115) distractions and inadequate surveillance) and decision error (includes driving too fast for conditions 2% 2% or the curve, illegal maneuver and misjudgment of gap or others’ speed) account for nearly three 2%

quarters of the accidents researched (see Figure 32).

7% | TIME IS MONEY IS TIME 10%

Survey — “Proportion of Drivers who reported distracted/risky driving behaviors at 39% least once in last 30 days“ (FIGURE 31) SOURCE: AAA FOUNDATION FOR TRAFFIC SAFETY’S 2018 TRAFFIC SAFETY CULTURE INDEX

Held and talked on cell phone 31% 52.0% 94%

Driving 15 mph over the speed limit 49.0% on freeway

Read a text/email on a cell phone 41.3% Driver Recognition error

Vehicles Decision error Driving 10 mph over the speed limit on a 40.1% residential street neighborhood Environment Performance error

Typed or sent a text message/ email 32.1% Unknown Critical Reasons Non-performance error on a cell phone In a CY 2014 task force report, the Casualty Actuarial Society (CAS) deconstructed the NMVCCS Speeding through a red light 31.4% and found only 32.5 percent of accidents were caused by human behavior, while 21.3 percent could be attributed to technology issues (figures do not add up to 100 percent to reduce duplication of Driving while being so tired that they had 27.1% 92 a hard time keeping their eyes open accidents with multiple causation). Despite the different conclusions drawn by NHTSA and the CAS, what both analyses highlight are the dangers of speeding, but also driver distraction. In the Aggressive driving 24.8% U.S., an estimated 9 to 10 percent of all fatal crashes between CY 2013 and CY 2017 were from distraction-affected crashes.93

Driving without wearing a seatbelt 16.7% Changes to vehicles themselves hasn’t necessarily helped either. According to data from IHS Riding in a car driven by someone Automotive, the number of vehicles built in North American with display screens measuring 7 13.1% 94 who has had too much alcohol inches or more has grown by 75 percent in the past five years from 6.3 million to 10.9 million. During that same time period the average screen size has also increased from 6.4 inches to 7.3 95 Driving after drinking enough alcohol 10.9% inches. AAA has found that wider use of touch screens that often replace more familiar knobs and that they may be over the legal limit buttons can be distracting.96

Driving shortly (within an hour) 6.6% after using marijuana Among the nearly 300 million vehicles in operation in the U.S., it is estimated roughly 80 percent Driving after using potentially of them are at Level 0 automation, where the driver is supposed to be in control of the vehicle impairing prescription drugs 5.6% 100 percent of the time. For drivers who tend to engage in dangerous and distracting behaviors, telematics also offers the ability to track and provide feedback (through scorecards or even 0% 10% 20% 30% 40% 50% 60% discounted insurance premiums) on driving behavior, which may prevent bad behavior in the first place. According to CY 2019 data from PTOLEMUS, among insurers selling usage-based

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 37 CRASH COURSE 2020 COURSE CRASH insurance, 70 percent track harsh braking, 65 percent track speeding, and 61 percent track rapid What is Monitored with Telematics Among Commercial Fleets Today (FIGURE 33) acceleration.97 Among the insurance company programs they analyzed that were smartphone- SOURCE: TELETRAC NAVAMAN. 2019 TELEMATICS BENCHMARK REPORT — U.S. EDITION. centric they found 80 percent use rapid acceleration and harsh braking; 57 percent use cornering; Vehicle/equipment tracking 74% 50 percent consider night time driving or peak time driving; and 27 percent use contextual data Hours of service/driver hours 66% 98 sets such as road type and traffic data. Speed 61%

Distance Driven 52%

Early data from telematics providers revealed one of the major benefits of telematics was Driver Performance 42% | supported by the self-selection theory that those drivers that opted in were less accident prone Idling 37% MONEY IS TIME due to being lower mileage drivers. More recent data also suggests that telematics can help reduce Harsh Braking 33% distracted driving, which, when accidents do occur, do so at lower speeds, leading to a lower cost Maintenance 32% of claim.99 Fuel Usage 30% Engine Hours 29% Ensuring the driver is aware of the risk he/she is taking behind the wheel has been shown to lead Proof of Service/Jobs Completed 26% Lone Workers 4% them to practice safer habits. This is particularly true within the commercial insurance sector. Other 3%

0% 10% 20% 30% 40% 50% 60% 70% 80% Telematics’ Opportunity to Help Commercial Challenges Abound for Commercial Auto Insurance Auto Insurance

The adoption of telematics among commercial fleets is advancing even faster than among Commercial Auto Insurance continues to experience difficult results. The combination of these personally owned vehicles. Commercial auto is best poised to use data such as driver behavior, factors drives commercial auto frequency and severity trends: the growth of e-commerce driving driver routes and route densities, weather, and other connected vehicle data to improve the demand for more vehicles, miles driven; difficulty finding experienced drivers; higher tort costs; accuracy of risk assessment and product pricing. With the significant growth in e-commerce, fleet deteriorating roads; increasing driver distraction; increases in medical inflation; increased rate traffic and demand for commercial auto insurance has increased, and the ‘last mile’ challenge has of high-speed survival with significant rehabilitation costs; and higher loss costs due to vehicle led to greater diversity among commercial fleets in terms of who is actually driving, when and technology.103 Since CY 2011, the direct combined ratio for commercial auto has exceeded where they are driving, and what they are driving. 100 percent (see Figure 34), although the industry did see some modest improvement in its profitability with the incurred loss ratio improving to 69.1 percent in CY 2018 from 70.2 percent in Telematics adoption within the commercial fleet has grown dramatically, with an estimated 20 CY 2017. million units expected in the next three years, up from 13 million today. A survey conducted by

Teletrac Navman in its CY 2019 U.S. edition of Telematics Benchmark Report revealed vehicle Commercial Auto Combined Ratio (before reinsurance) (FIGURE 34) tracking and driver hours are monitored most today with telematics (see Figure 33).100 In fact, SOURCE: ASSURED RESEARCH, LLC much of the growth in telematics adoption was driven by a federal mandate for incorporation 120% of Electronic Logging Devices (ELDs) and Automatic On-Board Recording Devices (AOBRDs). Future growth is anticipated to come from products such as asset monitoring of trailers, shipping 100% containers, and driver behavior management systems.101 More commercial fleets are adopting 80% telematics solutions with video and artificial intelligence integrated, which help to provide additional context, promote safer driving behavior among drivers, and potentially lower accident 60% frequency/severity and lower insurance premiums. 40%

It’s estimated that the annual accident rate for commercial fleets is around 20 percent, and the 20% average loss related to fleet vehicle accidents is approximately $70,000.102 It’s no wonder with 0% increased fleet miles in the U.S. that commercial auto has seen higher frequency and loss cost trends over the last several years. CY2000CY2001CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018CY2019E

In response, carriers raised prices, with price increases near or above double digits for 10 Fast Track Paid Claim Frequency by Calendar Year, Rolling 4 Quarters ending Q4 consecutive quarters as of Q4 CY 2019. Data from the American Transportation Research Institute Each Year (FIGURE 35) shows the average annual cost per mile for truck insurance premiums rose 42 percent between SOURCE: ISS FAST TRACK PLUS™ PERSONAL AUTO, AS OF JUNE 30, 2019 | NOTE: CY2019 ROLLING 4 QUARTERS ENDING Q2 2019 CY2010 and CY2018, from $0.06 per mile to $0.08.104 In the Commercial Property & Casualty PD Liability Collision PD Liability CY2007 Collision CY2007 Market Outlook Q4 2019-2020, USI predicts rates for primary auto liability with fleet lower than 10.0%

200 and good loss history will increase 10 to 15 percent, while those same fleets with poor loss history will see rates increased between 15 and 25 percent. Fleets in excess of 200 could see rates 9.0% |

8.0% MONEY IS TIME up between 15 and 30 percent.105 And while technology such as ADAS features and telematics will 7.0% eventually help improve frequency, for now benefits are small. 6.0% 5.0% Many fleet service vehicles tend to use the lowest base trim level package on trucks, and often 4.0% don’t have the option to order ADAS features, or often find the cost-benefit hard to prove.106 There 3.0% is also concern that the ADAS features may create a false sense of security among drivers — for 2.0% example relying on backup cameras versus walking around the vehicle prior to backup up or 1.0% looking over their shoulders. For fleets that do purchase vehicles with the ADAS features, rising 0.0% repair costs is also a concern. CY1981 CY1991 CY2011 CY1987 CY1997 CY1982 CY1983 CY1992 CY1985 CY1993 CY2017 CY1995 CY2012 CY1988 CY2013 CY2015 CY1989 CY1986 CY1998 CY2018 CY1984 CY1996 CY1999 CY2019 CY2016 CY1994 CY2014 CY1980 CY1990 CY2001 CY2010 CY2007 CY2002 CY2003 CY2005 CY2008 CY2009 CY2006 CY2004 CY2000

Looking forward to CY 2020, there is some hope the industry will see better results. Despite rising freight volumes, the cost of freight (measured as door-to-door spot rate index) fell sharply in CY 2019, driving the portion of the trucking industry closely tied to the factory sector into a recession. In the first half of CY 2019, around 640 trucking companies went bankrupt, and U.S. jobs data shows 5,100 trucking payrolls were slashed in August and another 4,200 were cut in September.107 Demand for freight services had weakened in part from ongoing trade policy disputes. While this Percent Change in Claim Frequency & Severity from Quarter One Year Prior (each is not great news for the trucking industry as a whole, it could result in improved frequency and quarter represents rolling 4 quarters data ending that quarter) (FIGURES 36-38) loss costs for the commercial auto insurance sector, which continues to trail the private passenger SOURCE: ISS FAST TRACK PLUS™ PERSONAL AUTO, AS OF SEPTEMBER 30, 2019 auto sector. Data from FMCSA for CY 2019 through July suggests there was a decline in crashes involving large trucks and buses for the first time in five years (3.2 percent drop versus 6.5 percent FIGURE 36 increase in CY 2018).109 COLLISION Frequency Severity Reduced Accident Frequency, but Rising Costs 7.0% Likely for Personal Auto in 2020 6.0% 5.0%

Private passenger auto claim frequency has followed a downward trend since the early 1980’s. 4.0%

While there has been fluctuation as the economy improved or slid into recession, recent increases 3.0% in CY 2015 and CY 2016 are beginning to reverse; paid claim frequency for both collision and property damage liability is at or below the levels from CY 2007 (see Figure 35).110 According to 2.0% the ISS Fast Track Plus™ reports, private passenger auto collision frequency closed out CY 2018 1.0% down 1.5 percent versus CY 2017, and the four quarters ending Q3 CY 2019 show collision frequency 0.0% down -1.8 percent from one year prior (see Figure 36).111 Private passenger auto property damage liability frequency also declined 3.9 percent in CY 2018 from CY 2017, and the four quarters ending -1.0% Q3 CY 2019 show liability frequency down another -3.8 percent (see Figure 37).112 Devastating 2012-Q42013-Q12013-Q22013-Q32013-Q42014-Q12014-Q22014-Q32014-Q42015-Q12015-Q22015-Q32015-Q42016-Q12016-Q22016-Q32016-Q42017-Q12017-Q22017-Q32017-Q42018-Q12018-Q22018-Q32018-Q42019-Q12019-Q22019-Q32019-Q4 -2.0% wildfires, hurricanes, and thunderstorm events (includes tornadoes, hail, and straight-line winds) in numerous parts of the U.S. in CY 2019 resulted in comprehensive loss frequency up 3.0 percent for -3.0% the four quarters ending Q3 2019 versus a decline of -2.1 percent in CY2018 (see Figure 38).113

FIGURE 37 from CCC customers reveals repairable claim counts fell just over 1 percent in 2019, but when total PROPERTY DAMAGE/ LIABILITY loss claim counts were included, claim counts were essentially flat for the full year (see Figure 39). Frequency Severity 8.0% Overall Claim Counts — Percent Change from Prior Period CY2019 vs CY2018 (FIGURE 39) SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

6.0% Collision/Liability Only All Loss Categories |

2.00% MONEY IS TIME 4.0%

1.50% 2.0%

1.00% 0.0% 0.50% 2017-Q1 2013-Q1 2015-Q1

-2.0% 2018-Q1 2019-Q1 2016-Q1 2014-Q1 2017-Q2 2017-Q3 2013-Q2 2013-Q3 2015-Q2 2015-Q3 2017-Q4 2018-Q2 2018-Q3 2019-Q2 2016-Q2 2019-Q3 2016-Q3 2012-Q4 2014-Q2 2013-Q4 2014-Q3 2015-Q4 2018-Q4 2019-Q4 2016-Q4 2014-Q4 0.00%

-4.0% -0.50%

-6.0% -1.0%

FIGURE 38 COMPREHENSIVE -1.50% Frequency Severity 1st Qtr 2019 2nd Qtr 2019 3rd Qtr 2019 4th Qtr 2019 Full Yr 2019

25% Police-reported crashes tracked by NHTSA (crashes in excess of a pre-determined dollar amount) 20% provide a mixed picture of overall motor-vehicle accident counts in the U.S. (see Figure 40). Between CY 2011 and CY 2015, police-reported accident counts in the U.S. were increasing. 15% However, in 2016 NHTSA replaced the National Automotive Sampling System General Estimates Systems (NASS GES) with the modernized Crash Report Sampling System (CRSS),114 which made 10% year-over-year changes from CY 2015 more difficult to understand. For the updated 2016, 2017 and 2018 reporting, NHTSA implemented a change in how injury estimates are derived. Today, they 5% are based on people injured in fatal crashes from FARS, and estimated people injured in non-fatal injury crashes from CRSS or NASS GESS. Previously NHTSA estimated people injured solely on 0% information from CRSS or NASS GES. -5% 2017-Q1 2013-Q1 2015-Q1 2018-Q1 2016-Q1 2019-Q1 2014-Q1 2017-Q2 2017-Q3 2013-Q2 2013-Q3 2015-Q2 2015-Q3 2017-Q4 2018-Q2 2018-Q3 2016-Q2 2019-Q2 2016-Q3 2019-Q3 2012-Q4 2014-Q2 2013-Q4 2014-Q3 2015-Q4 2018-Q4 2016-Q4 2019-Q4 2014-Q4 A change was also made to how property-only-damage (PDO) crashes were reported in 2018. -10% Previously at 6 of the 60 CRSS sampled data collection sites, police officers could report the PDO costs as “unknown.”115 Updates to the State’s reporting criteria for these sites removed “unknown” -15% as an option, converting many PDO crashes with unknown property damage costs into reportable crashes in 2018. Previously these were considered non-reportable crashes. Subsequently the 4.4 -20% percent increase measured in crashes overall for NHTSA — while statistically significant per NHTSA What remains to be seen is whether overall growth in the vehicles in operation in the U.S. — from — is primarily related to those changes. Hopefully, as data from CY 2019 becomes available, a both new vehicle registrations and slower scrappage growth — will counter declining auto claim clearer picture will emerge from the NHTSA data on the trend in motor vehicle crashes in the U.S. frequency, and still result in growth in the actual number of vehicles in accidents. Data collected

Police Reported Motor Vehicle Crashes by Year (FIGURE 40) Private Passenger Auto Insurance Combined Ratio and Annual Percent Change in Net SOURCE: NHTSA TRAFFIC SAFETY FACTS 2016 DATA AND NTHSA POLICE-REPORTED MOTOR VEHICLE TRAFFIC CRASHES IN 2018 DOT HS 812- (FIGURE 41) SOURCE: III 2020 INSURACE FACT BOOK 860, NOVEMBER 2019 Written Premiums CY2009-CY2018

Property Liability Combined Ratio Collision/Comprehensive Combined Ratio % Chg % Chg Damage % Chg % Chg Liability % Chg Net Premiums Writen Coll/Comp % Chg Net Premiums Written

Fatal Prior Yr Injury Prior Yr Only Prior Yr Total Prior Yr 115 10.0% | TIME IS MONEY IS TIME 110 8.0% CY2005 39,252 1,816,000 4,304,000 6,159,000 -1.5% -3.9% -2.7% -3.0% CY2006 38,648 1,746,000 4,189,000 5,973,000 105 6.0% -3.1% -2.0% 2.1% 0.9% CY2007 37,435 1,711,000 4,275,000 6,024,000 -8.7% -4.7% -3.0% -3.5% CY2008 34,172 1,630,000 4,146,000 5,811,000 100 4.0% CY2009 30,862 -9.7% 1,517,000 -6.9% 3,957,000 -4.6% 5,505,000 -5.3% -1.8% 1.6% -2.8% -1.6% 95 2.0% CY2010 30,296 1,542,000 3,847,000 5,419,000 Ratio Combined CY2011 29,867 -1.4% 1,530,000 -0.8% 3,778,000 -1.8% 5,338,000 -1.5% 90 0.0% CY2012 31,006 3.8% 1,634,000 6.8% 3,950,000 4.6% 5,615,000 5.2% 85 -2.0% CY2013 30,202 -2.6% 1,591,000 -2.6% 4,066,000 2.9% 5,687,000 1.3%

-0.5% 3.6% 7.9% 6.6% Prior Year versus Growth Premium Written Net CY2014 30,056 1,648,000 4,387,000 6,064,000 80 -4.0% CY2015 32,539 8.3% 1,715,000 4.1% 4,548,000 3.7% 6,296,000 3.8% CY2011 n/a n/a n/a n/a CY2011 CY2017 CY2017 CY2012 34,439 2,177,000 5,065,000 7,277,000 CY2012 CY2013 CY2013 CY2015 CY2015 CY2018 CY2016 CY2016 CY2014 CY2016* CY2014 CY2010 CY2010 CY2009 CY2009 CY2017* 34,560 0.4% 1,889,000 -13.2% 4,530,000 -10.6% 6,453,000 -11.3% CY2018* 33,654 -2.6% 1,894,000 0.3% 4,807,000 6.1% 6,734,000 4.4% How Does ADAS Impact Accident Frequency?

A review of new vehicle launches in the U.S. underscores the growing prevalence of advanced driver assistance systems. Advanced driver assistance systems (ADAS) include technologies such as front crash prevention or warning, lane departure warning (LDW), blind spot detection, park assist, back-over prevention, obstacle detection, and automatic emergency braking (AEB). While For the private passenger auto sector, however, average claim costs across all three lines of there is no requirement or mandate from NHTSA that requires automakers to add these features to coverage continue to climb, though not as sharply as the increases in CY 2016. Personal auto their vehicles, in March 2016, NHTSA and IIHS announced that 20 major automakers committed to insurance profitability suffered in CY 2015 and CY 2016 as increases in both frequency and loss equip all vehicles of curb weight less than 10,000 pounds with automatic emergency braking (AEB) costs exceeded premiums collected. Several years of increases in personal auto premium (7.3 and frontal crash warning (FCW). These automakers comprise 99 percent of all vehicles sold in the percent in CY 2016, 8 percent in CY 2017, and 6.5 percent in CY 2018) combined with some U.S. annually. Specifically, the automakers committed to make AEB standard on virtually all light- slowdown in frequency helped combined ratios fall below 100 again, after several years where it duty cars and trucks with a gross vehicle weight of 8,500 pounds or less no later than Sept. 1, 2022, was much higher (see Figure 41).116 Personal auto’s net incurred loss ratio also improved in CY and on virtually all trucks with a gross vehicle weight between 8,501 pounds and 10,000 pounds 2018, falling to 64.5 percent from 68.8 percent in 2017 and 72 percent in CY 2016.117 The industry no later than Sept. 1, 2025.118 This commitment, as well as consumer demand, is driving more of this can anticipate elevated claim costs to remain the norm in the near term. New vehicle technologies, technology into the overall vehicle fleet today. consumer preference for light trucks and higher trim levels are driving up vehicle loss costs, and technologies that are supposed to help prevent accidents, while effective, are not yet pervasive enough to result in significantly fewer accidents. Many vehicles on the road in the U.S. are still at Level 0, with no automated driving functions. But growth among vehicles with Level 1-2 automation is accelerating. The percentage of vehicles (curb weight 8500 pounds or less) with AEB standard across these 20 automakers grew from just under 30 percent for MY2017 (vehicles manufactured after 1Sep16 and before 31Aug17) to over 50 percent for MY2018 and to an estimated 64 percent for MY2019 based on those that reported actual volumes of AEB-equipped vehicles (see Figure 42).

Percent of Vehicles Manufactured For and Sold In the U.S. Conforming Potential Decline in Number of Vehicles in Accidents as ADAS Feature Adoption Grows to AEB Voluntary Commitment (FIGURE 42) SOURCE: NHTSA (HTTPS://WWW.REGULATIONS.GOV/ (FIGURE 43) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY DOCKETBROWSER?RPP=25&SO=DESC&SB=POSTEDDATE&PO=0&D=NHTSA-2015-0101) Projection Crash Course 2019 Projection Crash Course 2020 Veh Manuf after 9/1/16 Veh Manuf after 9/1/17 Veh Manuf after 9/1/18 and before 8/31/2017 and before 8/31/2018 and before 8/31/2019 0% 100.0% | -5%

90.0% MONEY IS TIME CY2011 CY1997 CY2017 CY2021 CY2031 CY1995 CY2013 CY2015 CY1999 CY2019 CY2019 CY2041 CY2001 CY2037 CY2027 CY2023 CY2033 CY2035 CY2025 CY2047 CY2039 CY2029 CY2043 CY2007 CY2045 CY2003 CY2005 CY2009 80.0% -10% 70.0% -15% 60.0% 50.0% -20% 40.0% -25% 30.0% 20.0% -30% 10.0% -35% 0.0%

Kia Tesla Volvo Audi BMW Ford Toyota Honda Mazda Nissan TOTAL Hyundai Subaru Porsche Mitsubishi Volkswagen Fiat Chrysler Mercedes-Benz General Motors Jaguar/Land Rover Maserati/Alfa Romeo In Crash Course 2018 and Crash Course 2019 we looked at early data on differences in crash frequency, crash characteristics, and crash severity among the same vehicle when equipped or not equipped with ADAS. Comparison of the vehicles with and without ADAS, and the Delta-V of Given the wide array of ADAS technologies in the marketplace today, the variation in how they are the crash (total change in vehicle velocity over the duration of the crash event), also implied ADAS packaged, the different designs and goals of each system, and the rate at which each is entering may be helping to reduce low-speed crashes altogether, and slows the speed of the vehicle prior the marketplace, it is difficult to project what the real long-term impact on auto claim frequency to impact for crashes that might otherwise have had higher Delta-V at crash impact. This year, we and auto loss costs may be. In prior issues of Crash Course, analysis gauged what percentage of compare the potential impact to accident types among a subset of automakers who have met the auto claims might benefit from ADAS features such as frontal crash avoidance, AEB, lane departure NHTSA AEB commitment for 90 percent or more of their vehicles by MY2019, versus a second warning, adaptive headlights, and blind spot monitoring given the mix of auto claims by age and subset of automakers who have met the commitment for 20 percent or fewer of their vehicles loss coverage. The Insurance Institute for Highway Safety/Highway Loss Data Institute (IIHS/ by MY2019 (see Figure 44). Collision loss vehicles in each of the two groups were segmented HLDI) predictions on the increase in percent of registered vehicle fleet equipped with each ADAS by vehicle age group, and comparison was made among each primary point of impact’s share of system to CY 2050 in the U.S. also was used. Finally, pooled data from analyses conducted across volume over a five-year period. For those automakers >=90 percent equipped, current model year automakers on the likely reduction in frequency was incorporated. vehicles saw a 2.9 percentage point drop in their share of losses that were “Front Primary Impact” for Q4’14-Q3’15 (MY2014-MY2016 vehicles) versus Q4’18-Q3’19 (MY2018-MY2020 vehicles). The implication is that as these automakers have more of their vehicles equipped with AEB, volume While research conducted over the last year reached similar conclusions to prior years’ research on share has shifted from direct front impacts to the rear and side impacts. This is reinforced by the ADAS efficacies, automakers have ramped up their incorporation of ADAS features like AEB earlier comparison to the collision losses for automakers <=20 percent equipped, whose current model than the MY2022 commitment date, and the predicted rate of adoption was updated from prior year vehicles saw only a 1.3 percentage point drop in their share of losses that were “Front” primary years’ analysis. As a result, our projection this year shows a slightly faster increase in the reduction impact. When the same comparison was performed for vehicles aged 7-years plus (where for both of vehicles in accidents annually, although decline eventually flattens as ADAS is added to nearly subsets of automakers, the presence of AEB was negligible), the percentage point change was also 100 percent of the vehicle fleet (see Figure 43). So, while this remains only a high-level estimate, 1.3. Finally, the 1.3 percentage point drop was also the very same change in volume share by impact it underscores the inevitable decline in frequency that these systems will have as ADAS market point for all collision loss appraisals across all automakers and vehicle ages over the same five-year penetration grows. This projection also does not account for the potential adoption of aftermarket period (see Figure 45). devices such as those created by Mobileye which have been shown to improve driver safety and could help lead to additional reduction in crash frequency.119

Shift Among Accident Impacts for ADAS Equipped Vehicles (FIGURE 44) Share of Claim Count by Primary Impact — Collision and Liability Losses Q4’14-Q3’15 SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY versus Q4’18-Q3’19 (FIGURE 45) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

Vehicles 90% Equipped with ADAS as of MY2019 - COLLISION LOSS CATEGORY LIABILITY LOSS CATEGORY Percent of Collision Repairable Claim Count per Primary Impact - Current Model Year Vehicles Q4'14- Q4'18- Pct Point Q4'14- Q4'18- Pct Point

Front Left Front Left Front Pillar Right Front Q3'15 Q3'19 Change Q3'15 Q3'19 Change Right Front Pillar Rear Side Other | Front MONEY IS TIME 100% 27.3% 26.0% -1.3% Front 12.5% 12.2% -0.3% Left Front 10.1% 10.1% -0.1% Left Front 7.6% 7.3% -0.3% 90% Left Front Pillar 3.1% 3.4% 0.3% Left Front Pillar 2.7% 2.9% 0.2% 80% Left Qtr Post 2.7% 3.1% 0.4% Left Qtr Post 3.7% 4.1% 0.4% 70% Left Rear 6.3% 6.0% -0.3% Left Rear 7.9% 7.9% 0.0%

60% Left T-bone 4.3% 4.4% 0.0% Left T-bone 5.2% 5.1% -0.1% Rear 18.3% 18.1% -0.2% Rear 39.6% 39.1% -0.5% 50% Right Front 10.8% 10.9% 0.1% Right Front 6.2% 6.1% -0.1% 40% Right Front Pillar 2.9% 3.4% 0.4% Right Front Pillar 2.1% 2.4% 0.2% 30% Right Qtr Post 3.2% 3.7% 0.5% Right Qtr Post 2.8% 3.2% 0.4%

20% Right Rear 5.5% 5.1% -0.4% Right Rear 5.4% 5.1% -0.2% -2.9% drop in Right T-bone 4.6% 4.8% -0.2% Right T-bone 4.0% 4.0% 0.1% 10% Front Impacts 0% MY14-MY16 MY18-MY20 Q4'14-Q3'15 Q4'18-Q3'19

Vehicles Minimally Equipped (<20%) with ADAS as of MY2018 - In February 2019, IIHS/HLDI published results of research conducted on police crash-report data Percent of Collision Repairable Claim Count per Primary Impact - Current Model Year Vehicles from 23 states during 2009-2016 for striking passenger vehicles with and without autobrake among models on which the system was optional. They wanted to identify the types of rear-end crashes Front Left Front Left Front Pillar Right Front in which vehicles with autobrake are involved to help identify where autobrake could be improved Right Front Pillar Rear Side Other to address a broader set of accident scenarios.120 The research found more than two-thirds of the 100% crashes occurred when the road surface was dry, the striking vehicle was moving straight, or the

90% struck vehicle was slowing or stopped. Of the remaining one third of crashes, the crash-involved vehicles with autobrake were turning, struck a vehicle that was turning or changing lanes, struck 80% a non-passenger vehicle, were driving on a snowy or icy road, or were driving on a road with a 70

70% mph or higher speed limit. This data helps underscore why AEB equipped vehicles may be seeing a larger drop in collision front impacts. 60%

50% There is hope that advances in vehicle technology such as ADAS will, over time, lead to fewer accidents, and subsequently fewer people and vehicles to fix. In the near term however, vehicles 40% have many more sensors, cameras, radar, lidar, etc. and complex material construction than 30% ever before. The wide variation in technologies used among the OEs is leading to challenges for repairers, where the importance of understanding each and every individual vehicle has become 20% critical. These changes also mean the industry is making major investments in tooling, training, -1.3% drop in 10% and data to ensure it is equipped to properly return vehicles to pre-accident condition. Near-term Front Impacts challenges abound, and certainly ADAS seems to be another factor driving the trend of fewer but 0% more expensive claims. MY14-MY16 MY18-MY20 Q4'14-Q3'15 Q4'18-Q3'19 © 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 49 CRASH COURSE 2020 COURSE CRASH nearly hit. The vehicle’s collision-mitigating repair experience. With Accident Advisor, braking system prevents the accident. OEMs can: Viewers exhale. • Guide drivers to quickly and safely Volkswagen, Subaru, and Volvo capture our document the accident scene and initiate attention by highlighting advanced driver- the claims process with their insurance

assisted systems (ADAS) in commercials, provider through a singular experience.

encouraging consumers to expect more from | TIME IS MONEY IS TIME their vehicles. Expectations that expand • Connect more seamlessly with repair beyond safety to include technology-aided facilities. Diagnostic scans inform repair assistance, prominently featured in Hyundai’s facilities while performing a thorough ABOUT THE AUTHOR “ghost car” ad, debuting in the OEM’s repair of vehicles. 2020 Super Bowl spot. The self-parking Manjaree Harwood and “unparking” features work to extend Senior Director, • Help drivers replace vehicles following consumer expectations of vehicle assistance. OEM Market an accident. Today, about 20 percent of Development, CCC accidents result in a total loss, and some What these ads, and the OEMs we speak consumers may decide to replace their with, want to make clear is that their vehicles after an accident. relationship with drivers is changing. OEMs are now companions on the driver’s journey, Providing drivers with support during a providing relief and comfort along the way. confusing and stressful time can help OEMs And, this seems to be good for business. distinguish themselves with unique services that drive customer satisfaction and brand BEYOND THE A 2019 survey by Fundera noted that 80 loyalty. percent of U.S. drivers would pay more for a vehicle if it was accompanied by a better Through telematics, OEMs can be a VEHICLE: OEM experience.¹ ADAS and telematics are the gateway to insurers to round-out the driving forces behind the driving experiences vehicle ownership experience. With driver of the future. OEMs are now TECHNOLOGY consent, CCC® VIN Connect connects vehicles from the OEM point of purchase companions on the However, the picture isn’t entirely rosy; to CCC to support insurers who may offer “ these same features increase the cost of personalized insurance rates or reward good driver’s journey, DRIVES ENHANCED vehicle repair. And, even with more new behaviors. CCC® VIN Connect Safety also providing relief technologies, not all accidents can be provides insurers with a clear view of vehicle and comfort along avoided. But, with the assistance of CCC, ADAS features, which is designed to help CONSUMER OEMs can help drivers navigate and recover underwriters factor in vehicle technology and the way." from unplanned events. accident-avoidance features when calculating EXPERIENCE rates. Connecting OEMs and by Manjaree Harwood Vehicle technology provides a platform Drivers for expanded relationships. By helping Heart-pounding music has long been the backdrop to the automakers connect with drivers beyond the sleek silhouette of vehicles gliding flawlessly down the road The relationship between OEMs and drivers initial purchase and — more importantly — in automotive advertising. And while the easy handling and no longer stops at the point of purchase; it after an accident, CCC can help OEMs build beauty of new vehicles remains important, automakers today now extends beyond the vehicle itself. CCC® deeper relationships with their customers. are highlighting much more — safety and simplicity. Think Accident Advisor keeps OEMs connected to about it: Honda’s “What If” advertising campaign, which drivers post-accident, providing a medium 1SHEPARD, MADDIE, “13 SURPRISING BRAND LOYALTY debuted in late 2019, features a new kind of heart-pounding to guide consumers through the claims and STATISTICS.” HTTPS://WWW.FUNDERA.COM/RESOURCES BRAND-LOYALTY-STATISTICS. JULY 23, 2019 moment when a man (a coworker, a father, a husband) is 51 CRASH COURSE 2020 COURSE CRASH

Of the four key megatrends, it is perhaps the connected car that offers the most

immediate benefits, particularly as |

automakers look for ways to remain top of MONEY IS TIME mind among vehicle owners throughout the CCC INSIGHT vehicle life cycle.

Auto sales in the U.S. remained strong in 2019 with 17.1 million new vehicles and nearly 40 million used vehicles sold. Analysts are predicting sales will slow in the coming years, but with vehicles lasting longer than ever before, scrappage rates low, and a decline in miles driven per vehicle annually, there will be many vehicles on the road in the U.S. for quite some time.

Analysis of the demographics of age suggest that the anticipated drop in auto sales and auto claim frequency from Millennials staying put in urban areas and using only ride-hailing instead of purchasing their own vehicle may be much smaller than anticipated. Numerous studies reveal that it is primarily differences in economics not preferences that have led to the delay in life choices by Millennials versus prior generations. The question ultimately becomes how new factors such as more online shopping, greater digital capabilities enabling more work-from- home, an aging population, and higher rates of migration to the West and South will drive auto claim frequency in the future.

Unfortunately, drivers face more congestion and potential for distraction, so despite lower per vehicle exposure to accidents from people driving their vehicles less, the overall count of motor vehicle accidents in the U.S. may not see any meaningful decline for some time. ADAS technology has been shown to be very effective at reducing crashes in the scenarios in which they are best designed to work, but most vehicles on the road are still not ADAS-equipped, and the technology doesn’t address every type of accident scenario. As we’ll discuss later, the ADAS technology is among the new features of vehicles that add to repair complexity, ultimately driving up claim costs for the industry in the near-term.

Automakers and other technology companies continue to make large investments in "ACES," but full-blown Level 5 autonomy will likely take longer than earlier predicted. Billions of dollars are being spent to gear up for electric vehicles, but sales remain tepid. Of the four key megatrends, it is perhaps the connected car that offers the most immediate benefits, particularly as automakers look for ways to remain top of mind among vehicle owners throughout the vehicle life cycle. Using data collected from the vehicle may be the best way yet to understand what customers really use among the numerous technologies now being added to vehicles, helping automakers potentially reduce the cost of unused features but also enable all industry players to better understand their customers. 53 CRASH COURSE 2020 COURSE CRASH

| TIME TO INVEST TO TIME TIME TO INVEST MODERNIZING REPAIR CAPABILITIES

02 Today’s vehicles are very different than vehicles manufactured even 10 or 20 years ago.

Automakers made numerous changes and enhancements to vehicle construction, material types, content and more to improve vehicle crash worthiness, crash avoidance, and vehicle emissions, and to offer new types of capabilities like telematics, adaptive cruise control, and over-the-air updates.

55 CRASH COURSE 2020 COURSE CRASH

All these modifications changed not only the how and what the vehicle is made of, but also resulted the industry can expect to see similar volatility in claim count and makeup in the coming years. in the addition of numerous new vehicle components. Figure 46 illustrates the growing share of These losses are unlikely to benefit from the ADAS technology being added to vehicles; if anything, electronics and software to vehicle costs over the last two decades. Changing consumer preference it will drive the replacement and repair costs of these vehicle claims up further. resulted in many more light trucks versus cars, and more consumers opting for higher trim levels than ever before. Average Total Cost of Repairs by Loss Category, Repairable Appraisal Statistics

(FIGURE 47) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

With the increase in vehicle complexity, repair complexity has also grown. Consumers increasingly | rely on new features such as ADAS, and all industry participants need to be invested in having Total Collision Comprehensive Liability INVEST TO TIME repairs deliver the vehicle back to pre-accident condition. Today that means making access to $4,000 training, tooling, and data available, and using them to provide the best repair outcomes.

$3,500 In this section of Crash Course 2020 we discuss how the industry is investing in capabilities to repair the vehicles of today and tomorrow, and what that means to repair cycle time and repair costs. $3,000

Software and Electronics Share of Vehicle Cost to Exceed Value of Mechanical $2,500 Components by 2020 (FIGURE 46) | SOURCE: ETI/MARTEC $2,000

Mechanical Electronics Software $1,500 2020 40% 30% 30% CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2049

2010 55% 30% 15% CY2019 Share of Overall Claim Count with Primary Impact of Hail/Water

(FIGURE 48) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY 2000 77% 18% 5% % of Claim Count Repair Costs Accelerate 18.6%

In CY 2019, the industry experienced a ramp-up in the annual rate of change in repairable claim 0.1% costs from prior years. The average total cost of repairs overall for all repairable appraisals was $3,225, up 4.9 percent from CY 2018. By comparison, overall inflation for All items not seasonally adjusted increased 1.8 percent in CY2019. Comprehensive and liability losses saw the largest increase in average costs over the last several years, followed by collision losses (see Figure 47). Severe weather in numerous parts of the country over the last several years led to significant swings in volume share of comprehensive losses related to hail and flooding (see Figure 48), as well as to vehicle age of comprehensive losses and the share of which were repairable versus total loss claims (see Figure 49). For example, in Q4’16-Q3’17, Hurricane Harvey struck Texas, flooding large volumes of vehicles, driving the age of comprehensive losses younger than in other periods, and resulting in higher total loss costs. With the severity of storms growing due to climate change,

Analysis of Comprehensive Claims by Vehicle Age and Total Loss versus Repairable Repairable Vehicle Damage Appraisals — Vehicle Mix Statistics by Calendar Year

(FIGURE 49) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY (FIGURE 50) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

Current Yr or CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 Percent Average Total Newer Group 1-3 Years Old 4-6 Years Old 7 Years and | Flagged Loss Adjusted Average Total % Overall % Overall % Overall Older % of $2,425 $2,497 $2,551 $2,597 $2,689 $2,752 $2,870 $2,941 $3,074 $3,225 Avg Repair Cost INVEST TO TIME Total Vehicle Valution Cost of Comp Claim Comp Claim Comp Claim Overall Comp % Chg from prior calendar year 1.1% 2.9% 2.2% 1.8% 3.5% 2.3% 4.3% 2.5% 4.5% 4.9% Loss Amt Repairs Cnt Cnt Cnt Claim Cnt Non-Driveable % 22.6% 21.9% 21.8% 21.1% 20.6% 20.6% 20.3% 20.5% 20.8% 20.5%

Q4'14-Q3'15 13.3% $8,752 $2,818 1.7% 9.8% 11.1% 77.3% % of Claims with Suppl(s) 47.1% 46.9% 47.2% 46.1% 46.5% 48.0% 49.2% 51.4% 51.2% 55.6%

Suppl % of Total Repair Cost 10.4% 10.7% 11.2% 12.1% 12.6% 13.3% 14.4% 16.1% 16.5% 18.1% Q4'15-Q3'16 16.8% $9,107 $3,069 1.8% 12.3% 13.8% 72.2% Avg Vehicle Age 5.79 5.99 6.11 6.19 6.10 6.00 5.90 5.90 5.90 5.94 Q4'16-Q3'17 20.3% $10,008 $3,124 2.5% 14.8% 16.6% 66.2% Avg CCC Regional Value Amt $12,851 $13,133 $14,280 $14,996 $15,022 $15,324 $15,275 $15,790 $16,071 $16,488 Avg Odometer 80,412 82,634 83,875 82,712 83,696 82,206 80,098 79,677 79,507 79,551 Q4'17-Q3'18 18.8% $9,479 $3,280 1.9% 12.7% 17.1% 68.3% Avg Mileage per Vehicle Year 13,899 13,792 13,726 13,372 13,615 13,620 13,576 13,505 13,476 13,392

Q4'18-Q3'19 19.2% $9,878 $3,475 1.7% 12.6% 18.6% 67.0% Parts % Total Repair Cost 38.2% 37.7% 38.2% 38.9% 39.3% 40.1% 39.8% 39.7% 39.9% 40.2% Avg # Parts Repl per Claim 7.9 8.0 8.3 8.7 9.0 9.3 9.4 9.6 9.9 10.5

OEM % of Total Part Amt 63.4% 63.3% 63.1% 62.8% 64.0% 63.7% 62.9% 62.6% 61.7% 61.8%

Labor % Total Repair Cost 42.6% 42.8% 42.3% 41.9% 41.7% 40.9% 41.2% 41.0% 40.4% 39.6%

Avg Labor Hrs per Claim 22.3 22.3 22.4 22.5 22.7 22.8 23.1 23.2 23.3 23.6

To better understand the dynamics behind the change in average appraisal cost, consider the Avg Hourly Body Rate $44.61 $45.01 $45.50 $46.13 $46.67 $47.27 $47.82 $48.87 $50.27 $51.45 primary components contributing to overall repair costs: parts and labor, and an analysis of the % Chg from prior calendar year 1.7% 0.9% 1.1% 1.4% 1.2% 1.3% 1.2% 2.2% 2.9% 2.3% changes within the population of repaired vehicles in terms of vehicle age, manufacturer and type 40.8% 42.2% 41.8% 41.3% 41.3% 40.7% 42.2% 41.8% 40.7% 39.8% (see Figure 50). Growth in the average number of parts replaced per claim as well as more labor Repair % Total Labor Amt hours per claim largely account for the rise in repair costs. Increasing vehicle complexity — both in Total Loss % Vol 15.0% 14.7% 15.1% 14.0% 14.1% 15.4% 16.7% 17.9% 18.4% 19.2% terms of the materials used in the construction of the vehicles, as well as an increased number of Collision Losses % Vol 52.8% 51.8% 52.2% 53.5% 53.9% 54.5% 53.3% 53.3% 53.9% 53.8% standard options, are helping to lift costs. Comprehensive Losses % Vol 16.2% 18.4% 17.5% 15.9% 15.9% 14.8% 16.5% 16.8% 16.0% 16.1%

Liability Losses % Vol 31.0% 29.8% 30.3% 30.6% 30.2% 30.7% 30.2% 29.9% 30.1% 30.0% Parts and Labor Drive Repair Costs Vehicles 7 Years & Older % Vol 36.5% 38.8% 40.7% 42.2% 42.8% 41.8% 38.3% 36.5% 35.4% 35.0% Light Truck % Vol 42.6% 42.6% 42.9% 43.1% 43.4% 44.2% 45.9% 48.1% 50.1% 52.4%

The distribution of repair cost dollars has changed little over the last 15+ years, the timeframe in Asian Vehicles % Vol 43.8% 45.0% 46.3% 47.5% 48.3% 49.0% 49.6% 50.1% 50.5% 50.8% which CCC tracked this information. The most significant change is the increase in both the share Domestic Vehicles % Vol 47.7% 46.2% 44.5% 43.1% 42.2% 41.4% 40.8% 40.5% 40.1% 39.6% of overall dollars spent on replacement parts and sublet/miscellaneous operations (see Figure 51). Replacement parts make up about 40.2 percent of overall repair dollars, but the distribution of European Vehicles % Vol 8.5% 8.8% 9.2% 9.4% 9.5% 9.6% 9.5% 9.5% 9.5% 9.6% overall parts dollars by part type has shifted towards greater non-OE utilization. Luxury % Vol 15.7% 15.7% 15.9% 15.8% 15.6% 15.5% 15.5% 15.5% 15.4% 15.6%

Repair Dollars Distribution by Category CY2010-CY2019 Average Parts Replaced/Repaired per Claim CY2004-CY2019

(FIGURE 51) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY (FIGURE 53) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY REPLACED REPAIRED

Non-paint labor Paint labor Paint materials OEM parts | 18.0 Non-OEM parts Misc Tax Towing Storage EPC INVEST TO TIME 16.0 CY2019 14.0 CY2018 12.0 CY2017 10.0 CY2016 8.0 CY2015 CY2014 6.0

CY2013 4.0

CY2012 2.0

CY2011 0.0 CY2010 CY2011 CY2017 CY2012 CY2013 CY2015 CY2018 CY2016 CY2019 CY2014 CY2010 CY2007 CY2005 CY2008 CY2006 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% CY2009

The average price paid per replacement part experienced only moderate increases at an aggregate level (see Figure 54); however, when looking at some of the top replaced parts — such as bumper At the close of CY 2019, the industry’s share of replacement part dollars was split at 61.8 percent covers, fenders, etc. — the impact of auto tariffs and material costs is more impactful (see Figure OEM versus 38.3 percent non-OEM (see Figure 52). Between CY 2018 and CY 2019 the average 55). As more newer model year vehicles enter the claim mix, there will be an increase in the share number of replacement parts increased from 9.9 parts to 10.5 parts, while the average number of of overall spend per part on OEM parts, as use of non-OE parts may be used less frequently on repaired parts per claim have also increased (see Figure 53). newer model year vehicles, and the actual cost of the parts tends to be higher for the newer model year vehicles as well.

Percent of Total Part Replacement Dollars by Part Type CY1998-CY2019 Average Cost per Replaced Part by Vehicle Age Group, CY2015-CY2019 (FIGURE 52) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY (FIGURE 54) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

OEM AFTERMARKET RECYCLED RECOND OPT OE CY2019 CY2018 CY2017 CY2016 CY2015 100% 90% 80% Total 70% 60% 50% 7 Yrs Plus 40% 30% 20% 4-6 Yrs 10% 0% 1-3 Yrs CY2011 CY2017 CY2012 CY2013 CY2015 CY1998 CY2018 CY1999 CY2016 CY2019 CY2014 CY2010 CY2001 CY2007 CY2002 CY2003 CY2005 CY2008 CY2006 CY2009 CY2004 CY2000 Current Model Yr

$- $20 $40 $60 $80 $100 $120 $140

Focused Parts Analysis Excluding Attachments Such as Clips, Hinges, Bolts, etc. Finding and keeping collision repair technicians remains a primary concern for the industry. Show Cost of Major Parts Still Increasing (FIGURE 55) According to the CREF/I-CAR Snapshot of the Collision Repair Industry 2019, 56.4 percent of the

SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY survey’s respondents reported hiring at least one entry-level technician during the previous 12 months, and the largest source of all technician hiring was hiring from another shop, followed by CY2015 CY2018 CY2019 hired from a non-automotive industry or as first job, and hired from a related automotive industry (see Figure 57).121 Even auto dealers are facing a critical shortage of service technicians — only

$350 about 37,000 auto mechanics are graduating from U.S. technical colleges and training programs |

$300 per year — barely fulfilling half of the roughly 76,000 technicians needed annually to keep pace INVEST TO TIME $250 with retirements and new jobs in the sector.122 This is particularly concerning given the need for $200 more vehicles to potentially require service at a dealership to complete some of the more complex $150 scan and/or calibration procedures, where a shortage of technicians with the needed skills leads to $100 longer overall cycle times. $50 $0 Source of Technicians Hired at Collision Repair Facilities (FIGURE 57) Grille Hood Fender SOURCE: CREF I-CAR SNAPSHOT OF THE COLLISION REPAIR INDUSTRY SURVEY 2019 Lift Gate Trunk Lid Windshield Front Lamps Pick Up Box Rear Lamps Front Bumper Quarter Panel Rear Bumper All Part Groups Radiator Support CY2016 Survey CY2019 Survey 60% Air Conditioner & Heater

50% The average number of labor hours per appraisal uploaded to CCC on behalf of insurance 40% companies in CY 2019 was up by one-tenth of an hour. These same appraisals showed moderate increases in the average hourly labor rate nationally. Combined, these factors help explain the 30% increase in average repair costs overall. The hourly labor rates saw moderate increases year-over- year during the recession, closing out CY 2019 up over 2 percent (see Figure 56). 20%

Average Labor Rates per Labor Category CY2015-CY2019 (FIGURE 56) 10% SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY 0% Body Body Hired from another shop Hired from non-automotive Hired from related Other industry or as ﬁrst job automotive industry (Sheet (Sheet Calendar Metal) Frame Mechanical Paint Paint Metal) Frame Mechanical Paint Paint Year Labor Labor Labor Labor Material Labor Labor Labor Labor Material

AVERAGE HOURLY RATE (WEIGHTED AVERAGE) % CHANGE FROM PRIOR YEAR Materially Different

CY2015 $47.27 $54.15 $80.47 $47.21 $27.78 1.3% 1.6% 2.9% 1.3% 2.5% To improve vehicle crashworthiness and occupant safety, automakers designed vehicles with 123 CY2016 $47.82 $55.20 $82.07 $47.79 $28.31 1.2% 1.9% 2.0% 1.2% 1.9% more energy transfer built into the vehicle design. Combine that with numerous materials, and vehicles are performing differently and handling crash energy differently.124 Stronger materials are CY2017 $48.87 $56.54 $84.98 $48.68 $29.07 2.2% 2.4% 3.6% 1.9% 2.7% being used in the front structure or in the upper and lower rails to reduce crush in the occupant CY2018 $50.30 $58.32 $87.34 $49.89 $29.92 2.9% 3.0% 2.6% 2.4% 3.0% compartment, often creating challenges when identifying where the energy from a crash traveled through a vehicle, and where that vehicle might be damaged. The materials used in vehicle CY2019 $51.45 $59.78 $89.91 $50.96 $30.70 2.3% 2.5% 2.9% 2.1% 2.6% construction also are changing. As automakers race to meet the ever-stricter corporate average fuel economy standards, they increasingly use lighter-weight materials such as aluminum, boron, ultra-high strength steel, and magnesium to reduce the overall weight of the vehicles themselves. As technology reduces the cost and speed of vehicle development through advances in electronic

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 63 CRASH COURSE 2020 COURSE CRASH modeling and CAD systems, manufacturers can select different materials and joining methods Most vehicle repairs today now recommend completion of a pre-repair scan, post-repair scan, within a single vehicle body structure. The result is a potential combination of materials all used in a and often an in-process scan to get a comprehensive understanding of what is damaged and to single vehicle structure. Figure 58 illustrates the growth in the percent of annual overall appraisal verify proper repair. Additionally, repairers must identify what electronic systems the vehicle has, volume where the vehicle itself was known to include at least one component comprised of a which of them has been damaged, and which components need to be removed, replaced, R&I’d, light-weight material such as aluminum, boron, magnesium, high-strength steel, ultra-high-strength disconnected, and/or refinished.125 Diagnostic trouble codes (DTC’s) triggered by the scan also need steel, hydroformed steel, or sandwiched steel. In CY 2007, only 13.4 percent of all the vehicles to be researched, as OE repair procedures are updated frequently.

for which an appraisal was generated included one or more of these lightweight vehicles in its | construction; by CY 2019 that number grew to 78.6 percent. In CY 2015, numerous OEs released position statements underscoring the critical nature of INVEST TO TIME performing these vehicle scans. As a result, industry awareness and the practice of scans has Percent of Overall Vehicle Appraisal Count for Vehicles with One or More grown. In first quarter CY 2016, less than a half-percent of appraisals included an entry denoting a Lightweight Material Part (FIGURE 58) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY “scan/diagnosis/health check/astech” operation (see Figure 59). By Q4 CY 2019 that percentage NOTE: APPRAISAL MAY OR MAY NOT HAVE INCLUDED REPAIR/REPLACEMENT/ETC. OF A LIGHTWEIGHT MATERIAL PART. grew to 49.5 percent. Note that the absence of a scan entry in the appraisal does not necessarily mean that the scan was not performed — just not recorded in the appraisal itself. As the completion Current Model Yr & Newer 1-3 Yrs 4-6 Yrs 7 Yrs Plus Total of the scan(s) now is a standard part of most repairs, the industry has adapted, and the average fee 100.0% per scan fell from $110 per scan in Q1 CY 2016 to just $68 per scan by Q4 CY 2019. 90.0% 80.0% Percent of Repairable Appraisals by Vehicle Age Group with Appraisal Line for “Scan/Health/asTech/Diagnose” (FIGURE 59) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY 70.0% NOTE: QUERY TO PULL DATA UPDATED FOR CRASH COURSE 2020 AND DATA MAY DIFFER FROM PREVIOUSLY PUBLISHED DATA. 60.0%

50.0% 2016-Q1 2016-Q2 2016-Q3 2016-Q4 2017-Q1 2017-Q2 2017-Q3 2017-Q4 40.0% 2018-Q1 2018-Q2 2018-Q3 2018-Q4 2019-Q1 2019-Q2 2019-Q3 2019-Q4 70% 30.0% 60% 20.0% 50% 10.0% 40% 0.0% CY2007 CY2011 CY2015 CY2016 CY2017 CY2018 CY2019 30% 20% With a greater number and variety of high-tech materials used in the construction of their vehicles, 10% automakers are releasing studies and bulletins that underscore the criticality of following the recommended repair procedures and methods so the vehicle is repaired safely and will perform 0% as designed in any subsequent collisions. Growth in electronic vehicle content — items added to Current Year or 1 - 3 Years Old 4 - 6 Years Old 7 Years and Older Total Newer Group address vehicle safety or convenience — adds to the overall cost and complexity of repair and the need to understand recommended repair procedures. As the number of connected cars grows, the opportunity to capture data — such as a vehicle health scan that includes diagnostic trouble codes — facilitates the ability for all parties involved in the Electronics and Software Growth Drives Need for repair to understand the nature of the vehicle damage. In early January 2020 CCC and American Additional Skills and Tooling Honda Motor Co. announced plans to integrate Honda’s scanning cloud technology into the CCC™ X data exchange. CCC® Diagnostics is designed to enable repair facilities to request and receive scan data, clear codes, and rescan directly from within the system they use every day, minimizing Consumer demand for ADAS to help avoid or mitigate an accident has grown dramatically. These hardware and software requirements and eliminating the need for manual entry of scan report technologies are driving the rapid addition of microprocessors, sensors, cameras, radar and other data. Finally, the ability to integrate the data into the overall repair file provides documentation technologies into vehicles, and numerous software programs to connect to this broad array of about the scans, should questions arise later. sensors and cameras.

According to MOTOR’s Guide to Estimating, the reset of an electronic module (of any kind) is not Percent of Repairable Appraisals by Vehicle Age Group with Appraisal Line for included in the replacement time for that module. Rather, the reset time is separate, and can vary “Calibration/Reprogram/Flash” (FIGURE 60) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY widely due to wide variation among individual vehicles and system complexity. Similarly, computer module re-learn and electronic memory functions reset (after battery disconnect) are also not 2019-Q4 2019-Q3 2019-Q2 2019-Q1 2018-Q4 2018-Q3 included with any labor operation. Many of the vehicles equipped with ADAS require “zero-point 2018-Q2 2018-Q1 2017-Q4 2017-Q3 2017-Q2 2017-Q1 calibration” following replacement of system parts and sensors.

| TIME TO INVEST TO TIME Calibration essentially tells a vehicle’s internal computers to readjust the cameras, sensors and Total other technology to function according to the vehicle’s OE specifications.126 A dynamic calibration may require the vehicle be test driven at certain speeds for a specified period — and that it 7 Years and Older be driven on certain types of roads with specific types of lane markings all while plugged into a scan tool.127 A static calibration often requires OE specific targets and equipment, such as 4 - 6 Years Old heating devices, lasers, jigs, cameras and Doppler devices.128 Conducting a calibration can vary by manufacturer, and requires OEM service information as well as scan tools. The time to recalibrate 1 - 3 Years Old these sensors and cameras may vary depending on the complexity of the systems, and differs from Current Yr or a scan, which is done to identify problems within a system. Perhaps most concerning is the fact Newer Group that there is no indicator — like a DTC or dash light — to inform the repairer that the system is not 0% 1% 2% 3% 4% 5% 6% 7% 8% calibrated, or the sensors are not aimed correctly. Following the vehicle-specific procedures and requirements, and documenting they have been completed properly, is critical to being able to demonstrate accurate calibration has occurred. Calibration Entries — Average Cost, Maximum Fee Identified, and Standard Deviation When a part such as a distance sensor is replaced, additional database labor time of 0.2 to 0.6 of Fees (FIGURE 61) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY hours is commonly added as an entry such as “FRONT BUMPER Add for distance sensor.” In addition to this, some repairs include separate manual entries for calibration, depending on the Average fee for calibration entry components damaged or the vehicle repair requirements. The specific parts requiring calibration (Includes ﬂat fees and labor) Maximum fee on single appraisal Std Dev of Fees are not always identified clearly or at all. Unsurprisingly, the majority of calibration line entries in appraisals identify a specific ADAS feature such as blind-spot monitoring sensor, distance sensor, 2017-Q1 $168 $4826 ”Window/Module reprogram” $197 camera, parking sensor, lane departure, adaptive cruise control, as well as mechanical parts such as occupant sensors, steering angle sensors and tire pressure monitoring sensors, and finally parts 2017-Q2 $172 $6250 “Reprogram Headlamps at dealer” $214 such as headlamps. 2017-Q3 $177 $4341 “Replace and Recalibrate Lane Ad” $229

2017-Q4 $182 $4149 “calibrate and repair” $223 In Q4 CY 2019, 2.6 percent of all appraisals included an additional entry for “calibration,” “reprogram,” “flash,” etc., up from 0.9 percent in Q1 CY 2017 (see Figure 60). As additional 2018-Q1 $182 $4270 “recalibrate windshield, clear” $226 supplements come in for repairs from Q4 CY 2019, the percentage likely will climb to meet or 2018-Q2 $189 $7791 “complete install/tune/calibrat” $233 exceed the 2.8 percent of Q3 CY 2019. There is, however, a wide range in these types of fees, with certain entries including not just the cost of calibration, or a fee to “Drive to and from calibration,” 2018-Q3 $196 $5852 “SRS system recalibration and S” $236 but the cost of additional components that may have been found to need replacement during the 2018-Q4 $210 $8941 “Replace and programm occupant” $257 calibration. Figure 61 illustrates the average fees per calibration estimate line (includes manual 2019-Q1 $217 $6094 “dealer install and calibrate f” $254 and database lines), and the maximum fee per quarter from Q1 CY 2017 to Q4 CY 2019. 2019-Q2 $219 $5316 “Dealer Distance Sensor Reprogr” $236

2019-Q3 $250 $6610 “RR DIFF CHANGE AND CALIBRATE” $281

2019-Q4 $232 $8564 "Re-calibrate distance sensor -" $264

Completing the proper diagnostics, determining which components are included in the repair that OEMs are increasingly focused on encouraging the repairer community to review the OEM repair might require calibration, and reviewing the OEM repair procedures can help a repairer incorporate procedures prior to repairing a vehicle. the calibration into the repair plan up front. These steps could potentially help avoid added costs and repair time identified later in supplements, that can also lead to lower customer satisfaction Over the last several decades, repairers have faced numerous challenges from changing vehicle with the repair. Analysis of calibration entries included in appraisals between CY 2017 and CY 2019 technology. The pace of change among vehicles today is accelerating, driving up the costs of suggest the industry has room for improvement here — less than 40 percent were included in the

operating a professional facility. Not surprisingly, the collision repair industry also is transforming. original estimate of record, and of those calibration with fees of $1000 or more, only 9 percent Between CY 2007 and CY 2013, nearly 4,000 collision repair facilities in the U.S. went out of | were included on the original E01 (see Figure 62). business, with just over 41,000 in operation as of CY 2018 (see Figure 63).131 In recent years, the INVEST TO TIME number of independent body shops has grown nearly 3 percent annually as economic recovery, Manual Calibration Entries Included in CY2017-CY2019 Appraisals — Percent miles driven, extreme winter weather, and improved new vehicle sales began to support increases Included per Estimate Version (FIGURE 62) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY in auto accident frequency. Overall, however, growth within the segment over the last 15 years has NOTE: ESTIMATE OF RECORD (E01); 1ST SUPPLEMENT (S01); 2ND SUPPLEMENT (S02); 3RD OR MORE SUPPLEMENT (S03-S99) seen a compound annual growth rate of -0.5 percent.132 These independent repairers and dealer body shops will do an estimated $41 billion in collision repair work in 2019, a number anticipated to All Manual Calibration Entries Manual Calibration Entries >=$1000 grow further over the next several years.133 The Romans Group estimates 89 percent of the overall collision repair revenue is insurance paid, while 11 percent is consumer paid.134 37.6% 36.7% 37.3%

31.7% U.S. Shops Specializing in Collision Repair CY2002-CY2018 (FIGURE 63) | SOURCE: AUTO CARE ASSOCIATION. AUTO CARE FACTBOOK 2020. 22.0% 47,000 16.4% 46,000 45,000 9.0% 9.3% 44,000 43,000 42,000 E01 S01 S02 S03-S99 41,000 40,000 39,000 38,000 Driving Change for Collision Repairers 37,000

CY2011 Changes to both the material composition of the vehicle and the added procedures to account CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009 CY2010 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 for repairs of new software and electronic components also mean repairers must make significant investments in added tooling and training and staff for their locations. Purchasing the tools and Analysis of the auto claims processed by CCC on behalf of its insurance and collision repair software specified by each individual OEM results in multiple software license fees, multiple customers reveals the percent of appraisals generated at a shop on behalf of an insurance company welders, rivet guns, scan tools, calibration targets and more. As shops struggle to keep up with the was 47.1 percent for CY 2019, after peaking at 49.7 percent in CY 2017 (see Figure 64). DRP share necessary investments, they face the reality that they may not be able to repair every vehicle from of repairable appraisal volume was 39.9 percent for CY 2019, down from 40 percent in CY 2016 every OEM in the future. This is just one of many major decisions that must be made by repairers and CY 2017 as more carriers shifted volume for customers among various methods of inspection today. Others include whether to participate in insurers’ direct repair programs, whether to expand, options such as photo/virtual estimating. sell or join a franchise or banner group, and whether to become an OEM certified shop, and if so, which OEM(s). Analysis of repairable DRP volume reveals national MSOs continue to see significant growth in share. Between CY 2000 and CY 2019 the share of overall repairable DRP appraisal count going Many more OEMs are developing their own OEM certified repair programs that require varying to national MSO’s quadrupled, growing from 5.8 percent to 39.1 percent (see Figure 65). Share levels of ASE certification, I-CAR training, I-CAR welding certification, and specific tool and of DRP appraisal volume going to dealers was 3.3 percent in CY 2019. According to Supplement equipment.129 The programs are designed to help drive brand recognition, customer satisfaction Advisory, dealerships share of the overall U.S. collision market dollars was just below 20 percent with service and repair, keep OEM warranties in place, and ensure eligible trade-ins and lease as of Q1 2019, trailing chain and franchise shops at 26.6 percent, and independent stores at 54.1 returns meet requirements for certified pre-owned programs.130 As vehicle complexity grows,

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 69 CRASH COURSE 2020 COURSE CRASH percent.135 With OEMs increasingly pursuing growth of their own certified collision repair networks, DRP Repair Cycle Times Continue to Grow While dealer share of volume may grow. Repairer Productivity Slows Share of CCC National Industry Repairable Appraisal Volume where Appraisal Generated at the Shop on Behalf of the Carrier as DRP, Open Shop or Service Center When an insurance vehicle claim is made, insurance carriers often have limited visibility into (FIGURE 64) SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY | whether their customer gets the vehicle repaired or not, and if so, where. The exception to this is

claims where the vehicle is repaired at a collision repair facility that is part of that carrier’s direct |

repair program (DRP). Through updates provided by the repairer to the insurer, data on repair cycle INVEST TO TIME DRP Svc Ctr Open Shop times, and repair satisfaction is available. 60.0% The percent of DRP appraisals written that convert to repairs tends to be higher as the repair cost rises (see Figure 66). For the period of Q4’18-Q3’19, 70.3 percent of all repairable DRP appraisals 50.0% converted to repairs. Separate analysis of the capture rate (estimates written to repair orders captured) for repairers’ walk-in business reveals just over 50 percent of walk-in business converts 40.0% to repairs, and the average cost per walk-in repair order averaged nearly 30 percent less than DRP repairs. 30.0%

Between Q4’16-Q3’17 and Q4’18-Q3’19, the average DRP repair cost rose from 8.3 percent from just 20.0% over $3,000 to $3,251. A comparison of repairs by repair cost dollar range shows the shift toward the higher dollar ranges over the same period (see Figure 67). Over the same two-year period, 10.0% the average number of days from vehicle drop-off to vehicle pickup, or ‘keys-to-keys’ has increased from 9.7 days to 10.2 days (see Figure 68). Consistent with prior analysis, the vehicle in to repair 0.0% started days average and the repair complete to vehicle out days average have improved or stayed CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 steady — the increase has occurred within the repair time itself.

Share of Annual DRP Appraisal Volume per Shop Type (FIGURE 65) DRP Appraisals Q4’18-Q3’19 — Percent Appraisals Written Converted to Repairs SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY (‘capture rate’) (FIGURE 66) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY NOTE: NATIONAL MSOS INCLUDE FRANCHISE; INDEPENDENTS INCLUDES 2&3-LOC MSOS. REGIONAL MSOS ARE REPAIRERS WITH 4-PLUS LOCATIONS IN A SINGLE STATE. Total Repair Cost Ranges 70.3% DEALERSHIP INDEPENDENT NATIONAL MSO REGIONAL MSO $10,000.01 & Up 59.0% 100% 90% $5,000.01-$10,000 72.0% 80% 70% $4,000.01 to $5,000.00 74.6% 60% 50% 40% $3,000.01 to $4,000.00 74.8% 30% 20% $2,000.01 to $3,000.00 74.1% 10% 0% $1,000.01 to $2,000.00 70.0%

CY2011 CY2000CY2001CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009 CY2010 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 $500.01 to $1,000.00 60.5%

$0.01-$500.00 53.6%

DRP Repairs by Repair Cost Dollar Ranges (FIGURE 67) Among the data that further illustrates the impact of greater complexity in vehicle repair are the

SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY following. The percent of vehicle repairs with at least one supplement has grown from 76 percent to 82 percent over the three year period from Q4’16-Q3’17 to Q4’18-Q3’19; the share of total repair cost $0.01-$500.00 $500.01 to $1K $1,000.01 to $2K $2,000.01 to $3K added during supplements grew from 12 percent to 14.5 percent; average labor hours per repair grew from 25 hours to 25.4 hours; and parts share of overall repair cost grew from 41 percent to 42 percent $3,000.01 to $4K $4,000.01 to $5K $5,000.01-$10K $10,000.01 & Up with average parts replaced per repair growing from 11.5 to 12.8 parts (see Figure 69).

Unfortunately, as complexity has grown, repairer productivity has declined. Labor hours per repair INVEST TO TIME Q4'18-Q3'19 10.4% 27.2% 20.7% 13.5% 8.8% 14.9% 2.9% 1.8% day and labor hours per shop day have both fallen over the last three years, with some of the most significant declines among the lower dollar repairs (seeFigure 70). Worth noting is the growing share of overall spend on the ‘Miscellaneous percent of Total Repair Cost’ where most of the additional fees for scan and calibration roll up — suggesting these fees are adding to costs across Q4'17-Q3'18 12.3% 28.2% 20.2% 12.8% 8.2% 13.5% repairs of all dollar ranges (see Figure 71). As noted above, as of Q4 CY 2019, 49.5 percent of all 2.2% 2.6% appraisals included an entry for a pre-, post-, or in-process vehicle scan, with an average fee of $68 per scan, and an average of $85 per claim.

Q4'16-Q3'17 13.5% 28.6% 19.7% 12.4% 7.9% 13.1% 2.4% 2.4%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% DRP Repairs Key Repair Metrics Illustrating Growing Repair Complexity (FIGURE 69) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

% Change % of Supplements Avg # Repair Cycle Time — Average Days — DRP Repairs (FIGURE 68) in Avg Appraisals % of of Parts SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY Avg Total TCOR with Total Avg Labor Parts % Replaced Cost of Versus Supplement Repair Hrs per of Total per

Q4’16-Q3’17 Q4’17-Q3’18 Q4’18-Q3’19 3 Year Change Period Repairs Prior Year (s) Cost Appraisal Repair Cost Appraisal

Repair Vehicle in Repairs Repairs Vehicles in Repairs Repairs Vehicle in Repairs Repairs Vehicle in Repairs Repairs to Repairs Started to Completed to Repairs Started to Completed to Repairs Started to Completed to Repairs Started to Completed Cost Started Days Repairs to Vehlicle Started Repairs to vehicle Started Repairs to Vehicle Started Repairs to Vehicle Q4'16-Q3'17 $3,003 n/a 76.3% 12.2% 25.0 40.94% 11.5 Range Avg Completed Out Days Days Avg Completed out Days Days Avg Completed Out Days Days Avg Completed Out Days Days Avg Avg Avg Avg Days Avg Avg Days Avg Avg Q4'17-Q3'18 $3,078 2.5% 75.1% 13.2% 25.0 41.30% 11.9

$0.01-$500.00 0.3 1.5 0.7 0.3 1.6 0.8 0.3 1.8 0.7 0.0 0.3 0.0 Q4'18-Q3'19 $3,251 5.6% 82.3% 14.5% 25.4 41.75% 12.8

$500.01 $1,000.00 0.3 2.7 0.8 0.3 2.8 0.8 0.3 2.9 0.7 0.0 0.3 0.0

$1,000.01 to $2,000.00 0.5 4.5 0.9 0.5 4.6 0.9 0.5 4.8 0.8 0.0 0.3 0.0

$2,000.01 to $3,000.00 0.8 6.8 1.0 0.7 6.9 1.0 0.6 7.1 0.9 -0.2 0.3 -0.1

$3,000.01 to $4,000.00 1.1 9.0 1.1 0.9 9.0 1.1 0.9 9.2 1.0 -0.2 0.2 -0.1

$4,000.01 to $5,000.00 1.4 11.1 1.2 1.2 11.1 1.2 1.0 11.2 1.1 -0.3 0.1 -0.1

$5,000.01-$10,000 2.1 15.6 1.4 1.8 15.4 1.3 1.6 15.5 1.3 -0.5 -0.1 -0.1

$10,000.01 & Up 3.7 27.5 1.9 3.3 27.5 1.9 2.9 26.9 1.9 -0.8 -0.6 0.0

Total 1.0 7.7 1.0 0.9 7.9 1.0 0.8 8.4 1.0 -0.1 0.7 0.0

Repairer Productivity Declines as Repair Complexity Grows Miscellaneous Percent of Total Repair Cost per Repair Cost Dollar Range

(FIGURE 70) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY (FIGURE 71) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

Q4'18-Q3'19 Q4'17-Q3'18 Q4'16-Q3'17

Labor Hours per Shop Day Total

Q4'16-Q3'17 Q4'17-Q3'18 Q4'18-Q3'19 $10,000.01 & Up |

4.0 $5,000.01-$10K INVEST TO TIME 3.5 $4,000.01 to $5K 3.0 2.5 $3,000.01 to $4K 2.0 $2,000.01 to $3K 1.5 1.0 $1,000.01 to $2K 0.5 $500.01 to $1K 0.0 $0.01-$500.00 Total 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% 8.0%

$0.01-$500.00 $500.01 to $1K $5,000.01-$10K $1,000.01 to $2K$2,000.01 to $3K $3,000.01 to $4K$4,000.01 to $5K $10,000.01 & Up As more repairs include additional steps previously included infrequently or not at all, the importance of clear processes to ensure these steps are completed properly and thoroughly, and clearly documented, grows. Blueprinting the repair, identifying up front what types of electronic Labor Hours per Repair Day components such as ADAS may require calibration, completing scans, etc. have all become more Q4'16-Q3'17 Q4'17-Q3'18 Q4'18-Q3'19 important up front to understand as much about the likely repair cost and repair time for each individual vehicle. As repair costs rise, supplement frequency and return rates rise, while on- 3.0 time delivery rates fall. Creating and following a standard set of procedures up front to remind 2.5 estimators and repair technicians to check for items such as ADAS features may help combat the 2.0 inevitable increases in each of these as repair costs continue to rise (see Figure 72). 1.5 1.0 Q4’18-Q3’19 DRP Repairs per Repair Cost Range -As Repair Costs Rise, On Time Repairs and Delivery Fall; Supplement Frequency and Return Rates Rise 0.5 (FIGURE 72) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY 0.0 $0.01-$500.00 $500.01 to $1K $1,000.01 to $2K $2,000.01 to $3K $3,000.01 to $4K Total $4,000.01 to $5K $5,000.01-$10K $10,000.01 & Up Total

100.00% $0.01-$500.00 $500.01 to $1K $5,000.01-$10K $1,000.01 to $2K$2,000.01 to $3K$3,000.01 to $4K$4,000.01 to $5K $10,000.01 & Up 90.00% 80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% % of On Time Delivery % of On Time Repairs % of Repairs with % Vehicles Returned Supplement(s) Post-Delivery

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 75 CRASH COURSE 2020 COURSE CRASH As Repair Costs Have Risen, So Too Has Total Loss With younger vehicles experiencing some of the largest growth in total loss frequency, the average age of total loss vehicles has started to decline. The younger mix, as well as continued shifts Frequency towards more expensive vehicles such as light trucks and non-domestic vehicles, drove up the average total loss adjusted vehicle value in CY 2019 (see Figure 75). Higher vehicle repair costs and longer-lasting vehicles are key factors behind the continued

increase in vehicle total loss frequency. The percent of vehicle appraisals flagged total loss has Total Loss Valuations — Vehicle Mix Statistics by Calendar Year (FIGURE 75) grown from 14 percent in CY 2013 to 19.2 percent in CY 2019 (see Figure 73). Total loss frequency | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY for non-comprehensive losses has risen nearly 5 full percentage points during that same period, INVEST TO TIME with vehicles four years and younger seeing the largest growth rate over the last five years (see Figure 74). CY 2017 less CY 2015 CY 2016 CY 2017 Harvey CY2018 CY2019

Percent of Vehicle Appraisals Flagged Total Loss CY2013-CY2019 (FIGURE 73) SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY Final Valuation Amt Avg $8,867 $8,943 $9,272 $9,125 $9,500 $9,861 % Chg from prior year 1.7% 0.9% 3.7% 2.0% 4.1% 3.8% Non-Comprehensive Losses All Loss Categories Avg Vehicle Age 10.01 9.87 9.64 9.73 9.66 9.60 25.0% Avg Odometer 119,079 116,839 113,965 114,768 114,196 113,566 Avg Mileage per Vehicle Year 11,894 11,834 11,817 11,800 11,822 11,833 20.0% Light Trucks % Vol 36.3% 36.6% 37.7% 37.5% 38.7% 40.0% SUV's % Vol 19.9% 20.7% 22.2% 22.0% 23.3% 24.9% Car Final Val Amt Avg $8,129 $8,042 $8,143 $8,041 $8,231 $8,437 15.0% % Chg from prior year 0.6% -1.1% 1.3% 0.0% 2.4% 2.5% Truck Final Val Amt Avg $10,164 $10,506 $11,133 $10,925 $11,508 $11,986 10.0% % Chg from prior year 3.3% 3.4% 6.0% 4.0% 5.3% 4.2% Collision % 62.9% 61.0% 59.9% 61.3% 62.8% 63.3% 5.0% Comprehensive % 25.9% 24.5% 16.7% 14.8% 14.1% 13.8% Liability % 11.2% 14.5% 23.4% 23.9% 23.0% 22.9% Collision Adj Vehicle Value Avg $9,817 $9,806 $10,000 $10,000 $10,399 $10,771 0.0% Comprehensive Adj Vehicle Value Avg $8,833 $9,080 $10,116 $9,357 $9,576 $9,916 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 Liability Adj Vehicle Value Avg $5,874 $5,906 $6,067 $6,067 $6,316 $6,584 Asian Vehicles % Vol 45.4% 47.2% 48.4% 48.5% 49.5% 50.1% Domestic Vehicles % Vol 46.2% 44.1% 42.7% 42.6% 41.6% 40.7% European Vehicles % Vol 8.4% 8.7% 8.9% 8.9% 8.9% 9.2% Percent of Non-Comprehensive Repairable Appraisals Flagged Total Loss by Asian Veh Final Val Amt Avg $9,028 $8,941 $9,051 $8,943 $9,202 $9,438 Vehicle Age (FIGURE 74) | SOURCE: CCC INFORMATION SERVICES INC. % Chg from prior year -0.2% -1.0% 1.2% 0.0% 2.9% 2.6% Domestic Veh Final Val Amt Avg $8,038 $8,270 $8,769 $8,608 $9,118 $9,590 CY2015 CY2016 CY2017 CY2018 CY2019 % Chg from prior year 3.2% 2.9% 6.0% 4.1% 5.9% 5.2% 50.0% European Veh Final Val Amt Avg $12,175 $12,162 $12,944 $12,580 $12,879 $13,273 45.0% % Chg from prior year 1.5% -0.1% 6.4% 3.4% 2.3% 3.1% 40.0% Theft % Vol 1.4% 1.4% 1.3% 1.4% 1.3% 1.3% 35.0% Vehicles Current Yr & Newer % 2.4% 2.5% 2.6% 2.5% 2.2% 2.1% 30.0% Vehicles 1-3 Years % 11.8% 13.2% 14.4% 14.1% 14.6% 14.3% 25.0% Vehicles 4-6 Years % 12.2% 14.0% 16.2% 16.0% 17.8% 19.2% 20.0% Vehicles 7 Years & Older % 73.6% 70.3% 66.8% 67.4% 65.4% 64.3% 15.0% 13.7% 14.1% 14.5% 14.5% 14.4% 14.7% 10.0% Luxury % Vol 5.0% 0.0% Current 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TOTAL Year or Year Years Years Years Years Years Years Years Years Years Years Years Years Years Years Years Plus Newer Old Old Old Old Old Old Old Old Old Old Old Old Old Old Old Old

A shifting vehicle mix also contributed to higher vehicle repair costs. Over the last 10 years, the share of vehicle claims volume per repair cost dollar range has shifted to the higher repair brackets just as it has for vehicle repairs discussed above (see Figure 76). Unfortunately, as the repair cost rises, so too does the likelihood that the vehicle will be a total loss. While less than 10 percent of vehicle appraisals with repair cost of $3,000 or less were flagged total loss, that number jumped to Michele Dufresne over 60 percent for those with repair cost of $10,000 or more. Chief Claims Officer |

Sentry Insurance INVEST TO TIME Crash tests conducted by organizations such as Thatcham, IIHS/HLDI, and automakers themselves, suggest that, over time, the prevalence of ADAS may help to reduce accident frequency, and reduce claim severity. Despite most of the ADAS designed to avoid specific types of accidents up to certain speeds, there is some hope that more of the higher speed, higher impact crashes could be mitigated by these systems. This could potentially stabilize repair costs and total loss frequency for newer models over time. Repairer productivity for ADAS-equipped vehicle repairs will improve over time, and the cost of the technologies may stabilize. Near term, however, the industry can expect fewer but more expensive repairs. What are we thinking about?

Bodily injury severity and social inflation continue to be Repairable Non-Comprehensive Appraisal Volume Shifts to Higher Cost Repairs concerns as we head into a new decade. The ability to

CY2010-CY2019 (FIGURE 76) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY implement solutions to moderate these severity trends will be key for any carrier hoping to grow market share CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 profitably. We have a duty to our policyholders to solve rate 35.0% inflation driven by extreme trends. One way: reaching out to legislators and working together to discourage legal 30.0% “ frameworks that magnify social inflation and bodily injury 25.0% severity.

20.0% Autonomous vehicles and ADAS technology will have 15.0% a significant impact on the insurance model. Quickly developing accurate pricing models will give a carrier a 10.0% significant competitive advantage. This technology also 5.0% creates an opportunity to build a telematic data-sharing environment that could reduce accident rates and drive an 0.0% improved experience for insureds and insurers alike.

$500.01-$1K Privacy concerns are real and must be accounted for as we $0.01-$500.00 $1,000.01-$2K$2,000.01-$3K$3,000.01-$4K $5,000.01-$6K$6,000.01-$7K$7,000.01-$8K$8,000.01-$9K $4,000.01-$5LK $9,000.01-$10K$10,000.01-$15K $15,000.01-$20K$20,000.01-$30K$30,000.01 & Up approach critical mass in our ability to leverage telematic data. There’s vast potential to scientifically and credibly predict the causes and likelihood of bodily injuries. Pricing, claims, and underwriting functions can all benefit from more accurately pricing risks and predicting driver behavior. The consumer can also benefit from paying rates which match their driving behavior and usage — and therefore, their risk."

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 79 CRASH COURSE 2020 COURSE CRASH I repaired in my dad’s shop 30 years ago. In it arrives on their premises. Conversely, the the 80’s, the industry wasn’t as advanced insurance carrier could assess if the vehicle and most of us were satisfied with the advent was likely a total loss based upon the images of computers and electronic estimating. captured at the accident scene. These are just a Remember when EzEst was the next big few examples as to how technology is helping thing? We have come a long way since those to deliver a quality repair and better customer days. experience.

When you hear the words “repair quality” Once the vehicle arrives at the shop, the INVEST TO TIME it is logical to automatically default to various touch points impacting the quality thinking of the vehicle repair at the collision of the repair grow exponentially. During the ABOUT THE AUTHOR repair facility. I am going to challenge that repair plan process, it is critical to identify the thinking and conditioned response. Repair advanced driver assistance systems (ADAS) Dan Risley quality doesn’t start when the technician adversely impacted and repair procedures Vice President, disassembles the vehicle. Repair quality for the damaged parts, while simultaneously Quality Repair and starts before the vehicle arrives at the documenting these steps for repair quality Market Development, collision repair facility. and consistency. The use of products such as CCC Repair Methods provides the user with repair When a vehicle is in an accident, the first procedures once the decision to make a repair notice of loss (FNOL) historically involves or replace a part has been made. the insured notifying their insurance agent and filing a claim. Today, the technology in Throughout the repair process, it has become many vehicles is changing that paradigm: an essential part of every repair to document A NEW ROLE FNOL is now the FNOA. CCC’s announcement the 5 Ws (who, what, when, where regarding our partnership with Volvo Cars and why). by Dan Risley is a great example of this shift. In these situations, the car manufacturer is the first Who performed repairs on the vehicle? entity notified of an incident. This is the • Identify the technician who performed the beginning of the “repair quality” chain and a scanning and calibrations. philosophical change to the claims handling During the repair When I was hired by CCC as the Vice President of Quality process. The industry was accustomed to • Identify the technician who performed the plan process, it is Repair and Market Development, the position did not “ being reactionary, waiting for the customer sectioning and welds. critical to identify previously exist. The company decided to create a role to notify the insurance company. Today, within the organization signifying the importance and CCC’s technology is enabling the car manufacturer What repairs were completed? the advanced driver commitment to support customers’ desire for consistent, to immediately reach out to the vehicle • Capture images of all the work performed quality repairs. What role can an information provider owner, verify if the occupant(s) are OK, assistance systems throughout the repair process. play in repair quality? The obvious and easy answer is and provide CCC products which capture (ADAS) adversely information, but the more complex answer is delivering pertinent information at the scene of the • Document quality control procedures and solutions that assist industry stakeholders — repairers, impacted and repair accident and allow the vehicle owner to identification as to who verified the quality. insurers, independent appraisers, OEMs, and part suppliers procedures for the notify their insurance carrier of their accident — throughout the process. It starts with the first notice of and schedule an appointment to begin When were the repairs completed? damaged parts, accident (FNOA) and continues through repair planning, repairs. All of this can be completed before • Document when or if pre- and post-repair repairing the vehicle and finishing with the delivery of while simultaneously the vehicle owner leaves the scene of the scanning was performed. the vehicle to the owner. Why was this role created? Our accident. In theory, the collision repairer • Document dates and time stamps of the documenting these customers’ needs have changed as vehicle technology has could order the parts for this vehicle before steps for repair quality rapidly evolved. Today’s car isn’t the 1983 Oldsmobile 88 that repair. and consistency."

81 Continued A NEW ROLE Scott Kohl Vice President & Manager, APD Field Operations, Global Retail Markets Liberty Mutual Insurance Where were the repairs completed? shop an opportunity to audit repairs and an opportunity to coach their technicians in real • Identify the location of the sectioning time. performed • Identify welds, welding and/or bonding We recognize the need for collision repairers procedures locations. to have documentation that supports the How do you/your company spend Why were specific repairs completed entirety of the repair. This documentation or not performed? is necessary at the time of the repair, and it may be even more important after their • Document the decision-making process your time? customer leaves the repair facility. including unrelated damage that was not repaired. Remember, repairers, insurers, independent At Liberty Mutual, we spend a lot of our time thinking about appraisers, OEMs, and part suppliers all To assist repairers in documenting the 5 Ws, innovation and how we can transform our customers’ help in providing a seamless experience in we offer a product called CCC® Checklist. In repairing vehicles. CCC remains committed to experience with us. Consumer expectations for service and my opinion, every shop in the country should building, enhancing and delivering products personal transportation are changing and along with that, be using a product like Checklist. It gives that meet the needs of the industry today the shop owner information when questions we’re changing to modernize what we can deliver and how with a watchful eye to the future. arise regarding a repair and provides the “ we deliver it."

83 CRASH COURSE 2020 COURSE CRASH

The sharp drop-off in new vehicle sales that occurred with the last recession slowed the

pace at which new automotive technology |

hit the street. INVEST TO TIME

With new vehicle sales exceeding 17 million each year for the past five years, many more insurance claims and vehicle repairs are for newer model year vehicles. As of CY 2019, vehicles CCC INSIGHT of model years 2016 through 2020 accounted for 34.4 percent of all appraisals generated (including those flagged total loss), and for 42.7 percent of those repaired (see Figure 77). Many of these vehicles include options such as automatic braking, crash warning or avoidance systems, parking assist, lane departure warning, and more. Unfortunately, when a vehicle is equipped with these features, additional repair requirements related to them may also be driving up repair costs and subsequently total loss frequency.

More than ever before it is essential for the automotive claims and collision repair industries to stay current on new technologies, tooling, and training. Ensuring the vehicle repair is completed in a manner that follows recommended repair procedures can also head off any potential unplanned returns of the vehicle and leave the ever-more demanding customer with a more expensive and longer repair time than anticipated.

Repairing a vehicle and ensuring it is returned to its owner in pre-accident condition has never been easy. As more computers, sensors, substrate metals, new welding and joining techniques, and more are added to vehicles, the repair process has become even more challenging. Each vehicle accident is different and has its own unique characteristics. Repairers must subsequently develop a unique set of steps or repair plan for each individual repair. With additional requirements based on the construction and electronic content of the vehicle itself, the insurance company, OE certified program, and/or any regulatory requirements for documentation, there could be potentially millions of combinations of process steps required to properly complete a repair. Ensuring processes exist to check-off and document the specific steps required for each type of repair has become more important than ever — not only to complete a proper repair, but to also have the documentation to prove it.

CY2019 Vehicle Claim and Repair Volumes by Vehicle Age Group (FIGURE 77)

SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

Current Yr or Newer Group 1 - 3 Years Old 4 - 6 Years Old 7 Years and Older

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% All Vehicle Repairable Vehicle Repaired Vehicles Total Loss Vehicles Appraisals Appraisals 85 CRASH COURSE 2020 COURSE CRASH

| TIME FOR SERVICE FOR TIME TIME FOR SERVICE HUMANIZING THE EXPERIENCE

03 Rising repair costs, higher total loss frequency and costs, and longer repair times have added more pressure to an industry struggling to keep up with growing consumer demand to turn an unpleasant, disruptive experience into a satisfying one.

87 CRASH COURSE 2020 COURSE CRASH Insurers Use Digital to Improve the Data from numerous surveys on customer service delivery across all industries reveals 75 percent of consumers feel self-service is a convenient way to address customer service issues, and 67 percent Claims Experience prefer self-service over speaking to a company representative (see Figure 79).138 However, there is a key difference depending on the type of interaction: customers may prefer the use of self-service The numerous components, steps, and players involved in a vehicle accident and claim can be tools for transactional-type activities, but still value personal contact with an agent or service representative for high-value interactions regarding price and policy coverage and first notice

confusing, disconnected, and subsequently dissatisfying. Process improvements alone don’t close 139 this satisfaction gap. Research continuously shows the power of insurance companies’ customer of loss. | communication and engagement. Over the last several years, carriers invested a great deal in SERVICE FOR TIME technology designed to improve these areas. Improvements to their websites, more self-service Consumers Prefer Self-Service Over Speaking to Company Representative and mobile capabilities are just some of the areas of focus. (FIGURE 79) | SOURCES: NUANCE ENTERPRISE SURVEY AND COLEMAN PARKES FOR AMDOCS STUDY (HTTPS://WWW.ZENDESK.COM/ RESOURCES/SEARCHING-FOR-SELF-SERVICE/)

According to the J.D. Power 2019 Auto Insurance StudySM, the most important components of the insurance companies’ customer experiences are clear delivery and transparency around A recent consumer survey commissioned by Nuance Enterprise found that a majority of respondents thought positively about self-service premium increases, delivery of functional digital self-service tools, and quality customer service.136 Consumers expect support of a digital experience from their insurance companies in claims, policy updates, insurance shopping, etc., and improved access to these has helped drive auto insurance customer satisfaction to its highest level since CY 2000 (see Figure 78).137 75% 67% of survey resondents of respondents said they said self-service is a preferred self-service convenient way to over speaking to a J.D. Power 2019 U.S. Auto Insurance StudySM Auto Insurance Customer Satisfaction address customer company representative. Index (on a 1,000 point scale) (FIGURE 78) | SOURCE: LAJDZIAK, ROBERT M. “HIGH CUSTOMER SATISFACTION LEADS TO NEW service issues.

CHALLENGES FOR INSURERS.” SEPTEMBER 18, 2018. HTTPS://WWW.CARRIERMANAGEMENT.COM/FEATURES/2018/09/18/184296.HTM. While consumers are eager to embrace their favorite brands’ online self-service capabilities, another study by Coleman Parkes for Amdocs points out self-service is only an option if it’s done right.

840 831 826 819 820 818 810 811 804 800 794 91% 40% 790 787 of survey respondents of customers contact a 777 said they would use an call center after they 780 772 772 768 online knowledge base have looked for answers 765 765 766 if it were available and to their question via 760 761 760 tailored to their needs. self-service. 747

740 This means the majority of calls to the call center can be dramatically reduced, if brands provide specialized self-service, which customers prefer. 720

700

CY2011 CY2000CY2001CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009CY2010 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019

As auto premiums increased, many consumers opted to shift to higher dollar deductibles. CY2019 Collision Losses — Share of Volume per Deductible Dollar Range and A comparison of all vehicle claims submitted as a collision loss between CY 2001 and 2019 Estimate Method of Inspection (FIGURE 81) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY underscores this shift as claims with a deductible greater than $500 has grown from 3.8 percent to 22.5 percent (see Figure 80). $0.01 to $250 $250.01 to $500 $500.01 & Up 100.0%

Volume Share per Deductible Dollar Range — Vehicle Claims, Collision Loss Category 90.0% C2001-CY2019 (FIGURE 80) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY |

80.0% SERVICE FOR TIME $0.01 to $250 $250.01 to $500 $500.01 & Up 70.0% CY2019 CY2018 60.0%

CY2017 50.0% CY2016 CY2015 40.0% CY2014 30.0% CY2013 CY2012 20.0% CY2011 CY2010 10.0% CY2009 0.0% CY2008 Sta Appraiser Independent Photo Open Shop DRP Total CY2007 Appraiser CY2006 CY2005 CY2004 Repairable Appraisal Volume by Estimate Method of Inspection and Repair Cost CY2003 Dollar Ranges, CY2019 (FIGURE 82) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY CY2002 CY2001 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% $0.01-$500.00 $500.01 to $1K $1,000.01 to $2K $2,000.01 to $3K $3,000.01 to $4K $4,000.01 to $5K $5,000.01-$10K $10,000.01 & Up 100% As deductibles rise, and even minor repairs become expensive, many consumers are unsure at the 90% time of loss whether to have their vehicle repaired. A comparison of collision claims by the method of inspection (MOI) for the initial estimate of record (E01) and deductible range raises an interesting 80% question. Are consumers with higher dollar deductibles more willing to have their initial inspection 70% done by the insurer versus a shop? And, conversely, are those consumers with lower deductibles 60% more willing to take their vehicle to a shop from the outset (see Figure 81)? While we don’t know how much the deductible factors into consumers’ choice of method of inspection, comparison of 50% the volume of appraisals written per dollar range by inspection type suggests nearly 30 percent of 40% photo appraisals fall below $1,000 (including any supplement changes), and nearly 60 percent fall below $2,000 (see Figure 82). Customers choosing photo MOI in CY 2019 received their estimate, 30% on average, in 1.4 days, versus 4.5 days across all MOIs. With photo MOI, insurers provide expedited 20% information to those insurance companies’ customers who may be unsure about repair of their 10% vehicle due to deductible or other reason (see Figure 83). Transparency provided with fewer steps is a key benefit of this digital experience. 0% DRP IA Sta Open Shop Photo Total

CY2019 Percent of CCC Repairable Appraisals where Last Estimate Assignment Growth within the hyper-competitive private passenger auto insurance market remains a challenge. Sent to Estimate Upload Within Specified Hour Range and Overall Average Days Last Market share continues to shift to the top 10 carriers (see Figure 85), and data prepared by CB Estimate Assignment Sent to Estimate Upload Insights reveals the number of property & casualty insurance filers between 2008 and 2018 fell (FIGURE 83) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY nearly 10 percent (see Figure 86), and only 71 carriers have net written premiums of more than $1 billion.142 % <= 12 Hrs % 12-48 Hrs % 48-120 Hrs % > 120 Hrs

Overall Last Estimate | Total 20.9% 34.4% 20.8% 23.9% Assignment Sent to

Estimate Sent Days Avg SERVICE FOR TIME PHOTO 52.0% 32.6% 8.7% 6.8% Share of Private Passenger Auto Direct Premiums Written (FIGURE 85) Total 4.5 SOURCE: NAIC, AUTO INSURANCE REPORT OPEN SHOP 18.1% 25.9% 22.8% 33.2% PHOTO 1.4 STAFF 22.7% 47.8% 20.7% 8.7% OPEN SHOP 6.2 STAFF 2.0 IA 8.4% 25.1% 31.7% 34.8% CY2000 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 IA 4.8 100.0% DRP 11.5% 24.4% 22.8% 41.3% DRP 7.8 90.0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 80.0%

70.0% Customer retention is a huge benefit gained from improving digital channel experiences, yet there 60.0% are additional benefits that help insurance carriers with their loss adjustment expenses. Data from 50.0%

J.D. Power 2019 U.S. Auto Claims Satisfaction Survey reveals that insurers improved or remained 40.0% consistent on all factors that comprise the total claim satisfaction score: first notice of loss, claim 30.0% servicing, estimation process, repair process, rental experience, and settlement.140 Faster cycle time 20.0% across the entire claim — from making the second contact after first notice of loss, making the 10.0% initial estimate, sending the vehicle to the repair facility, informing the customer of a settlement 141 0.0% offer, and making the initial payment. From the 2017 and 2018 studies, J.D. Power found Total Share Top 5 Total Share Top 10 Total Share Top 15 Total Share Top 25 satisfaction with the estimation process rose from 859 to 871 if insurers used the photos or videos provided by the customer, but if the carrier still had to send out an adjuster, satisfaction fell to 839 (see Figure 84).

Number of property & casualty insurance filers, CY2008-CY2018 (FIGURE 86) SOURCE: NAIC, A.M. BEST (CB INSIGHTS “FUTURE OF INSURANCE”) J.D. Power Auto Claims Satisfaction by MOI (2017 + 2018) (FIGURE 84) SOURCE: SCHMITT, KYLE. “WHY CARRIERS STRUGGLE TO DIFFERENTIATE WITH DIGITAL AND END UP PAYING A PENALTY.” WWW.JDPOWER. COM, MAY 31, 2019. 2,842 2,831

2,794 Photo Estimate Only Traditional MOI Tried Photo Estimate, but needed traditional MOI 2,769 2,743 880 2,706

869 869 871 871 870 2,666 867 866 867 862 861 863 2,639 859 858 2,620 2,606 2,600 851 853 852

841 839 837 834

CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 Index FNOL Estimation Body Shop Rental Car Claim Servicing Settlement

Digital experiences infused with photo and AI can help insurers at first notice of loss better identify “Who Do U.S. Internet Users Trust for Recommendations When Shopping?” whether the vehicle is likely repairable or not, and identify what is the most appropriate MOI based, (% of respondents, by generation, February 2019) (FIGURE 87) in part, on their customer and vehicle characteristics. For example, CCC Smart Total Loss™ solution SOURCE: ORACLE, “ONE SIZE DOESN’T FIT ALL,” JULY 30, 2019. WWW.EMARKETER.COM. can help insurers more accurately predict if a vehicle is likely repairable or a total loss with a single photo. And, the CCC® Smart Estimate solution applies CCC’s estimating logic and AI to vehicle Gen Z (18-24) Millennials (25-39) Gen X (40-54) Baby Boomers (55-75) collision photos to predict repair requirements and suggest estimate lines — including parts likely 100%

required to complete the repair — for human estimators to review, edit, and advance. As vehicles 90% | become more connected, the data from the vehicle itself may supplement digital technology such 80% SERVICE FOR TIME as photos, and AI, to enhance decision-making precision and further streamlined claims processes. 70% Insurers want and need to make things simple for their customers but have limited opportunities 60% given the nature of the product they are selling. Claims is a key opportunity — but the challenge is 50% there are numerous entities that must be relied on to deliver a quality experience — whether that is the repairer, the rental car company, the parts supplier, etc. 40% 30%

Understanding which technologies to implement and how to implement to help with the handoff 20% between parties and to manage the performance of those business partners is a challenge. Enabling 10% processes to support those handoffs becomes critical as insurance companies’ consumers expect 0% Friends, family, colleagues Employee in a store Fellow consumers online Inﬂuencer the claims experience to be as seamless as ordering something on Amazon. The key difference is that the accident was not an experience the consumer chose. Technology helps reduce the time spent by claims processors on the mundane, administrative nature of claims, frees them up and Comparison of Repair KPIs for Vehicles Returned Post Vehicle Pick-up versus enables them to spend more time on the human-touch, customer-service-oriented aspects of claims Vehicles Not Returned (FIGURE 88) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY processing that technology can’t yet (and may never) replace. Providing the human support and NOTE: COMPARISON CONDUCTED OF Q1 CY2019 DRP REPAIRS WITH COLLISION FRONT IMPACT RETURNED/NOT RETURNED empathy at the right points while facilitating the digital experience remains job one. Not Returned Returned

Transparency of Repair Ave Total Cost of Repairs $3,835 $5,260 88.2% 93.1% Repairers are central to delivering consumer claim satisfaction and continue to embrace tools and % of Appraisals with Supplement(s) processes to meet expectations. With National Safety Council data reporting accident frequency Supplement % of Total Repair Cost 19.0% 22.9% per driver in the U.S. at 11 percent in 2017,143 very few consumers have regular or repeated Labor % of Total Repair Cost 34.6% 32.2% experience with a vehicle accident and subsequent repair. Among the top reasons that a consumer Avg Labor Hrs per Appraisal 25.8 32.6 selects a collision repairer per data from IMR Inc.’s are ‘Good Prior Experience’ (32.6 percent), Repair % of Total Labor Amt 35.5% 34.2% ‘Recommendation’ (13.1 percent), and ‘Convenient Location’ (12.4 percent).144 Compare this to Parts % of Total Repair Cost 48.7% 50.4% broader consumer survey data showing who internet users trust for recommendations and what Ave # of Parts Replaced per Appraisal 19 24.6 becomes clear is repairers must rely more than ever on good reviews from prior repairers’ customers by word of mouth or via online reviews (see Figure 87). Analysis of customer satisfaction data Non-OEM % of Part Amt 41.8% 38.0% conducted by CCC shows quality repairs and few returns lead to higher scores; yet, if the shop wants Misc % of Total Repair Cost 5.5% 7.0% positive customer recommendations, it needs to make sure the service is great and customers are Vehicle In To Repairs Started Days Avg 0.7 1.0 kept informed. In other words, quality is table stakes — service gets repeat business. Repairs Started to Repairs Completed Days Avg 9.0 12.6 Repairs Completed to Vehicle Out Days Avg 0.8 1.3 In order to better understand whether there are characteristics of repairs where the customer Vehicle In To Vehicle Out Days Avg 10.5 14.9 returns the vehicle for additional work post repair pickup, CCC completed analysis of collision loss Labor Hrs per Repair Day 2.9 2.6 front impact DRP repairs from Q1 CY 2019 (time period selected to allow for customers to have time Labor Hrs per Shop Day 2.5 2.2 to use vehicles post vehicle pick-up). Figure 88 provides a comparison of metrics that illustrate several differences among the 10.6 percent of the repairs returned to the repairer after the vehicle... % of On Time Delivery 32.8% 29.5%

Transparency of Repair is continued on page 99

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 95 CRASH COURSE 2020 COURSE CRASH and to be differentiated in the marketplace? It’s easy to understand why trust is such an Several things, including: overwhelming component of satisfaction in collision repair. First, repair is a personal • Being customer-centric and evaluating experience. For 90 percent of repairs, the effectiveness of every touchpoint– the vehicle is the primary vehicle in the online, on phone, and in-store. household, 75 percent are owned (not leased), and 70 percent of consumers are • Having a mobile-first (digital) approach

using a shop for the first time. It’s disruptive | to engagement that seamlessly connects and inconvenient, and with new cars costing SERVICE FOR TIME to in-store and on phone engagements. more than $35,000 and being the second • Collecting and analyzing customer most expensive household expense, getting ABOUT THE AUTHOR feedback and integrating processes to it repaired right is important. close the gaps on failed engagements Michael Kolb and continually reinforce good behavior. Time needs to be taken with each repair to Vice President, • Developing ways for consumers to help help the owner/consumer understand: Consumer Product themselves, meeting the 67 percent of Management, CCC consumers who prefer self-service over • What does the consumer know about speaking to a company representative. today’s repair process and claims • Focusing on outcomes that result process? The consumer experience in quality, trust, transparency, and should build knowledge and confidence. convenience. Accidents and repair happen once every seven years for a vehicle. Taking time And, in this day of mobile, digital, AI-infused to help an inexperienced consumer engagement, this last element often means understand the process will benefit the CUSTOMER a human touch. As much as customers repair by setting proper expectations. appreciate a seamless, connected, digital experience — they want it personal and • What is damaged, how it will be EXPERIENCE TAKES curated and they want to have a human repaired, and why is that important? The accessible in the event things break down consumer experience needs to deliver transparency. Providing an estimate The (Magid) study CENTER STAGE is the first step but explaining the estimate and answering every question a found that ‘trust’ was Establishing Trust “ by Michael Kolb consumer has is critical. An experienced ranked as the most In 2019, CCC partnered with Magid, a estimator can proactively answer important element In 2010, Gartner reported that 36 percent of companies consumer experience research firm, to questions and take the time to ensure the perform a benchmark study on the state of customer understands the severity and in selecting a repair competed primarily on the basis of customer experience. In 2020, that number is 90 percent. Nine of ten companies collision repair. This study included 2,000 complexity of repair — and the intricacies facility, and the compete on the experience they deliver to a customer every primary decision makers who experienced a of restoration. element of trust also day. This means that how you deliver on your customer’s collision repair in the previous 24 months and • Why is the shop qualified to do the expectations is what matters, and it’s more important than was representative of the U.S. population. repair and who will be repairing it? The carried through to the the price you charge. From the study, there was one attribute that consumer experience needs to integrate repair and return of floated to the top in terms of importance in authenticity and familiarity. Shops creating a satisfying consumer experience — In the aftermath of a vehicle accident, the idea of customer spend a lot of time training technicians, the vehicle." Trust. The study found that ‘trust’ was ranked experience resonates. Consumers want and expect the claims securing certifications, and investing as the most important element in selecting vehicle repair or settlement process to be empowering, in equipment to perform high-quality a repair facility, and the element of trust also seamless, convenient, transparent, trustworthy, and repairs. It’s worth it to take time to carried through to the repair and return of responsive. So, what do collision market stakeholders need explain what the relevant plaques on the the vehicle. to be thinking about to deliver a great consumer experience

97 CRASH COURSE 2020 COURSE CRASH

Continued Transparency of Repair continued ...was picked up by the customer, versus those repairs that were not. For repairs returned, higher average repair cost, higher supplement frequency, higher parts replacements, longer overall cycle time, and lower shop productivity suggest these were more complex repairs than those repairs CUSTOMER EXPERIENCE not returned. Not surprisingly, distribution of repairs returned by dollar range skewed towards the

higher dollar repairs (see Figure 89). |

TAKES CENTER STAGE SERVICE FOR TIME Share of Repair Volume per Repair Cost Ranges: Vehicles Returned Post Vehicle Pick- wall for their vehicle means and maybe • Integrating the proper technology that up versus Vehicles Not Returned (FIGURE 89) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY even talk with a technician who has complements the people and process NOTE: COMPARISON CONDUCTED OF Q1 CY2019 DRP REPAIRS WITH COLLISION FRONT IMPACT RETURNED/NOT RETURNED repaired a similar make or model. and enables the organization to achieve Returned Repaired Cnt Not Returned Repaired Cnt • When it will be repaired and is the repair desired outcomes. The latest trends are $30,000.01 & Higher on-track? The consumer experience systems and applications that connect $20,000.01 to $30,000.00 needs to be proactive and timely with seamlessly. As first notice of loss becomes a claim, and a claim becomes $15,000.01 to $20,000.00 communications, notifications, and $10,000.01 to $15,000.00 a repair, and repair becomes a restored messaging. We live on our phones, yet $9,000.01 to $10,000.00 vehicle, how are all of those activities consumers today rarely make or accept $8,000.01 to $9,000.00 and touchpoints between insurer, repair phone calls. Find ways with intelligent $7,000.01 to $8,000.00 texting and notifications to inform your facility, and OEM handled? How is the $6,000.01 to $7,000.00 customers about the repair, status, consumer empowered and connected $5,000.01 to $6,000.00 payment, and enable them to review across the journey? $4,000.01 to $5,000.00 your business post-repair. $3,000.01 to $4,000.00 An industry executive once told me that $2,000.01 to $3,000.00 $1,000.01 to $2,000.00 successful digital transformation to evolve Delivering on a customer experience that $500.01 to $1,000.00 the consumer experience is like trying to checks all of these boxes requires a customer $0.01 to $500.00 change all four tires on a car that’s rolling experience strategy that simultaneously 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% improves your business in three areas. down the highway at 65 miles an hour. There are lots of moving parts and you can’t pause Interestingly, the average vehicle model year for vehicles returned was 2013.8 versus 2013.7 for your business to make the changes. You those not returned, and distribution by vehicle age group was also nearly identical. Roughly 39 • Hiring and developing the right people are trying to align your people, evolve your percent of repairs for both vehicles returned and not returned were for model years 2016-2019. at all levels of the organization. Your processes, and integrate your technology — A pre-, in process-, and/or post-repair scan was included in 50 percent of the vehicle repairs not people are the lifeblood of your all while innovation throws new things at you returned, versus 53 percent for those returned. A calibration was included 4.5 percent for those organization. Create a culture of trust at an ever-increasing pace. vehicles not returned versus 8.5 percent for vehicles returned. And while the specific details and accountability and empower them to prompting each return were not available for analysis, a comparison of where in the appraisal fulfil their role. process the calibration occurred helps underscore the overall vehicle and repair complexity that An area of focus at CCC is to help our are potentially driving up return rates. Figure 90 compares the frequency which calibration was • Creating and rigorously following the customers keep the arrow aimed at their added in the orirignal estimate of record versus later supplements, and also shows what percent of right processes to deliver a consistent customers experiencing a collision repair. the calibrations per estimate version included ‘dealer’ in their description: 16.7 percent of calibration experience whether you are one location By using research to better understand the entries were added in the third or higher supplement for vehicles returned post pick-up, 84 percent or hundreds of locations. Every business consumer mindset, we can help our repair higher than the 9.1 percent added at the same time for vehicles not returned. Additionally, of the has its secret sauce for how it operates customers implement technologies and best- 16.7 percent added in the supplement S03-S99, 20.6 percent of those included ‘dealer’ in the line to deliver exceptional experiences. in-class processes to deliver their desired description. While the vehicles returned may not have been returned for reasons related to an Scaling process to deliver consistent consumer journey. experiences may be difficult as location electronic system that needed calibration, the data does minimally suggest that repairs returned numbers increase, acquisitions happen, tend to have a higher repair cost and include more specialized repair tasks than those not returned. and employee turnover occurs.

Share of Repair Volume with Calibration Enries per Estimate Version: Vehicles Customer Satisfaction Metric Comparison: Vehicles Returned Post Vehicle Pick-up Returned Post Vehicle Pick-up versus Vehicles Not Returned (FIGURE 90) versus Vehicles Not Returned (FIGURE 91) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY NOTE: COMPARISON CONDUCTED OF Q1 CY2019 DRP REPAIRS WITH COLLISION FRONT IMPACT RETURNED/NOT RETURNED NOTE: COMPARISON CONDUCTED OF Q1 CY2019 DRP REPAIRS WITH COLLISION FRONT IMPACT RETURNED/NOT RETURNED Not Returned Returned

Percent of Calibration Entry(s) Percent of Calibration Entries Survey % of On Time Delivery 89.6% 71.2% | Added per Estimate Version including ‘dealer’ in line desc Repair Satisfaction % 98.3% 68.4% TIME FOR SERVICE FOR TIME Estimate Version Not Returned Returned Not Returned Returned Kept Informed % 96.2% 86.4% Avg Recommend Shop Score 9.7 8.1 Estimate of Record (E01) 35.9% 23.7% 19.0% 11.8% Ave Recommend Insurer Score 9.5 9.0 1st Supplement (S01) 41.2% 34.7% 42.0% 47.1% Avg Quality Score 9.8 8.3 2nd Supplement (S02) 13.9% 24.9% 24.0% 20.6% Avg Serivice Score 9.8 8.7 Avg Cleanliness Score 9.9 8.8 3rd Supplement + (S03-S99) 9.1% 16.7% 12.0% 20.6% Avg Insurer Handle Claim Score 9.6 8.9 Vehicle Out to Survey Response Days Ave 6.4 8.6

Figure 91 provides comparison of customer satisfaction for the vehicles returned versus not Role of the Automaker Beyond Vehicle Purchase returned. Perhaps not surprisingly, the repairs that were returned had signficantly fewer satisfied customers across every category surveyed, including Kept Informed and Avg Service Score. As Despite another year of record sales, OEMs continue to face a future of potentially fewer auto sales, mentioned above, these categories correlate most to whether or not the customer will recommend and very different types of consumer expectations re: personal mobility. Many automakers have the shop. Unexpected additional work with today’s more complex repairs may be unavoidable, begun to transform from strictly vehicle manufacturers to companies that will facilitate all types of but anticipating them and creating a better process to accompany the customer through the personal mobility in the future. As discussed above, significant investment in the "ACES" continues. process could result in better service and recommend shop scores. Including alerts or reminders for estimators and technicians based on the anticipated repair to help set customer expectations up front and throughout the repair could potentially help better manage the experience of customers Customer retention is just as critical to an automaker and its dealers as it is to all other companies who may have to return their vehicle for additional work post-repair pick-up. While costly in terms in every industry. According to Black Book and The Appraisal Lane, when factoring in all potential of additional repair dollars, time and lost productivity, these repairs don’t also have to lead to revenue from total lifetime gross per unit profit including F&I and referral business, the lifetime unhappy customers. value of a customer to a dealership is more than $50,000.145 Creating an owner experience that encourages they stay within the brand with subsequent vehicle purchases is key. Unfortunately, a vehicle accident may not only be an extremely disruptive experience for the vehicle owner but may Many collision repairers today use technology to “digitize the customer journey” through the also potentially be the reason why a consumer might defect. Research has shown 50 percent of OE vehicle appraisal and repair processes. This technology offers repairers’ consumers the type of customers who disposed of their vehicle within 18 months of an accident did so for repair-related experience they have come to expect — automating many of the manual steps needed to keep the reasons (27 percent for damage suffered in the repair and 21 percent for quality of the repair).146 customer informed throughout the repair — and it frees up the collision repairer’s time so they can focus more on the actual repair of the vehicle. With vehicle complexity growing, repairers will face further demands on their resources, making access to information and analytics mission critical to With CCC® Accident Advisor, the OE is central to making a bad experience less bad. The ability their success. Using data and digital capability to manage customer expectations, even for difficult to use data from the vehicle, data on the accident itself, and known data about the vehicle owner repairs, could help repairers set and better manage customer expectations. In the consumer survey can help facilitate an end-to-end claims experience that is digitally enabled, transparent, with conducted with the research firm Magid, the ability to delivery transparency throughout the repair fewer and smoother hand-offs among all participants. After confirmation is made that the driver process was central to customer satisfaction. and passengers are uninjured and do not need emergency services, a link is immediately texted to their smartphone, which opens a web-based application that takes drivers through important steps following an accident. Accident Advisor prompts the driver to capture facts and photos of the accident scene, offers them the option to digitally notify their insurer and to search for collision repair shops, helping to speed claims handling, reduce repair cycle time, and better manage the overall post-collision process.

3. Identifying a shop or forgo the repair altogether. This last option 4. Waiting for the shop to open is a reality for all shops, but that’s where customer service comes in. 5. Calling the shop to schedule an appointment 6. Waiting for appointment day/time

What Repair

7. Driving to shop |

8. Waiting for estimate to be delivered Customers Say SERVICE FOR TIME Last year, CCC commissioned Magid, a When shops leverage photo estimating, consumer-centered research firm, to survey ABOUT THE AUTHOR several steps can be eliminated, and consumers about the repair experience. The results reveal that body shops should Bart Mazurek consumers get the added value of consider investing in soft skill development Vice President, convenience — requesting and receiving for the front of the shop. Magid’s mystery ASG Consulting, estimates from anywhere, any time of day or night, while building a trusting relationship shopping research found that most auto shop CCC with the shop virtually. employees are missing key opportunities to earn trust from their customers, including:

The entire process can take hours instead • Not calling customers by name of days. • Not reliably returning communication (online specifically) Are Consumers • Not asking consumers if they have PHOTO ESTIMATES questions or concerns Ready? • [Providing] little guidance regarding next For several years, my colleagues at CCC steps CAN BUILD TRUST have been tracking photo-estimating In order to build trust, the front office usage and adoption. Our data shows that employees need to think more like service for Q3 2019, 14.9 percent of all insurance providers and less like technical experts. When shops leverage VIRTUALLY vehicle appraisals were written using a Customers will respond better if they are photo estimating, virtual or photo method of inspection, up guided through the process. Staff should “ by Bart Mazurek from basically 0 percent in 2016. An even assume that customers do have questions, several steps can stronger example of the trend toward photo since they are out of their comfort zone be eliminated, and For thousands of consumers with light vehicle damage from estimating is that some early adopters of the in a repair facility, but they may not feel driveway bumps and parking lot taps, the process for getting technology are using the photo channel for empowered to ask them. According to consumers get a repair estimate can be cumbersome. It’s also wildly different up to 70 percent of the auto estimates. Magid, “providing consumers with answers the added value than the digital experiences they enjoy in other areas of their to questions they didn’t even know they needed to ask will help build trust aspect and of convenience — lives. Every day, many of us use our smartphones to pre- For shops, photo estimating holds a different order and pay for our morning coffee at Starbucks®, book value proposition: photo estimating can be increase their confidence in the process.” requesting and restaurant reservations through OpenTable®, and order just a powerful lead generation tool. It works receiving estimates about anything through mass online retailers. Think about the during and after-hours. Several steps are Technology Helps process for getting an estimate from a body shop. It’s a multi- automated, including requesting an online from anywhere, any step, multi-day process, including: estimate and sharing damage photos to Automate Smaller digitally. This keeps the process moving for time of day or night, Repairs while building a 1. Doing research eager consumers who want to quickly gauge 2. Reading reviews the damage to their vehicle so they can A poor experience with the virtual interaction trusting relationship decide if they want to make a claim, self-pay diminishes the likelihood of converting with the shop virtually." 103 Continued PHOTO ESTIMATES CAN Kevin Prettner AVP Service Operations BUILD TRUST VIRTUALLY American Honda Motor Co. it to a scheduled repair. But a thorough, generation of photo estimates in a timely professional approach to virtual estimates and consistent manner. For example, have can increase overall shop revenue (and build estimators write the photo-based estimates trust with the consumer). While it is possible during known down times and reply to that a shop could see an increase in the the customer in 24 hours (or less). Include How do you/your company spend number of overall estimates being written, a standard letter that answers the most the total time spent with vehicle owners will common customer questions such as how be greatly reduced. to schedule a repair appointment and the your time? shop’s hours. It’s possible that shops will have Photo estimating makes the process virtual fewer low dollar appointments in the shop, and leaves the estimators more time to but more overall low dollar estimates due to In a rapidly evolving industry such as ours, we’re keeping interact with live customers, guiding them the online channel. Be prepared to handle the close to the data and trends that will help inform how we through the repair process, answering volume effectively. According to JD Power, questions, nurturing leads, and focusing on grow and change to meet the needs of tomorrow’s drivers. “Expectation setting and clearly articulating For American Honda, investments in enabling technologies more complex, high-dollar repairs. Many next steps are the two most important shops take the trust-building to a new level factors in making the customer feel at ease.”¹ that can drive efficiencies and help our business partners by leveraging digital checklists to ensure that — collision repairers, dealers, etc. — make better, safer quality repairs procedures were followed Photo estimating is an inevitability — “ decisions is paramount." throughout the repair process. Read more consumers will come to expect the about Checklists in my colleague Dan Risley’s convenience and control it provides. article on page 80. Preparation is required for shops to embrace it and realize meaningful benefits from Preparing Your it. Estimating and customer interaction processes will need to be reviewed for Operations professionalism and thoroughness. Photo- based estimates should be turned around Introducing a new technology or process quickly with a focus on customer service. requires planning to help make your shop Today’s digital consumers have high operationally ready to leverage the capability. expectations, and the shop must be prepared Photo estimating may lead to an increase to provide a solid, informative customer in the volume of requested estimates, experience or face the consequences of a and processes should be structured to negative review, which they will be willing to incorporate the new tools to reach high write. customer satisfaction levels.

1JD POWER, WHY CARRIERS STRUGGLE TO DIFFERENTIATE WITH The key is to have your standard operating DIGITAL AND END UP PAYING A PENALTY, procedures structured to support the KYLE SCHMIDTT, MAY 31, 2019.

Data from the vehicle itself may offer the capability to engage with the vehicle

owner at the time of the loss and provide |

critical support and can also help inform SERVICE FOR TIME CCC INSIGHT the repair.

Technology within the vehicle provides our industry with new opportunities and challenges. Vehicles involved in an accident are more difficult to repair and may often require a customer to return the vehicle after delivery for additional work. Analytics can help anticipate those and provide insurers and repairers with the needed information to set proper expectations with the customer. AI and photos can help identify whether the vehicle is likely repairable or a total loss, and if likely repairable help carriers recommend the best method of inspection for that customer based on a carrier’s criteria. Data from the vehicle itself may offer the capability to engage with the vehicle owner at the time of the loss and provide critical support and can also help inform the repair. Finding the right balance of technology and human touch on a per-customer basis helps the industry provide their customers with the best possible experience.

107 CRASH COURSE 2020 COURSE CRASH

| TIME FOR EFFICIENCY FOR TIME TIME FOR EFFICIENCY SIMPLIFYING THE COMPLEX

04 Cycle time and customer satisfaction are closely aligned — as repair costs and cycle time increase, customer satisfaction is lower (See Figure 92).

Growing vehicle complexity brings unexpected delays in the repair process. Understanding where and how technology can be used to remove time from the overall days between date of loss and vehicle delivered is key to free up personnel to focus on delivering a quality customer experience.

109 CRASH COURSE 2020 COURSE CRASH

In this section of Crash Course 2020, we discuss ways in which efficiencies are found by simplifying Share of Annual Repairable Appraisal Volume for Losses Between $1K-$4K, CY2010- historically complex steps within the overall vehicle claims and repair processes. CY2019 (FIGURE 93) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

$1,000.01-$2K $2,000.01-$3K $3,000.01-$4K 60.0%

DRP Repairs Q4’18-Q3’19 Repair Satisfaction and Cycle Time by Repair Cost Range

(FIGURE 92) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY |

50.0% EFFICIENCY FOR TIME

LossRpt to Vehicle Out Days Avg Repair Satisfaction % 40.0% 45.0 97.5% 40.0 35.0 30.0% 95.0% 30.0 25.0 20.0% 92.5% 20.0 15.0 10.0% 10.0 90.0%

5.0 0.0% CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 0.0 87.5% $0.01-$500.00 $500.01 to $1K $1,000.01 to $2K $2,000.01 to $3K $3,000.01 to $4K $4,000.01 to $5K $5,000.01-$10K $10,000.01 & Up Total

Biding Their Time Overall Claim Cycle Time from Date of Loss to Date Vehicle Out

Over the last 15 years, repair costs for the industry have become more expensive, but roughly (FIGURE 94) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY

56 percent of non-comprehensive losses annually fall between $1000 and $4000 (see Figure NOTE: INCLUDES ONLY COLLISION CLAIMS WITH REPAIR COST BETWEEN $1K-$4K THAT WERE DRIVEABLE FRONT IMPACT LOSSES. 93). These are historically the claims/repairs with the highest productivity and highest CSI. Understanding the factors behind overall cycle time for these losses can highlight opportunities Loss to Loss Reported Days Avg Loss Reported to Last Assignment Sent Days Avg present across all losses. Analysis of these losses was completed, where the losses were filtered Last Assignment Sent to Assignment Retrieved Days Avg Assignment Retrieved Days Avg to Estimate Start Days Avg further to include only drivable, front impact, collision losses for Q4’18-Q3’19 to arrive at a common Estimate Start to Estimate Sent Days Avg Estimate Sent to Vehicle In Days Avg Vehicle In to Repair Start Days Avg Repair Start to Repair Complete Days Avg set of claim characteristics, and to identify opportunities to simplify the claim and repair process Repair Complete to Vehicle Out Days Avg independent of the repair itself.

The total number of days from date of loss to vehicle delivery date for this subset of losses was on Q4'18-Q3'19 average over 58 days in Q4’18-Q3’19, an improvement from 61.4 days in Q4’14-Q3’15 (see Figure 94). The entire process as we know involves multiple steps and multiple stakeholders — and while repairers and insurers have made improvements in cycle time in certain parts of the process, Q4'14-Q3'15 much of the overall time is still tied up in parts of the process in which the repairer and insurer are dependent on consumers making numerous decisions about what is, for them, an unfamiliar process. 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0

Some examples of how the overall cycle time for this subset of claims was reduced: Potential Cycle Time Improvements With Digital Capabilities Implemented (FIGURE 95) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY • The average days between date of loss and date loss reported fell from 11.7 days in Q4’14-Q3’15 NOTE: INCLUDES ONLY COLLISION CLAIMS WITH REPAIR COST BETWEEN $1K-$4K THAT WERE DRIVEABLE FRONT IMPACT LOSSES. to 10.1 days in Q4’18-Q3’19. As more consumers select UBI auto policies, or buy connected cars, the ability to identify the accident immediately and contact the customer is more prevalent. A comparison of claims where CCC’s telematics-enabled claim solution detected the crash versus Loss to Loss Reported Days Avg Loss Reported to Last Assignment Sent Days Avg

Estimate Sent to Vehicle In Days Avg the non-telematics enable repairable losses conducted by CCC in CY 2019 revealed significant Last Assignment Sent to Estimate Sent Days Avg | Repair Start to Repair Complete Days Avg

reduction in time from date of loss to loss reported for the telematics-enabled claims. The Vehicle In to Repair Start Days Avg EFFICIENCY FOR TIME number of days from loss to loss report was nearly 90 percent lower, for telematics-enabled Repair Complete to Vehicle Out Days Avg claims. As connected car volumes grow, and more OEMs and insurers take advantage of this technology to drive customer engagement, this metric will likely see further reduction in the 70.0 future. 60.0 50.0 • Figure 83 provided a comparison of the time to produce the estimate during the initial 40.0 estimate inspection, a key step in the process needed to help the consumer make the next decision about their vehicles. The average days from last assignment sent to estimate 30.0 complete/sent was 0.6 days for those losses among the $1,000-$4,000 sample that used 20.0 photo MOI versus all others at 5.1 days — a nearly 90 percent reduction. 10.0 0.0 ORIGINAL WITH CYCLE TIME IMPROVEMENTS • Previous analysis of claims data conducted by CCC revealed the overall time between when the insurance company’s customer has an estimate to when they schedule their vehicle in for repair is less for those claims where the consumer has scheduled their appointment online themselves. For consumers equipped with information about the shop and its schedule, the Part of the Solution number of days between when the consumer receives their estimate to the date they schedule their vehicle in for repair was also nearly 11 percent lower. Access to key information on the Historically, the process of identifying the best part, from the best supplier and delivery times, repair shop such as proximity, DRP program participation, on-line reviews, customer referrals, ordering those parts, matching up orders once delivered, keying in invoices, and finally paying or availability of OE certification all within the same platform may be helping consumers suppliers was extremely cumbersome and time consuming. Insurance appraisers, shop estimators, choose their repair shop quicker. And, perhaps more important, our repair experience survey and parts personnel made numerous phone calls, identifying whether catalogued inventory was revealed an easier scheduling process is a key variable driving overall customer satisfaction available, and matching up orders and invoices upon receipt. When a part didn’t work, the part with the repair experience. return process was even more cumbersome. Suppliers such as OEM dealers struggled with high call volumes, and ensuring their current inventory was being accessed by the largest possible audience. These are just several ways we see the industry using technology to improve and simplify the overall customer experience, where the results are meaningful reductions in cycle time and With CCC® Parts, repairers are electronically connected with parts suppliers, whose live inventories, subsequently improvements in overall customer engagement. By providing digital capabilities to pricing and delivery times are displayed while creating the estimate. By moving parts purchase the consumer in key parts of an unfamiliar process in a way that provide clear easy-to-understand upstream to the estimating/repair planner, and to delivery availability and margin visibility, moves information, where they can use their smartphone to review that information and make decisions the point of decision upstream in the repair process. Analysis of repairers utilizing the system shows regarding their claim and repair, the overall cycle time industry-wide could potentially see a 12- improvements in supplement rate and labor hours per shop day and labor hours per repair day day improvement (see Figure 95). As vehicle complexity increases, repair costs and repair cycle from improved efficiencies. Supplier preferences are automatically applied to parts quotes and times are expected to rise further. The ability to integrate technologies into the process to help the orders and, once orders are confirmed, invoices flow back through the system, reducing the need consumer make better, more informed and faster decisions can help reduce the overall time from to manually key invoice information. Removing redundant non-value tasks such as manual purchase loss to vehicle out, potentially providing a significant counterweight to growing repair times and lift orders and invoices frees personnel up to focus on those orders that may require additional focus. to overall customer engagement. Industry analysts project online parts sales reached $12 billion in 2019, and are growing at a rate of about 12 percent annually.147 A survey conducted by JDJS Consulting found that dealers who use online platforms that provide real-time pricing and procurement data report a ‘win’ rate of 55 percent on their quotes compared to 41 percent who stick with fixed prices.148 Finally, with

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 113 CRASH COURSE 2020 COURSE CRASH electronic purchase orders (ePOs), repairers see greater efficiencies, and 15-20 percent fewer When the Vehicle is Not Repairable, How Can returns. Repairers and suppliers are freed up to focus time on more complex orders, quality control, and customer service. Complexity in the Process Be Streamlined?

Today’s repairs include more part replacements than ever before — 12.8 parts were replaced per Another area where new car dealerships have seen their revenues grow is within the F&I segment repair in Q4’18-Q3’19, up 1.3 parts in two years. With replacement parts accounting for a growing of their business. As of CY 2018 dealership F&I penetration for new car sales was 89.6 percent, and

155

share of the overall repair cost, and share of labor time, the ability to identify the most appropriate 73.2 percent for used vehicle sales. F&I income as a percent of new vehicle sales was 2.9 percent |

parts up front, to know that they are available, and to be able to place the order seamlessly can for new-vehicle sales and 3.8 percent for used in CY 2018, up from 2.8 percent and 3.4 percent EFFICIENCY FOR TIME help counter increases in overall cycle time from more complex repairs. in CY 2014.156 With the average price of a new vehicle now over $35,000 — up nearly 30 percent since CY 2005 — more buyers purchase a new or used vehicle today with financing (see Figure As auto dealerships look toward a future with fewer new vehicle sales, they will rely more on other 97). According to Experian, as of Q3 CY 2019, the U.S. had a $1.220 billion open automotive loan areas of their business such as parts and service. In CY 2018, data from NADA reveals U.S. new car balance up from $968 billion in Q3 CY 2015.157 Loan balances have grown as both vehicles sold, and dealership’s total parts sales were $68.61 billion, of which $18.33 billion of which was wholesale.149 amount financed have grown. 31 percent of all new vehicle loans and 30 percent of used vehicle Parts & Service as a percent of the average U.S. new car dealership’s sales was 12.8 percent as of loans had a loan term of 73-84 months, up from 27.5 percent for new and 16.2 percent for used five June CY 2019 — up from 11.4 percent in CY 2000 (see Figure 96).150 In CY 2009 with auto sales years prior (see Figure 98).158 at record lows, parts and service accounted for an even higher share of total sales. According to a survey of 650 franchised dealers conducted by Kerrigan Advisors, sales of service and parts, as well Experian Percentage of New and Used Vehicles with Financing by Quarter as used vehicles, remain the key drivers of profitability. Service and parts sales were up 27 percent CY2010-CY2019 (FIGURE 97) | SOURCE: EXPERIAN “STATE OF AUTOMOTIVE FINANCE MARKET” since 2015 and continue to contribute most to the dealers’ overall gross profits.151 NEW USED NADA Parts & Service Percent of Total Sales — U.S. New Car Dealers CY2000-CY2019 95.0% (FIGURE 96) | SOURCE: NADA DATA 85.0%

18.0% 75.0% 16.0% 65.0% 14.0% 55.0% 12.0% 45.0% 10.0% 35.0% 8.0%

Q1 2011 6.0% Q1 2010Q3 2010 Q3 2011Q1 2012Q3 2012Q1 2013Q3 2013Q1 2014Q3 2014Q1 2015Q3 2015Q1 2016Q3 2016Q1 2017Q3 2017Q1 2018Q3 2018Q1 2019Q3 2019 4.0% 2.0% New and Used Vehicle Financing Distribution of Volume by Loan Term 0.0% (FIGURE 98) | SOURCE: EXPERIAN “STATE OF THE AUTO FINANCING MARKET” Q3 2015 AND Q3 2019

50.0% CY2011 CY2000CY2001CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009CY2010 CY2012CY2013CY2014CY2015CY2016CY2017CY2018 45.0% 40.0% CY2019 to June Finally, companies within the aftermarket and recycler segments are looking to expand their 35.0% markets via digital. According to studies on e-tailing and other trends conducted by the Auto 30.0% Care Association, the e-tailing channel in 2014 was roughly $6 billion or roughly 6 percent of the 25.0% replacement parts market.152 By 2017, the association’s data showed e-tailing had reached $22 20.0% billion, with a CAGR of 15 percent.153 And while that CAGR is expected to slow to 8 percent through 15.0% 10.0% 2020, the association anticipates the industry will continue to change — integrating online and 5.0% offline shopping/purchasing experiences.154 0.0% New Q3 2015 New Q3 2019 Used Q3 2015 Used Q3 2019 49-60 Mos 61-72 mos 73-84 mos

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 115 With loan term lengths extended, more customers end up owing more for their vehicle than it’s worth. According to Edmunds, an estimated 33 percent of individuals who traded in cars to buy new ones in the first nine months of CY 2019 had negative equity, compared to 28 percent five years ago, and 19 percent a decade ago.159 Unfortunately, this also means that when a customer has a total loss vehicle, the likelihood that he/she still has an outstanding vehicle loan increases. Ellen Tierney

Figure 74 illustrated the growth in total loss frequency across all vehicle ages and showed how Vice President, Chief Claims Officer over the last five years the largest percent increase occurred among vehicles ages current year MetLife Auto and Home to four years of age. Many of these vehicles likely still have outstanding loan balances, given the average loan term length as of Q3 2019 per Experian was 69 months. The percent of vehicle claims that result in a total loss has risen over the last several years and is anticipated to grow further in the near future given the older vehicle fleet and higher vehicle repair costs. Given the growing share of consumers with longer loan term length, more total loss claims than ever require engagement of the lien holder, which can result in longer overall cycle time from loss to final settlement and salvage lot closed. Among confirmed total loss claims in CY 2019, the average days from the date What is one thing you/your company of loss to the date salvage lot closed was 73.7 days, with only 11.6 days of those from date of loss to date of last vehicle valuation. spend time thinking about? While heavily damaged vehicles or older model year vehicles may be more easily identifiable as likely total loss claims, an estimated 2-5 percent of claims are borderline losses, where the total loss determination is less obvious. CCC customers who enabled the suite of tools CCC® Quick Estimate Today’s technology in vehicles — our overriding concern -> CCC Smart Total Loss -> CCC® Quick Valuation have seen improvements in their ability to is the cost and reparability of newer cars. The fact that identify borderline total losses that went through the photo estimating channel. In these instances, consumers used Quick Estimate to take photos of their damaged vehicle, and when uploaded, enhanced technology is now not only in high-end vehicles, Smart Total Loss is automatically run. Consumers are then, using Quick Valuation, prompted to but also in more affordable vehicles, brings a whole new confirm options and take additional photos to document things like vehicle condition and mileage. level of complexity and cost to repairing these vehicles. This helps process claims more quickly, and significantly faster than via a physical inspection As vehicle technology continues to evolve, it will challenge (see Figure 99). “ shops’ ability to assess and repair this technology, which will With increased share of claims non-repairable, the ability to more accurately and quickly identify result in needing to take an entirely different approach to whether a loss vehicle is a total loss earlier in the process, and to process the claim through pricing and the claim repair process." settlement will become more important to help counter the growing dissatisfaction among insurance companies’ consumers as more owe more on their vehicle than it may be worth at time of loss.

Total Loss Cycle Time Comparison by Inspection Method CY2019

(FIGURE 99) | SOURCE: CCC INFORMATION SERVICES INC., CCC NATIONAL INDUSTRY Customers using CCC’s Quick Initial Vehicle Estimate -> CCC’s Smart Inspection Method Total Loss -> CCC’s Quick Valuation Total Industry

Loss Report to Last Photo Estimate 12.6 17.3 Valuation Days Avg

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 117 CRASH COURSE 2020 COURSE CRASH This reality runs counter to consumer artificial intelligence (AI) innovations, such as expectations for how an insurance claim would CCC Smart Total Loss™, which assists carrier in be handled, which today means consumer-led making near-instant predictions as to whether and digitally enabled — the same way nearly a vehicle is likely a total loss or repairable using every other transaction is managed in their daily a single photo, which may help inform whether life. And, because a vehicle is often the second the vehicle should be sent to a repair shop or most costly investment a consumer makes, salvage yard. Next, we introduced CCC® Quick

the stakes are high when an unplanned and Valuation to enable a vehicle owner to capture |

disruptive experience, like a vehicle accident, additional photos and input options as requested EFFICIENCY FOR TIME leaves them feeling anxious and confused. by an insurance carrier for the insurer’s total loss claims processing. The combined AI and digital ABOUT THE AUTHOR experience have been shown to result in a 20 Total Losses on Active percent reduction in cycle time from first notice Heidi Ford of loss to settlement. Senior Director, Loans Product Management, According to 2018 U.S. census figures, 93 percent Now, we’re taking the next step in total loss CCC of households reported having access to at transformation to address the complexities least one vehicle.¹ Vehicle ownership is woven of lien release and title paperwork with the into the fabric of our lives but comes at a higher introduction of CCC® Total Loss Care. Total Loss price. Over the last 10 years the average MSRP Care mitigates and simplifies the cumbersome, of a new vehicle has increased from $23,900 manually intensive tasks associated with these to $29,000, and the cost of consumer selected processes for the insurance carriers and auto options increased from $6,500 to $10,000.² Over finance lenders. this 10-year period, vehicle prices have surged 34 ADDRESSING By extending CCC platform capabilities for percent, outpacing the mean wage increase of 4 insurance carriers to new market segments such percent.³ as auto finance lenders, CCC’s solutions can help TOTAL LOSS reduce wait time on needed information for lien Consumers are paying more for their vehicles, and taking longer to pay them off — if they pay-off information from days to seconds. This CLAIM PROCESS are able to do so at all.⁴ Edmunds reports that single platform approach provides visibility for ...a vehicle is often through the end of Q3 2019, 33 percent of insurers and lien holders to their role within the total loss claim life cycle, reducing manual tasks the second most consumers trading in vehicles on a new purchase “ COMPLEXITY had negative equity, compared with 19 percent and providing analytics that unlock actionable costly investment a a decade ago.⁵ On average the consumer owed insights to drive continued improvements. consumer makes, by Heidi Ford $5,000 after trading in their vehicle, before taking on a new loan. Total Loss Care also enables insurance carriers to the stakes are high The total loss process exemplifies the definition of complex: “… simplify and enhance the vehicle owner journey when an unplanned consisting of many different and connected parts.” The numerous Subsequently, more consumers than ever before through provision of proactive digital status steps in the total loss claim life cycle (notification of loss, method of with a total loss claim still have an outstanding updates and process guidance, providing clarity and disruptive inspection, total loss determination, valuation, settlement, salvage, vehicle loan balance. The process of settling the and comfort during a potentially worrisome experience, like a etc.) can extend across multiple participants: insurance carrier, total loss with the lien holder may be manual or event. Powering simplification of insurance claims through adoption of digital innovations…the time vehicle accident, lender, salvage yard, gap insurance provider, and others. Each somewhat automated but often involve other participant has specific requirements, workflows, documentation, companies who have historically been very paper is now. intensive. Acquisition of the vehicle title for title leaves them and customer touch points that can result in inefficiencies, long 1. HTTPS://WWW.VALUEPENGUIN.COM/AUTO-INSURANCE/CAR- feeling anxious and cycle times, and lost productivity. Many of the participants still transfer also involves numerous steps, some of OWNERSHIP-STATISTICS#NATIONAL. have paper-intensive processes that are slowly being digitized. which are automated and support e-signature 2. HTTPS://WWW.CNBC.COM/2019/10/22/CAR-PRICES-ARE-RAPIDLY- INCREASING-HERES-WHY-THATS-BAD-FOR-AMERICANS.HTML. confused." Additionally, given the number of stakeholders involved, the and electronic forms, but most do not. 3. WWW.BLS.ORG. customer and insurance claims adjuster often have to jump between 4. HTTPS://WWW.MARKETWATCH.COM/STORY/MORE-BORROWERS- CCC is applying its 40 years of experience to ARE-GOING-UNDERWATER-ON-CAR-LOANS-2019-11-09-111032935. numerous touchpoints and applications, re-entering the same 5. HTTPS://WWW.MARKETWATCH.COM/STORY/MORE-BORROWERS- information many times. The process can take weeks if not months improve, simplify and streamline the process for ARE-GOING-UNDERWATER-ON-CAR-LOANS-2019-11-09-111032935. to complete and can lead to a less than optimal customer journey. participants. CCC brought to market digital and 119 CRASH COURSE 2020 COURSE CRASH

TIME FOR DATA FOR | TIME TIME FOR DATA INFORMING AND INCITING ACTION

Despite advancements in technology in vehicle 05 design, crash worthiness, crash avoidance, millions of people in the U.S. each year are injured in motor vehicle accidents, and tens of thousands are killed.

The economic costs of providing the necessary medical care for these individuals continues to grow, as inflation in medical cost grows. In this section of Crash Course 2020 we will discuss the state of motor vehicle safety and insurance casualty claims.

121 CRASH COURSE 2020 COURSE CRASH

Motor Vehicle Injuries and Fatalities Remain a USDOT NHTSA U.S. Motor Vehicle Fatality Composition, CY2009 and CY2018 Serious Problem (FIGURE 101) | SOURCE: USDOT NHTSA 2018 FATAL MOTOR VEHICLE CRASHES: OVERVIEW CY2009 CY2018 In CY 2018, 36,560 people were killed in motor vehicle traffic crashes. Early data from NHTSA for the first nine months of 2019 suggest fatalities fell further in CY 2019, with the fatality rate per Pedestrians, Bicyclists and Pedestrians,

100 million vehicle miles traveled dropping to 1.10 versus 1.13 in CY 2018 and a high of 1.17 in CY Other Non - Bicyclists and DATA FOR | TIME 160 Occupants 14% 2016. While this data suggests fatalities may fall another 2.2 percent for CY 2019, after declining , Other Non - 2.4 percent in CY 2018 versus CY 2017, the shift in where those fatalities are occurring remains Occupants 20% Passenger Car Passenger Car Occupants 39% Motorcyclists Occupants 35% a concern. According to the U.S. Census Bureau, the urban population increased by 13 percent 13% between CY 2008 to CY 2017, while the rural population fell by 12 percent; and the urban fatality Motorcyclists 14% rate per 100 million vehicles miles traveled increased by 18 percent between CY 2009 and CY 161 2018 while the rural fatality rate declined by 14 percent (see Figure 100). As more people move Light-Truck Light-Truck Occupants to urban areas, motor vehicle fatalities are increasing there. This shift explains in part why the Large-Truck, Occupants 30% 27% composition of fatalities has shifted to more pedestrians, bicyclists and other non-occupants (see Bus and Other Vehicle Occupants 3% Figure 101). Large-Truck, Bus and Other Vehicle Occupants 4% Unfortunately, while fatalities fell 0.9 percent between CY 2016 and CY 2017, the National Safety Council reports medically consulted injuries resulting from motor vehicle crashes rose from 4.3 million in CY 2016 to 4.6 million in CY 2017, with an estimated cost of $433.8 billion.162 This figure includes wage and productivity losses, medical expenses, administrative expenses, motor-vehicle National Safety Council Share per Type of Crash for Fatal and Injury Crashes property damage, and employer costs. And while 41 percent of motor vehicle fatalities occur when (FIGURE 102) | SOURCE: NATIONAL SAFETY COUNCIL. INJURY FACTS® 2017 EDITION, PP. 106-107 the accident type was a collision between motor vehicles, 77 percent of medically consulted injuries come from collisions between motor vehicles (see Figure 102). 100%

90%

80% USDOT NHTSA U.S. Motor Vehicle Fatalities by Land Use, CY2009-CY2018 70% (FIGURE 100) | SOURCE: USDOT NHTSA 2018 FATAL MOTOR VEHICLE CRASHES: OVERVIEW 60%

50% Rural Urban 40,000 40%

35,000 30%

30,000 20%

25,000 10%

20,000 0% Deaths (NSC estimated 37,757 occurring within 30 Medically consulted injuries (4.3 million) 15,000 days of accident)

10,000 Collision between motor vehicles Collision with ﬁxed object Collision with pedestrian Noncollisions (ex. overturn, jackknife) 5,000 Collision with pedalcycle Collision with railroad train Other collision (ex. Animal hit) 0 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018

The good news is that more recent data from NHTSA suggests injury rates per million vehicle miles percent-100 percent received an “A” grade; “B” = 60 percent-69 percent; “C” = 50 percent-595; “D” traveled may be trending down. All three measures — ‘all injury crashes,’ ‘vehicles involved in all = 40 percent-49 percent; and “F” = below 40 percent. The range in scores (a low of 11 percent for injury crashes,’ and ‘persons injured in all crashes’ per one hundred million vehicle miles traveled Montana to a high of 78 percent for Illinois) as well as other important demographic differences saw a slight decline in CY 2017 and CY 2018 (see Figure 103). Unfortunately, due to NHTSA’s help explain why the outcome of motor vehicle accidents can be so different by state — particularly replacement of the reporting system used to the Crash Report Sampling System (CRSS) in CY 2016, as it relates to drivers and passengers. comparison to earlier years is not straightforward.

National Safety Council’s Risk Factors for Fatal Injuries on the Road DATA FOR | TIME Crash Rate per 100 Million Vehicle Miles Traveled by All Motor Vehicles, CY1997- (FIG 104) | SOURCE: NATIONAL SAFETY COUNCIL. THE STATE OF SAFETY: A STATE-BY-STATE REPORT. WWW.NSC.ORG/STATEOFSAFETY. CY2018 (FIGURE 103) | SOURCE: ANALYTICS DIVISION, FEDERAL MOTOR CARRIER SAFETY ADMINISTRATION, LARGE TRUCK AND BUS Vulnerable Road Users 6 33% CRASH FACTS 2017, FMCSA-RRA-18-108, TRENDS TABLE 9. Alcohol 29% All Injury Crashes Vehicles Involved in All Injury Crashes Persons Injured in All Crashes No Seat Belt 4 28% 180.0 Speeding 27% Fatigue 17% 160.0 Teen Drivers 5 13% 140.0 Distraction 2 10% Older Drivers 3 5% 120.0 Child Passengers (Ages 14 and younger) 1 2% 100.0 1) Despite laws in all 50 states for restraining infants and children in vehicles, over 1/3 of children killed in trac crashes are not buckled up. 80.0 2) Currently, no state has a law that completely bans all electronic device use, including hands free, behind the wheel. NSC believes a full ban is the best way to prevent crashes involving distracted drivers. 60.0 3) 1 in 5 drivers on the today today are ages 75 and older, and although older drivers comprise only 7% of two-car crashes, they are more likely to be seriously injure or killed than any other age group. 40.0 4) In fatal crashes, 50% of fatalities occur among people who are unbuckled.

20.0 5) Teen drivers are 3 times as likely to be involved in a crash at night versus during the day. Just a single young passenger can increase a teen’s crash risk by 44%. Unfortunately, most states have age or driver education loopholes that permit these safety 0.0 measures to be circumvented. 6) Vulnerable road users include motorcyclists, bicyclists, and pedestrians.

CY2011 CY1997CY1998CY1999CY2000CY2001CY2002CY2003CY2004CY2005CY2006CY2007CY2008CY2009CY2010 CY2012CY2013CY2014CY2015CY2016*CY2017*CY2018* National Safety Council Road Safety Score (FIGURE 105) | SOURCE: NATIONAL SAFETY COUNCIL. THE STATE OF SAFETY: A STATE-BY-STATE REPORT. WWW.NSC.ORG/STATEOFSAFETY.

A large majority of accidents each year are caused by the driver. Driving under the influence of drugs or alcohol, driving fatigued, speeding, distracted driving, and other factors continue to play a role. According to a study from the U.S. Centers for Disease Control and Prevention, in 2018 an estimated 12 million drivers (4.7 percent) aged 16 and older drove while stoned (under the influence of marijuana), and 2.3 million said they drove after using other illicit drugs.163 21 million drove drunk during the same period.164 IIHS has completed numerous studies revealing crash rates have risen in states that legalized retail sales for recreational use of marijuana.165

The National Safety Council’s State of Safety report provides a state-by-state score of road safety, looking at key issues and factors that impact it. Nine key issues were examined, and in total were assigned an overall risk factor contribution for fatal injuries on the road (see Figure 104) (numbers do not add up to 100 percent due to overlap).166 Figure 105 shows the individual weighted score per state based on policy and legislation pertaining to 24 different indicators across eight safety issues: alcohol impaired driving, child passenger safety, distracted driving, older drivers, seat belt use, speeding, teen drivers, and vulnerable road users.167 States with a score between 70

As more vehicles are equipped with ADAS, early data suggests we will start to see a reduction in A comparison of the Bureau of Labor’s CPI data reveals the medical care index since CY 2001 auto crash rates, and resultant injuries and fatalities, because the vehicle speed and subsequent show some of the fastest growth in the last decade occurred in CY 2016, and costs for CY 2019 are impact are slowed by ADAS features, even if the accident is not avoided altogether. For example, increasing faster than overall CPI (see Figure 107). The indices for hospital services, prescription IIHS/HLDI studies of the efficacy of various ADAS features in reducing claim frequency and costs drugs, and services by other medical professionals have seen larger increases over the last several found FCW without autobrake reduced bodily injury claim frequency by 15 percent; adding AEB years than overall inflation (see Figure 108). pushed the reduction up to 19 percent.168 Most of the benefits of these safety systems come to

vehicle occupants. Unfortunately, tests conducted by IIHS and AAA indicate that many of the Annual Percent Change, CPI — Medical DATA FOR | TIME pedestrian crash prevention systems are lacking, particularly when a vehicle is taking a right-hand (FIGURE 107) | SOURCE: INSURANCE INFORMATION INSTITUTE, BLS CPI turn, driving at night, or driving at speeds of 30 mph.169 In order to promote development of better 4.6% 4.7% pedestrian crash prevention, IIHS has announced tougher new requirements for its 2020 Top Safety 4.4% 4.4% 4.2% Pick and Top Safety Pick+ designations. Automakers must achieve good or acceptable ratings for 4.0% 4.0% 3.7% 3.7% 3.8% the IIHS tests of pedestrian crash prevention system test, headlight evaluations, and passenger-side 3.4% 3.2% small overlap crash worthiness in order to achieve the designations.170 3.0% 2.8% 2.6% 2.5% 2.4% 2.5% Improved vehicle safety and availability of ADAS may help to reduce fatalities and injuries of vehicle 2.0% occupants, but until features such as pedestrian airbags are standard and pedestrian detection systems improve, the non-occupants’ share of motor vehicle crashes may continue to trend higher. State of the Union — U.S. Medical Costs Continue to Rise

U.S. spending on health care rose to $3.65 trillion in CY 2018, rising above $11,000 per person for CY2011 CY2017 CY2012 CY2015 CY2013 CY2018 CY2016 CY2019 CY2014 CY2010 CY2001 CY2007 CY2002 CY2005 CY2003 CY2008 CY2006 CY2009 the first time (see Figure 106).171 Growth in per capita spending increased 4 percent in CY 2018, CY2004 up from 3.5 percent in CY 2017, as faster growth in medical prices more than offset slower growth in the use and intensity of health care goods and services.172 Higher medical prices accounted for 2.1 percent of the overall 4 percent per capita increase in CY 2018; growth in the residual use and Healthcare Costs Continue to Rise Faster Than Inflation — BLS CPI Price Level intensity of health care goods and services accounted for 1.3 percentage points, and the changing Change Annual CPI Percent Change from Prior year — U.S. city average, all urban age and sex mix of the population accounted for 0.6 percentage points.173 Overall, spending rose customers, not seasonally adjusted 4.6 percent in CY 2018 versus 4.2 percent in CY 2017 even as the number of Americans without (FIGURE 108) | SOURCE: WWW.BLS.GOV health coverage rose to 30.7 million individuals, up 1 million for the second consecutive year.174 A CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 slight increase in generic drugs helped drop the price of retail prescription drug prices by 1 percent — the first drop since CY 1973, but total spending on prescription drugs rose 2.5 percent to $335.175 16.0% 14.0% U.S. Health Care Spending in CY2018 12.0% (FIGURE 106) | SOURCE: HEALTH AFFAIRS 39, NO. 1 (2020) 10.0% CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 8.0% National Health Expenditures, in billions $2,791.1 $2,875.0 $3,025.4 $3,199.6 $3,347.4 $3,487.3 $3,649.4 6.0% Annual Growth 4.0% 3.0% 5.2% 5.8% 4.6% 4.2% 4.6% 4.0% U.S. Census Bureau U.S. Population 313.3 315.5 317.9 320.1 322.5 324.6 326.6 (in millions) 2.0% Annual Growth 0.7% 0.7% 0.7% 0.7% 0.8% 0.7% 0.6% 0.0% U.S. GDP per capita $51,695 $53,200 $55,143 $56,932 $58,025 $60,128 $63,004 -2.0% Annual Growth 3.4% 2.9% 3.7% 3.2% 1.9% 3.6% 4.8% All All Items Medical Physicians' Services by Hospital Prescription Medical care Hospital and Health Items Less Food Care Services Other Services Drugs services related Insurance -4.0% NHE per capita $8,908 $9,113 $9,518 $9,995 $10,379 $10,742 $11,172 and Energy Medical services Professionals -6.0% Annual Growth 3.3% 2.3% 4.5% 5.0% 3.8% 3.5% 4.0%

State of the Union... is continued on page 132

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 127 CRASH COURSE 2020 COURSE CRASH 1. Prevent auto accidents; or 2. Increasing the severity of an impact, also in terms of impact force, increases the 2. Protect occupants from injury when in an auto accident; or risk of injury to the vehicle’s occupants. 3. Facilitate better medical care to an injured occupant after an accident. Please refer to the Physics Insight breakout section in this article for more information about the relationship of accident forces and Today, vehicle manufacturers offer a accident outcomes. DATA FOR | TIME variety of technologies within each of these categories. For example, Automated Driver Assistance Systems (ADAS) such as Forward If we can measure acceleration in a collision, Collision Warning, Automatic Emergency we can more precisely determine the collision ABOUT THE AUTHOR Braking, and Blind Spot Detection are force or severity and gain better insights Scott Palmer designed to prevent accidents. Occupant into auto damage and the potential for Vice President, Product protection systems such as airbags, seat occupant injury. A few auto manufacturers now offer what is commonly referred to as Management, CCC belts, and collapsible steering columns are designed to prevent injury when an Advanced Automated Collision Notification accident occurs. Lastly, today’s telematics Systems (AACN) in the U.S. These systems systems connect the car to telematics use data from on-board sensors that measure service providers who can use data from collision accelerations to detect accidents the vehicle to improve emergency medical and other on-board data elements. These responses to the accident scene. Virtually data elements are transmitted via telematics all of these technologies benefit from and AI is used to predict the severity of injury robust vehicle data capture and artificial outcomes in near real time. VEHICLE DATA intelligence. Interestingly, data captured for accident prevention and occupant protection Over a decade ago, the Centers for Disease technologies can be used to improve medical Control and Prevention (CDC) published OFFERS THE NEXT care to injured parties after an accident. We forward looking recommendations from an will explore the use of this data and how it expert panel on AACN and the triage of the can also benefit casualty claim processes. injured patient. The expert panel found that If we can measure FRONTIER TO AACN systems showed promise in improving outcomes in severely injured crash patients. acceleration in a “ MITIGATE ACCIDENT Impact of Telematics Using AACN, severe, life threatening injuries collision, we can more Systems and Vehicle can be predicted almost immediately from precisely determine vehicle data using telematics and AI. Armed INJURIES AND Data on Injury with these injury predictions, first responders the collision force Outcomes with the appropriate level of training can or severity and gain be deployed to the accident scene, the IMPROVE CASUALTY location of which is also provided by the better insights into Let’s start with a review of some important vehicle. Additionally, when severe injuries auto damage and the background to understand why and how are predicted and are not easily observable CLAIM RESPONSES vehicle data is important for understanding by first responders (e.g., internal injuries), potential for occupant injury outcomes. The professional disciplines injured occupants should be directed to Level injury." by Scott Palmer of accident reconstruction and injury 1 trauma centers rather than local hospital causation have clearly established axioms. All emergency rooms. A study published in the Logically, for automakers there are three types of technology other factors being equal, the following are New England Journal of Medicine showed strategies to reduce serious injury and related consequences in true: that if someone is severely injured in an auto accidents. These three strategies are designed to either: 1. Increasing the severity of an impact, accident and they are transported to a Level in terms of impact force, increases the 1 trauma center, their chances for survival amount of damage to a vehicle, and increase by 25 percent.¹ 129 CRASH COURSE 2020 COURSE CRASH Continued 3. Are there any injury claims that are Add the earlier stated benefits of improving inconsistent with the vehicle data? accident scene responses and saving lives, valuable vehicle data such as acceleration 4. Is the developing treatment path and seat belt use promise to drive significant VEHICLE DATA OFFERS THE consistent with the expected injury? changes in the accident response and

casualty claims landscape. Today’s AI Armed with these insights, insurers can initiatives at CCC aim to leverage real DATA FOR | TIME NEXT FRONTIER... accurately segment casualty claims based time vehicle crash data to inform process on their complexity and straight through improvements for casualty claims. And, these process the claims that experience shows will improvements will be about time. The expert participating in the and likely more accessible from the vehicle not benefit from increased human touches. aforementioned CDC study identified a The end result? Significant time savings, 1 MACKENZIE, E., ET AL. “A NATIONAL EVALUATION OF THE through telematics. EFFECT OF TRUAMA-CENTER CARE ON MORTALITY.” THE NEW variety of vehicle-based data points that elimination of related process expense, and ENGLAND JOURNAL OF MEDICINE 2006 JAN 26;354(4):366-78.

would be helpful in predicting serious injury. improved customer service. 2 It should also be noted that an initial AI NHTSA. 49 C.F.R. § 563 Two helpful data elements are not surprising: model designed to predict severe injury 1) acceleration, and 2) seat belt use. based on vehicle data was published over 20 years ago. And in 2015, an AACN system Interestingly, NHTSA requires that all started operation in Japan. The point of passenger vehicles and pick-up trucks and these observations is that the new frontier of other lightweight vehicles manufactured injury prediction from vehicle data is a more on or after September 1, 2013 capture and advanced capability than most appreciate. record the following information: longitudinal Physics Insight — Breaking Down Accident acceleration data during a collision event that exceeds a 5 mph speed change in What are the Force and Its Relationship to Outcomes about 1/7th of a second and driver seat belt usage immediately preceding the collision. Implications to Auto As we may remember from our high school physics class, force is equal to the mass of an Additionally, under certain circumstances, Casualty Claims? object times its acceleration. Physics also tells us that the mass of an object can be derived NHTSA requires the capture of lateral from its weight and the gravitational pull of the planet. So, in an auto collision, the weight acceleration data during a collision event, of the vehicle, and accordingly, its mass can be easily estimated (published curb weight can It stands to reason that if real time vehicle and front-seat passenger seat belt usage be adjusted for varying fuel levels, occupants and cargo). However, acceleration in an auto crash data is predictive of injury outcomes immediately preceding the collision.² collision, is the more elusive variable which must be measured or estimated to determine the for early responders to accident scenes, it Interestingly, the capture of this type of impact force of any collision. Acceleration is defined as the change in velocity of an object should also be predictive of the nature and vehicle crash data began in the early 1990’s with respect to time. Both acceleration and velocity are vector quantities which means both extent of ensuing casualty claims. and one of the likely purposes for the use of must be defined in terms of magnitude and direction. this data was to evaluate the performance of occupant protection systems. For insurers, what are the potential benefits Let’s examine a couple of examples of these relationships. A high speed (the magnitude), of a detailed understanding of expected frontal (the direction) collision that causes the accident vehicle to completely stop very injuries immediately after an accident when The major takeaways here are: 1) auto quickly (e.g., within 1/7th of a second — which is the time element) would be classified as a vehicle crash data is available? For starters, manufacturers already have sensors and high severity (and high acceleration) frontal impact and likely cause significant auto damage the following casualty claim questions can be other technologies in place that capture and serious injuries to occupants. Conversely, a low speed frontal collision involving a vehicle better informed: the important data elements identified of the same weight that causes the accident vehicle to completely stop over the same time by the CDC’s expert panel to predict 1. What are the likely diagnoses and frame would be described as a low severity (and low acceleration) frontal impact, causing injury outcomes, and 2) as vehicle safety treatments? much less auto damage and far less serious injuries, if any, to occupants. technologies become more advanced, these 2. Will the injury claim costs likely exceed data elements will become more utilized policy limits?

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 131 CRASH COURSE 2020 COURSE CRASH Private Passenger Auto - PIP State of the Union... continued Frequency Severity Frequency for bodily injury (BI) coverage has continued its downward trend, declining -2.3 percent 12.0% for the four quarters ending Q3 2019 (see Figure 109), while paid claim frequency for personal injury protection (PIP) coverage has continued to increase since early CY 2018 (Figure 110). 10.0% Annual rates of change in BI claim severity had risen steadily between CY 2014 and CY 2017, then 8.0%

tapered in CY 2018, but have since accelerated again. BI claim severity was up 6.4 percent for the four quarters ending Q3 2019 versus the same period a year prior. PIP severity also increased but DATA FOR | TIME 6.0% only 1.8 percent during the same period. Unfortunately, increases in claim severity have offset 176 declines in claim frequency, resulting in higher claim costs per insured vehicle. In fact, data from 4.0% the Insurance Research Council shows insurer payments for BI and PIP claims between CY 2008 and CY 2017 grew 31 and 26 percent respectively, despite the seriousness of claim-related injuries 2.0% appears to be decreasing.177 0.0%

Historically, most of the trends in PIP were driven by three of the largest states with no-fault -2.0% approaches to compensating auto injuries — Florida, Michigan, and New York. Michigan finally passed a major no-fault reform law in 2019 that goes into effect July 1, 2020. For the first time in -4.0% nearly 50 years, residents of Michigan will no longer be required to purchase unlimited, lifetime -6.0% personal injury protection.178 -8.0% Percent Change in Claim Frequency & Severity from Quarter One Year Prior (each quarter represents rolling 4 quarters data ending that quarter) 2015-Q1 2017-Q1 2018-Q1 2013-Q1 2016-Q1 2019-Q1 2014-Q1 2015-Q2 2015-Q3 2017-Q2 2017-Q3 2018-Q2 2013-Q2 2013-Q3 2015-Q4 2016-Q2 2016-Q3 2016-Q4 2017-Q4 2018-Q3 2018-Q4 2019-Q2 2019-Q3 2019-Q4 2012-Q4 2013-Q4 2014-Q2 2014-Q3 2014-Q4 (FIGURES 109-110) | SOURCE: ISS FAST TRACK PLUS™ PERSONAL AUTO, AS OF SEPTEMBER 30, 2019 Among the provisions of the reform law are those that enable the use of managed care networks Private Passenger Auto - Bodily Injury for treating auto accident injuries; allow utilization review to ensure the level and quality of care 10.0% Frequency Severity is appropriate; provide new restrictions on attorney fees; codify the Anti-Fraud Unit; and make changes to the Assigned Claims Plan that covers accidents involving uninsured motorists.179 The hope is the reform will help improve the system, reduce overall costs, and reduce the average auto 8.0% premium for residents of Michigan, where the average annual premium was nearly 80 percent higher than the national average. 6.0% Analysis of bodily injury claims data reveals very little change in the highest ranked diagnoses in 4.0% terms of total dollars billed/claimed per diagnosis (either standalone or in combination with other diagnoses) over the last several years. Data from NHTSA continues to show rear-end collisions are 2.0% the most common type of accident, accounting for nearly one-third of all motor vehicle accidents, so it is not surprising that neck injuries and treatment plans including chiropractic care continue to 0.0% top the list. In fact, among third-party casualty claims referred for causation analysis, low-impact crashes (change in velocity of 10 MPH or less) as a percent of all crashes had remained relatively stable over the last several years at 72 percent but jumped to 79 percent in CY 2019 (see Figure -2.0% 111). There are multiple transportation trends which may be influencing the 2019 increase in low- velocity impacts, including increasing presence of ADAS systems across U.S. vehicles and shifts in -4.0% urban versus rural driving patterns.

-6.0% According to the Insurance Research Council’s Countrywide Patterns in Auto Injury Insurance Claims: 2018 Edition, attorney involvement in auto injury claims also continues to climb (see Figure 112).180 Insurance Research Council research has shown claims with attorney involvement tend to 2017-Q1 2018-Q1 2019-Q1 2013-Q1 2014-Q1 2015-Q1 2016-Q1 2017-Q2 2017-Q3 2017-Q4 2018-Q2 2018-Q3 2019-Q2 2019-Q3 2012-Q4 2013-Q2 2013-Q3 2013-Q4 2014-Q2 2014-Q3 2015-Q2 2015-Q3 2015-Q4 2016-Q2 2016-Q3 2018-Q4 2019-Q4 2014-Q4 2016-Q4 have higher utilization rates for chiropractic treatment, physical therapy and expensive diagnostic

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 133 CRASH COURSE 2020 COURSE CRASH procedures such as MRIs and CTs, even when comparing claimants with similar injuries.181 Growing Medical inflation is in large part being driven by a number of key variables. use of ‘lien-doctors’ (typically referred to injured by plaintiffs’ attorneys) that agree to provide medical treatment for the injured and get paid after the litigation is resolved is also believed to be a factor driving up litigation costs overall as well. In fact, there is growing belief that increases in casualty claims severity exceeding the CPI medical inflation rate can be attributed to increasing State Higher Safety Diagnostic 183 Dierences Speeds Systems Intensity attorney representation fees. TIME FOR DATA FOR | TIME Low Impact (change in velocity of 10 MPH or less) as a Percent of All Crashes for Third Party Casualty Claims Referred for Causation Analysis, CY2012- CY2019 (FIGURE 111) | SOURCE: CCC INFORMATION SERVICES INC., AIS

US Age & CLAIM Age & Treatment Healthcare Driving Clinical Complexity Insurers SEVERITY Care 79% 75% 74% 71% 72% 72% 72% 72%

Evidence- Alcohol Opioid Unsafe Based Marijuana Epidemic Behavior Medicine

To understand some of the key factors driving the above trends in auto BI and PIP claim frequency and loss costs, Auto Injury Solutions, Inc. (AIS), a CCC company, completed an analysis of over 4 million auto BI and PIP /Medpay claims processed through CCC’s casualty products for the period CY 2015-CY 2019. The results of this analysis are outlined below. CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 CY2018 CY2019 Bodily-Injury Claims Review Attorney Involvement in Auto Injury Claims (FIGURE 112) | SOURCE: INSURANCE RESEARCH COUNCIL, “COUNTRYWIDE PATTERNS IN AUTO INJURY INSURANCE CLAIMS: 2018 EDITION. For the period of study (CY 2015 to CY 2019), the top diagnoses for bodily injury claims in terms of overall dollars charged have remained consistent, with neck pain (Cervicalgia) and neck sprain and CY2017 CY2012 CY2007 strain among the top one or two positions in the last four years (see Figure 113). Subsequently, there were only moderate changes in both the procedure utilization by category (see Figure 114), and the medical procedures billed for treatment of bodily injury claims 39% (see Figure 115), underscoring the consistency in the types of injuries, diagnoses, and treatment over that period. Worth noting is a comparison of the top diagnoses based on dollars PIP Claims 36% billed versus the top procedures based on dollars billed underscore the fact that the highest 32% frequency diagnoses are not directly related to the highest cost procedures. In other words, the majority of injuries from high frequency lower speed accidents involve treatments with lower overall costs, while lower frequency high-speed accidents involve treatments with higher overall costs. A noteworthy change in ranking procedures based on dollars-billed is the increase in the 52% ranking for emergency department visits and the appearance of “CT head/brain w/o contrast material” among the top 10 procedures billed in the last five years. There is some debate that BI Claims 50% greater public awareness of traumatic brain injuries among athletes who have sustained numerous 49% concussive head injuries has led to more individuals injured in vehicle crashes concerned about similar injuries, subsequently leading to more claims with related medical procedures. According to data from the Centers for Disease Control and Prevention, TBI-related emergency visits, hospitalizations and deaths increased 53 percent between CY 2007 and CY 2014.184

BI Claims: Top Diagnoses Raking based on Total Dollars Billed for Period BI Claims: Top Medical Treatment Procedures Ranked Based on Total Dollars Billed CY2015-CY2019 (FIGURE 113) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY for Period CY2015-CY2019 (FIGURE 115) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY

CY2019 CY2018 CY2017 CY2016 CY2015

CY2019 CY2018 CY2017 CY2016 CY2015 Therapeutic pc 1/> areas each 15 min exercise 1 1 1 1 1

Cervicalgia 1 1 1 2 4 Operating room/other 2 2 2 2 2

DATA FOR | TIME Low back pain 2 2 2 5 5 Emergency department visit high/urgent severity 3 3 3 5 5

Sprain lig cerv spine initial enc 3 3 3 3 Ct cervical spine w/o contrast material 4 5 5 7 6

Strn musc fasc tendon neck levl int 5 6 6 1 1 mri spinal canal cervical w/o contrast matrl 5 6 7 8 7

Sprain ligaments t-spine initial 6 5 5 10 3 Mri spinal canal lumbar w/o contrast matrl 6 8 8 9 8

Radiculopathy cervical region 7 7 8 Manual therapy tqs 1/> regions each 15 min 7 4 4 3 3

Pain in thoracic spine 8 8 10 8 Chiropractic manipulative tx spinal 3-4 regions 8 7 6 6 4

Radiculopathy lumbar region 9 9 ct head/brain w/o contrast material 9 9 9 10

Headache 10 10 9 7 Ct abdomen & pelvis w/contrast material 10

Physical Medicine & Rehabilitation Procedures

Radiology Services Procedures

Facilities Procedures

Other

Procedures in Third Party Auto Casualty CY2015–CY2019 All Closed Claims (FIGURE 114) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY

NOTE: RADIOLOGY PROCEDURES INCLUDE CTS, MRIS AND OTHER DIAGNOSTIC XRAY STUDIES. Overall, however, when reviewing summary level procedure data, we see just how little change CY2015 CY2016 CY2017 CY2018 CY2019 there has been over the last five years in ranking of the most common procedures based total 25% amount paid (see Figure 116), with steady growth in average fees per procedure for many of the most common (see Figure 117). 20% 15% BI Claims Top Medical Summary Procedures Based on Total Dollars Paid 10% CY2015-CY2019 (FIGURE 116) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY 5% Procedure Code 0% Summary Category CY2019 Rank CY2018 Rank CY 2017 Rank CY2016 Rank CY2015 Rank

Radiology 1 1 1 1 1 Surgical 2 2 2 2 3 Physical Medicine/Chiro 3 3 3 3 2 %Ambulance Evaluation & Management 4 4 4 4 5 % E M Procedures %Pharmacy Drugs Other 5 5 5 5 4 % Surgical Procedures % Path Lab Procedures %Radiology Procedures % Psychiatry Procedures Durable Medical 6 6 6 6 6 % Physical Medicine Chiro… % Neurology Procedures % DME Supplies Procedures Equipment/Supplies Inpatient 7 7 7 7 7 Pathology/Lab 8 8 9 8 9 Pharmacy & Drugs 9 9 8 9 8 Ambulance 10 10 10 10 10 Neurology 11 11 11 11 11 Psychiatry 12 12 12 12 12 Dental 13 13 13 13 13

Average Billed Amount per Procedure for Common Third Party Procedures So why are third-party medical costs increasing? The newer drivers of rising cost are twofold: a) the same types of injuries are being treated with more expensive treatment approaches (e.g., CY2015-CY2019 (FIGURE 117) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY emergency room and hospital related treatments); and b) cost-increases in these more expensive CY2015 CY2016 CY2017 CY2018 CY2019 care modalities are increasing at a much faster rate than traditional modalities such as chiropractic and physical therapy related procedures. Spinal Injections Lumbar/Sacral

Spinal Injections Cervical/Thoracic TIME FOR DATA FOR | TIME Arthroscopy PIP/Medpay Claims Review Anesthesia X-Ray PIP/Medpay claims analyzed from CY 2015 to CY 2019 reveal very little change in the diagnoses Ultrasound of injuries: neck pain (Cervicalgia) or neck sprain and strain held the top spots in four out of the Magnetic Resonance Imaging last five years, while low back pain or lumbar sprain and strain was typically in second place (see Computerized Tomography Figure 118). Medical procedures for PIP/Medpay claims — both in terms of frequency (see Figure Psychiatry 119) and when ranked in terms of dollars billed (see Figure 120) — saw greater use of emergency Passive Physical Medicine/Chiro room and neurology procedures. Active Physical Medicine/Chiro Opioid Nonsteroidal Anti-inﬂammatory Anxiolytics/Sedatives PIP/Medpay Claims: Highest Ranked Diagnoses in Terms of Dollars Billed/Claimed for Drug Testing Period CY2015-CY2019 (FIGURE 118) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY Electromyography/Nerve Conduction Inpatient Procedure Codes DIAGNOSIS RANKING BASED ON DOLLARS CHARGED Moderate Level E&M Low Level E&M CY2019 CY2018 CY2017 CY2016 CY2015 High Level E&M Cervicalgia 1 1 2 4 2 Durable Medical Equipment/Supplies 2 2 1 3 Dental Sprain lig cerv spine initial enc Non Emergency Basic Life Support Strn musc fasc tendon neck levl int 3 3 4 6 4 Non Emergency Advanced Life Support Low back pain 4 4 6 9 3 Emergency Basic Life Support Emergency Advanced Life Support Radiculopathy cervical region 5 5 9 Ambulance Supplies Neck sprain and strain 6 9 5 1 1 Ambulance Air Lift Car occ injured head initial enc 7 7 8 $0 $500 $1,000 $1,500 $2,000 $2,500 $3,000 $3,500 $4,000 $4,500 $5,000 Radiculopathy lumbar region 8 8 7

What has seen significant change is the average number of days between the first treatment date Oth iv disc displacement lumbar rgn 9 9 5 of service to the last treatment date of service, which grew from 140 days in CY 2015 to 211 days in Ct abdomen & pelvis w/contrast material 10 10 CY 2019, while the median number of days grew from 85 days to 91 days over the same period.

Historically, increases in utilization were a principal cause for the rising costs in medical dollars for BI claims. However, for the observation period from CY 2015 to CY 2019, the number of unique procedures has essentially been flat (21 across all five years), as has the average number of procedures: 175 in CY 2015, 172 in CY 2016, 168 in CY 2017, 172 in CY 2018 and 174 in CY 2019.

This also underscores a trend to monitor — ADAS may be helping reduce the frequency of low speed Procedures in PIP/Medpay Auto Casualty CY2015–CY2019 All Closed Claims crashes with less severe injuries, leaving carriers with more crashes with higher velocity collisions (FIGURE 119) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY where injuries are more severe. Additionally, rising costs are coming with more costly procedures, CY2015 CY2016 CY2017 CY2018 CY2019 higher costs per procedure and rising number of visits. Figure 121 shows how the average number 30% of unique procedures for first-party claims was unchanged at 10 procedures throughout the period DATA FOR | TIME 25% analyzed, but the average number of procedures grew from 98 to 119 between CY 2015 and CY 2019. 20% 15% Finally, a comparison of the average charge-per-claim and the average charge-per-claim excluding 10% duplicates reveals the increase in medical inflation, where the same types of treatments and 5% procedures are simply being billed at higher rates (see Figure 122). 0% First Party Casualty — Average Total Number of Procedures Increasing While Average Number of Unique Procedures Remains Flat All Closed Claims CY2015-CY2019 %Ambulance (FIGURE 121) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY % E M Procedures %Pharmacy Drugs

% Surgical Procedures % Path Lab Procedures %Radiology Procedures % Psychiatry Procedures% Neurology Procedures 119 % Physical Medicine Chiro… % DME Supplies Procedures 105 108 98 101

PIP/Medpay Claims: Top Medical Procedures in Terms of Dollars Billed/Claimed for

Period CY2015-CY2019 (FIGURE 120) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY

DIAGNOSIS RANKING BASED ON DOLLARS CHARGED 10 10 10 10 10 CY2019 CY2018 CY2017 CY2016 CY2015

Therapeutic pc 1/> areas each 15 min exercise 1 1 1 1 1 CY2015 CY2016 CY2017 CY2018 CY2019

Emergency department visit high/urgent severity 2 2 2 2 6 Avg # of unique Procedures Avg # of Procedures Manual therapy tqs 1/> regions each 15 min 3 4 4 4 3

Ct cervical spine w/o contrast material 4 3 3 3 2 PIP/Medpay Increasing Average Charges per Claim CY2015-CY2019

Chiropractic manipulative tx spinal 3-4 regions 5 5 5 5 5 (FIGURE 122) | SOURCE: AUTO INJURY SOLUTIONS (AIS), A CCC COMPANY

ct head/brain w/o contrast material 6 6 6 6 4 Avg Charge per Claim Avg Charge per Claim (excl. dups) Mri spinal canal cervical w/o contrast matrl 7 7 8 8 9 $20,000 Mri spinal canal lumbar w/o contrast matrl 8 8 9 9 10 $18,000 Operating room/other 9 9 7 7 8 $16,000 Emergency dept visit high severity&threat funcj 10 10 10 $14,000 $12,000 Physical Medicine & Rehabilitation Procedures $10,000 $8,000 Radiology Services Procedures $6,000 Facilities Procedures $4,000 $2,000 $0 CY2015 CY2016 CY2017 CY2018 CY2019

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 141 CRASH COURSE 2020 COURSE CRASH and no motor on the bed. What comes to Across the U.S., the National Highway mind? Maybe it’s a thought, like “Where’s my Transportation Safety Administration car?” Or, possibly, “Am I going to make it to documented fewer single and multivehicle work tomorrow?” collisions occurring at less than 30 mph in police-reported crashes, with about 76 As the experience drags on, clinicians percent of all these crashes happening at eventually explain their findings and 30 mph or greater. More than 86 percent of

recommendations. A nurse comes with injury crashes were reported at 30 mph or DATA FOR | TIME paperwork. Finally, return home is near. In greater. the days to follow, an auto repair facility will address vehicle damage. But it may As Scott Palmer notes in this publication, ABOUT THE AUTHOR take weeks to feel better — physically and there is a known relationship between mentally — or not. Perhaps an outpatient change in velocity, or Delta V, at collision Mary Capelli- chiropractic visit will become necessary, or impact and injury potential. Consequently, Schellpfeffer, MD, a physical therapy treatment will be needed. higher speed collisions can result in more MPA is VP Chief What is certain is that soon medical bills will trauma, surgeries, and complications, as Medical Officer for arrive soon. well as longer rehabilitation periods, thereby AIS, a CCC Company. increasing casualty claims severity. Casualty claims management makes a difference for consumers in the increasingly complex healthcare environment. Through Marijuana Legalization the combination of clinical expertise and Impaired driving risk is non-uniform across technology, auto casualty solutions optimize the U.S. For example, states have varied in the digital processing of medical bills causally their actions on marijuana legalization. As TRENDS related to motor vehicle collisions. While of early 2019, Alaska, Washington, Oregon, every accident is unique for the people, California, Nevada, Colorado, Michigan, time, place and circumstances involved, Vermont, Massachusetts and Maine had IMPACTING AUTO accumulating data-driven insights show legalized consumer recreational and medical how macro-trends impact auto injury claims marijuana use, while 22 other states had severity and challenge auto insurers. Through the INJURY CLAIMS authorized prescribed medical marijuana. In its 2019 report, the U.S. Congressional combination“ of Jurisdictional Trends Research Services summarized the U.S. clinical expertise SEVERITY AND National Academy of Sciences’ significant State jurisdictions differ in auto personal finding of statistical association between and technology, auto injury protection insurance regulations; marijuana use and increased risk of motor casualty solutions CHALLENGING however, recent data reveals less-apparent vehicle accidents. state-based influences on auto casualty optimize the digital claims risk. Regulations on speed limits, Alcohol-Associated processing of medical AUTO INSURERS marijuana legalization, alcohol-related driving bills causally related law enforcement, and the opioid epidemic Impaired Driving vary significantly across the U.S., reshaping to motor vehicle by Dr. Mary Capelli-Schellpfeffer patterns of auto casualty claim severity. Differences in state alcohol-associated collisions." impaired driving patterns have been noteworthy as well. As reported by the After a car accident, we might find ourselves in an emergency Road Speed Limits Department of Transportation’s National department, dressed in an exam gown and covered by a thin Center for Statistics and Analysis based on blanket on a hospital gurney. In that moment we have no way Over the last decade road speed limits have 2017 data, Texas, Wyoming and Montana had to speed through a doctor’s evaluation, diagnostic testing, increased in many localities, including the 35 percent or more alcohol-impaired driving or clinical treatment. There’s no steering wheel to grab on to, states of Wisconsin, Illinois and Indiana.

143 CRASH COURSE 2020 COURSE CRASH

drivers (20.5 percent) was aged 65 years or trauma (BAT) has been in part due to Continued older. Of these, 3.4 percent were accountable generally accepted EBM guidelines. for 50,807 of Michigan’s total number of 314,921 crashes in 2017, suggesting a highly • Attention to diagnoses commonly disproportionate risk of collision for senior associated with motor vehicle accidents

TRENDS IMPACTING AUTO drivers. has increased. For example, heightened

monitoring of traumatic brain injury DATA FOR | TIME Substantiating the recognition of age-related diagnoses has evolved in many states in INJURY CLAIMS... driving risk for seniors, a published 2017 tandem with state emergency medical national report of age-related differences in systems (EMS) and state trauma fatal intersection crashes shows teens and registries. As a result, there is wider drivers aged 55 years and older were “over- awareness the association between involved” in fatal intersection crashes, while motor vehicle accidents and the fatalities, while Rhode Island’s experience neighborhood and socioeconomic group, drivers aged 20 to 54 years were “under- occurrence of clinical conditions such as was at 41 percent, Connecticut at 43 percent law enforcement and healthcare providers involved. Remarkably, the study showed that headaches, concussion, and traumatic and Washington D.C. at 51 percent. continue to struggle to meet the escalating the fatal intersection driving risk across all brain injury or TBI. First responders and epidemic’s fallout. There is little documented driver age groups was doubled by age 85. emergency clinical providers are trained Alcohol-impaired driving fatalities are an indication that the opioid epidemic has to evaluate for these conditions in auto indicator of the prevalence of driving under abated, suggesting continued driving risk accidents, just as today team coaches the influence in the state of record, as associated with prescribed and illicit opioid Treatment Complexity are expected to manage player risk of coroner involvement and death certificate use. concussion in sports situations across all The third major trend related to auto data are rigorously addressed. These age groups. casualty claims severity is greater treatment fatalities have been understood as the “tip complexity. Treatment complexity in of the iceberg” when it comes to the scale Aging Populations • Age-related treatment complexity, for auto casualty reflects broader healthcare of each state’s challenge to combat drunken pediatric and senior populations, is now A second macro trend shifting U.S. auto initiatives related to evidence-based medicine driving behaviors. widely recognized. In this regard, the casualty claims severity is seen with the (EBM) guidelines, diagnostic intensity, and American College of Surgeons and the growth of aging U.S. populations. A key data aging-related care considerations. Pediatric Trauma Society’s guidelines Opioid Associated insight is that the generational transition have been influential, as these are with the maturing of “Baby Boomers” has the basis for EMS protocols in most Driving Fatalities not been distributed evenly across all states. • EBM guidelines, protocols, and states. Clinically relevant concerns Between 2007 and 2016, the number of recommended practices influence clinical Impaired driving attributable to use of include medical frailty, eligibility for people aged 65 years or older in the United evaluation, treatment and rehabilitation opioids also intensified across the last certain types of care, or difficulty with States increased by 30 percent (males by nationally, by jurisdictional and on a decade. For the U.S., the percentage of post-injury rehabilitation protocols. 35 percent and females by 26 percent), with specialty basis. As electronic medical fatally injured drivers who tested positive Regarding seniors, data show that the total population across all age groups record systems penetrate large health for prescription opioids rose seven-fold from in injury treatment scenarios, older increased by 7 percent. States most recently system operations, EBM’s incorporation 1 percent in 1995 to over 7 percent in 2015. patients are likely to have more than affected by this demographic shift include into patient evaluation protocols, According to the U.S. Centers for Disease one pre-existing medical condition, Texas, California, Florida, Georgia, North prescribing systems, diagnostic testing Control and Prevention, between 2016 and take multiple medications, and be at Carolina, Washington and Arizona. algorithms and care paths is more 2017, Oregon, Arizona, Minnesota, and North risk for complications even with simple routine. For example, the consistent and Carolina experienced significantly increased procedures. As generational aging has impacted state expanded use of diagnostic imaging with numbers of opioid deaths, signaling epidemic populations, the distribution of licensed suspected head injury or blunt abdominal crises in those areas drivers has trended older as well. For example, of Michigan’s 2017 population, 72.3 While public awareness has increased percent were reported as licensed drivers. about the opioid use prevalence in every In 2017, one out of five Michigan licensed

TRENDS IMPACTING AUTO Head of Auto Claims INJURY CLAIMS... Farmers Insurance DATA FOR | TIME

Summary more efficiently. Efficiency gains rely on the How do you/your company spend use of artificial intelligence, machine learning, and data analytics to advance digital claims Jurisdictional differences, aging populations management. your time? and increasing treatment complexity are impacting auto casualty risk and claims These three key innovations link the sciences management. These trends are likely to of engineering and medicine. As a result, the continue into 2020 and beyond. The good We continue to focus on our future through the lens of time-in-process for resolving the reimbursed news is that technology solutions can assist an ecosystem, which revolves around our customers and relationships in an auto casualty claim carriers in managing locational, generational between providers, healthcare facilities, employees. The digital realities of artificial intelligence, and healthcare-related changes in auto and allied health resources can be reduced. machine learning, robotic processing, telematics and digital casualty claims severity. Furthermore, the consumer can benefit from process mapping are all escalating the rate of change. the knowledge that their insurer is addressing Moreover, by aligning deep medical insights This has also increased the possibilities of connectivity their claim with expert processing relying about how auto crashes might result in and new joint ventures within the industry. As a company, on the facts and details personal to their “ injuries, along with state-specific automated accident. we’re continually evaluating how we evolve the skills of our processing of healthcare transactions, employees so they can more easily embrace the changing medical bill payments can be addressed environment; the industry collaborators we work with to help create the change, and ability for all of it to be seamless for the customer."

Despite the prevalence of improved head restraint systems, seat backs and other

related occupant safety features, treatment DATA FOR | TIME has become more complex and the overall cost of treating these same injuries has CCC INSIGHT increased.

The types of auto-related BI and PIP/Medpay diagnoses most frequently seen in the last three years have remained relatively unchanged and are predominately by nature soft tissue neck and back injuries. Despite the prevalence of improved head restraint systems, seat backs and other related occupant safety features, treatment has become more complex and the overall cost of treating these same injuries has increased.

At the same time, the average age of the claimant continues to increase in both BI and PIP claims. According to Pew Research, 29 percent of Baby Boomers aged 65- 72 years older were still employed or looking for work in CY 2018, a higher share than for the Silent Generation or Greatest Generation when they were the same age.185 With employment high among an age group traditionally retired, it also means more individuals are on the roads during peak commuting hours. Data from the 2014-2018 American Community Survey (ACS) shows labor force participation rates among those ages 65 to 74 years increased to 25.7 percent from 25.2 percent in the 2009-2013 ACS and increased from 5.9 percent to 6.6 percent for those ages 75 years and older.186 Continued increases in this metric over the longer-term point to higher medical costs per patient as the treatment of older patients, all things considered equal, can become more complex and thus more costly relative to their younger counterparts. With advanced age, individuals may appear during crash triage with inter-current health conditions, more use of medications, and physical or psycho-social vulnerabilities in the setting of acute trauma, thereby

149 CRASH COURSE 2020 COURSE CRASH CCC INSIGHT CONT.

complicating their initial clinical evaluation and management (E&M) and requiring more initial Additionally, many of the top BI and PIP/Medpay treatment procedures have ranges to be TIME FOR DATA FOR | TIME medical, surgical or diagnostic procedures. According to the U.S. Census Bureau projections, by CY considered when being billed. These ranges include the level of service selected or the number of 2030, all Baby Boomers will be older than age 65 — meaning one in every five residents in the U.S. units that are employed. The percentage of claims that include these types of medical services has will be of retirement age.187 grown; in fact, the Insurance Research Council recently published research that showed medical utilization rates exceeding national norms were a primary contributor to high claim costs in eight 190 Recent data from Gallup’s annual Health and Healthcare survey reveals 43 percent of U.S. of the 12 least affordable auto insurance systems. Other factors included: High injury claim households have at least one member in the family with a “long-term medical condition, illness or frequency rates, excessive attorney involvement, and high rates of claim abuse. disease that would be considered a ‘preexisting’ condition by a health insurance company,” down slightly from 46 percent from the survey one year prior.188 Preexisting conditions among adults As more people move to high-deductible insurance plans or plans with co-insurance provisions that aged 65 and older and aged 50-64 report are reported at a rate twice as high as the rate reported mean higher out-of-pocket expenses, the temptation to find a way to have that covered elsewhere by young adults aged 18-29 (see Figure 123).189 grows.191 Results of a survey conducted by the Kaiser Family Foundation and the L.A. Times reveal annual deductibles in employer-based health plans nearly quadrupled in the last decade, averaging Americans’ Reports of Preexisting Medical Conditions in Their Household, 2018-2019 about $1300.192 Yet, more than 40 percent of workers in the high-deductible plans lacked the (FIGURE 123) | SOURCE: MCCARTHY, JUSTIN. “43% OF U.S. HOUSEHOLDS REPORT PREEXISTING CONDITIONS.” WWW.GALLUP.COM, DECEMBER savings to cover such an expense, choosing instead to cut back on household spending, delaying 193 6, 2019. or skipping needed medical care. And the growing number of people who work within the gig economy means fewer covered under group health insurance plans. Do you, personally, or does anyther member of your family living with you, have a long-term medical condition, illness or disease that Since CY 2016, numerous efforts by the Republican-controlled U.S. government to repeal the ACA would be considered a “pre-existing condition” by a health have resulted in several changes that weaken the bill. Among these changes were the following. In insurance company? December 2017, Congress reduced the individual mandate penalty to $0 effective in 2019.194 In 2018, new rules were announced that allow more loosely regulated plans. Additional changes expanded Respondent or family member Respondent personally the use of health reimbursement arrangements (HRAs) which let employers use tax-exempt funds to help workers pay for medical plans in entirety, essentially subsidizing workers buying entire 50% health plans on the individual market instead of offering a group plan.195 These changes may enable 65+ years old 33% healthy people to find lower cost coverage, but that coverage may be less comprehensive, and can exclude some pre-existing conditions.196 Finally, in mid-December 2019, Congress passed a repeal of three taxes meant to help pay for the Affordable Care Act, and the 5th U.S. Circuit Court of Appeals ruled that the individual mandate is unconstitutional and unlawful under federal taxation 49% 197 50-64 years old rules. While this ruling weakens the bill, California is appealing the case to the Supreme Court. 31%

A January CY 2019 report by the Kaiser Family Foundation reported the number of uninsured in the U.S. rose in 2017 for the first time since the implementation of the ACA to 10.2 percent of the 40% nonelderly population (see Figure 124).198 As a result, there are higher numbers of uninsured 30-49 years old 23% Americans, which will add to the pressure to reduce healthcare expense, particularly by Medicaid.199 Subsequently, with less money available in the system, the P/C insurance industry might become an even bigger target of cost-shifting in the future. 38% 18-29 years old 16%

Number of Uninsured and Uninsured Rate Among the Nonelderly Population (ages 0 President and CEO to 64), CY2008-CY2017 Fix Auto USA (FIGURE 124) | SOURCE: KAISER FAMILY FOUNDATION ANALYSIS OF 2008-2017 AMERICAN COMMUNITY SURVEY (ACS), 1-YEAR ESTIMATES.

Uninsured (in millions) Uninsured Rate 50 25% 46.5 45 45.7 45 44.2 44.8 44.4 40 20% 35.9 35 What do you think the biggest evolution 17.8% 17.4% 30 17.1% 17.3% 17.0% 16.8% 29.1 15% 26.7 27.4

25 13.5% 20 10% in the industry will be? 10.9% 10.0% 10.2% 15 10 5% 5 Today’s independently owned and operated body shops 0 0% are finding themselves in a landscape that didn’t look the CY2008 CY2009 CY2010 CY2011 CY2012 CY2013 CY2014 CY2015 CY2016 CY2017 same even two years ago; it’s evolving that rapidly. Because vehicles are becoming more complex, technologically advanced machines, the need for vehicle manufacturers, Finally, growth of ADAS-equipped vehicles within the registered vehicle population is expected to insurers, and body shops to come together to ensure safe reduce auto injury frequency and severity based on analysis from IIHS/HLDI and others. Research “ and proper repairs has never been more imperative. It conducted in our Crash Course 2019 revealed the Delta-V, or severity of a crash, may be less for just seems unrealistic that any one body shop will be able ADAS equipped vehicles involved in a crash. The Delta-V (or speed loss or gained in a crash) as well as the speed of the vehicle prior to the crash and the direction the impact came from, are to serve every customer. Therefore, the need to combine important in determining the mechanics and severity of the injury. If ADAS is in fact helping to resources and provide a complete repair platform has never avoid more low speed crashes, the remainder of injury claims would have the potential to be from been more important. This is really the next major evolution higher speed crashes where injury severity is higher. Other variables such as the age of the driver, of our industry and we’re positioning our enterprise to be as well as driver’s understanding of ADAS true capabilities (i.e. mode confusion, role confusion, ready for this, and to also influence how these programs are and misplaced trust) may mean ADAS could have very different benefits/outcomes depending on the demographics of the driver. So, despite all the potential benefits to BI and PIP frequency and designed, implemented, and successfully managed.” severity expected from ADAS, people are still involved, so it becomes ever more critical to focus on the experience that is distinct to each individual, and to understand the implications of those differences to your business. CRASH COURSE 2020 COURSE CRASH

TIME TO PREPARE TO | TIME TIME TO PREPARE EMBRACING THE NEXT WAVE OF TECH

06 Until now, the utilization of mobile, AI, and IoT were part of the strategy for pioneers and leaders within our industry.

So, what is next? As we look to 2020 and beyond, it’s important to remember that while we cannot predict the future, we can prepare for it. In this final section of Crash Course 2020, we look at some of the biggest challenges facing the industry, and what opportunities technology of today and the future may offer.

155 CRASH COURSE 2020 COURSE CRASH

Many of the most impactful new technologies developed over the last several decades came Moving forward, those same technologies will play a transformative role and will become building from the consumer world.200 The invention of the smartphone released a wave of new mobile blocks to the other technologies such as autonomous vehicles, voice technology, IoT hardware, 5G technologies that enabled consumers to connect whenever and to whomever they wanted or connectivity, intelligent automation, industrial software, robotics, semiconductors, extended reality, needed, through new social channels that never existed before. According to data from Pew and quantum computing.204 Research, smartphone adoption grew from 35 percent of U.S. adults in CY 2011 to 81 percent by CY 2019, with nearly 50 percent of all U.S. adults saying they mostly go online using their We have moved into the post-digital world — which Accenture writes now means that “… digital

201 smartphone. technologies are now so integrated into people’s lives that they have become part of consumers’ PREPARE TO | TIME identities … technology identities are part of an emerging feedback loop ... resulting [in] ‘snapshots’ Many of the biggest ideas in technology over the past of insight into customer needs and goals to deliver personalized products and services … [and now] decade have centered on how people communicate, individualized experiences.”205 consume, transact and travel. Over the next decade, however, the most profound innovations — and “ investment opportunities — could be on factory Transformation of Risk Management floors, in operating rooms, at mining sites and energy The insurance of tomorrow will continue to pay claims but may provide more technology to avoid facilities." claims altogether. Already carriers offer customers IoT and sensor technologies to do things like monitor leakage from washing machines and hot water heaters, smart fire detectors that alert Morgan Stanley emergency services, and health and fitness trackers. As the digital ecosystem becomes more SOURCE: “THE SECOND MACHINE AGE HITS THE TIPPING POINT.” AUGUST 5, 2019. intelligent and understands more about each person’s background, preferences, interests, and intentions, digital experiences will become ever more personalized. The addition of sensors in devices like activity trackers, smartphones, smart speakers, vehicles, and more enable companies to provide their customers what they needed with greater convenience and While continuing to be a payer, the insurer of the personalization. The number of IoT devices has exploded, outnumbering individual human beings by nearly 400 percent (see Figure 125).202 According to Deloitte, smart home devices alone are future will take a holistic approach to customer risk expected to see compound annual growth rates of 60 percent, which translates to 600 million management by evolving to become both a partner smart home devices by 2021.203 “ and a preventer."

IoT Devices Outnumber Human Brains SOURCE: CAPGEMINI AND EFMA. 2019 WORLD INSURANCE REPORT. (FIGURE 125) | SOURCE: THE ECONOMIST, UNITED NATIONS (AUTONOMOUS RESEARCH REPORT, APRIL 2018).

IoT Devices (in billions) Humans (in billions) 28.1 25.2 In the future, technology such as location detection, smart sensors and augmented reality will do 22.2 more than simply collect and deliver information. In CY 2017, researchers used Google Street View images to study the distribution of car types in the U.S. and then used that data to determine the 19.2 demographic makeup of the country — including things like income level, education, occupation, 16.3 and even the way you vote in elections.206 More recently, researchers at Stanford and the University 13.1 of Warsaw in Poland used Street View images of people’s houses to determine how likely they are to be involved in a car accident. Their research found that features visible on a picture of a house were predictive of car accident risk, independent of variables such as driver age or zip code.207 7.3 7.4 7.5 7.5 7.6 7.7 While numerous issues would need to be resolved, the wealth of data available to potentially collect, analyze and utilize will continue to grow rapidly in the future.

2015 2016 2017 2018 2019 2020

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 157 CRASH COURSE 2020 COURSE CRASH For insurance company customers with claims, the integration of numerous new data sets with Climate Change is Costly artificial intelligence and machine learning will help inform recommended course of action.208 In many cases that may mean a claim gets created and routed automatically to the appropriate The University of Cambridge Centre for Risk Studies’ Global Risk Index predicts as much as channel, essentially ‘touchless.’ For the smaller share of claims that fall outside the norm or appear $584 billion, or 1.55 percent of 2020 global GDP could be erased from the global economy by to be a situation where a human interaction would be beneficial, the claim is routed to a human catastrophic events in 2020 — an increase of 3 percent from the prior year.210 Key to how the world with the appropriate skills and most importantly, the empathy needed for that particular claim.

manages its exposure and its recovery are insurance, reinsurance, and insurance-linked securities

Repetitive tasks like providing status updates and monitoring treatment schedules for injured PREPARE TO | TIME (ILS) capital.211 Allianz SE estimates fighting climate change will cost companies worldwide nearly claimants can be automated — yet personalized, and free up people’s time to focus on the issues $2.5 trillion over the next ten years.212 that benefit most from the human touch or expertise. Significant storms and catastrophic events continued to batter the U.S. in CY 2019. In early July, In the 2019 World Insurance Report, Capgemini’s research shows insurance company customers the meteorological phenomenon ‘Ring of Fire’ hit numerous parts of the Planes, Midwest and are increasingly interested in new insurance models that could help cover new risk scenarios Northeast. Several days after baseball-sized hail hit Denver and Fort Collins, as well as other part stemming from changing macro trends such as technology, demographics, and the environment of WY, NE and SD. In total, these storms as well as others in July alone led to an estimated $2 (see Figure 126).209 billion in damage.213 In total, the U.S. experienced three billion-dollar events related to convective storms — economic losses for the first half of the year in the U.S. was 14 percent above the 21st century average.214 Hurricane Dorian slammed into the Bahamas in early September as a Category 5 hurricane with 185 mph winds — one of the strongest cyclones ever recorded in the Atlantic Capgemini 2019 World Insurance Report — “Customers’ interest in new insurance 215 models, individual and business customers (%), 2019” Ocean. Also in September, Tropical Storm Imelda dumped more than 40 inches of rain in areas east of Houston into Louisiana, resulting in over 10,000 vehicle flood claims.216 As of late October, (FIGURE 126) | SOURCE: CAPGEMINI FINANCIAL SERVICES ANALYSIS, 2019; CAPGEMINI VOICE OF THE CUSTOMER SURVEY, 2019. CY 2019, California wildfires had caused $25 billion of damage to properties.217 On a single day — October 20, 2019 — nine tornadoes were confirmed to have touched down around Dallas county resulting in losses of around $2 billion — the costliest tornado outbreak in Texas history.218 Business customers Individual customers Changes to the environment bring significant challenges to all — the ability to underwrite the

34.0% risk, cover the losses, and reduce the exposure from business interruption and property loss will Microinsurance 15.4% require transformation of traditional insurance models. There is some hope that technology can help address some of these risks. For example, there are several companies (Descartes Labs, Planet 34.8% Labs, and Jupiter Intelligence) that use data from satellite imagery and sensors, integrate with Peer-to-peer insurance 14.9% other data on the weather, terrain, vegetation, humidity, wind, etc. to provide real-time data to firefighters and even model potential dangers from one hour to 50 years out.219 34.3% Parametric insurance 15.5% As the number and magnitude of risks grow, insurers’ success is dependent on the ability to use data and technology to deeply understand the risks, to help mitigate losses, to expand coverage 40.9% Proactive value-added services and offer extreme personalization. As vehicle, personal tracking devices, and home technologies 21.1% develop, the ability to use the data collected and data available from numerous other sources will be central to risk management. 44.2% On-demand insurance 28.8%

46.7% Usage-based insurance 35.0%

As we discussed at the start of this publication, automakers and other technology companies continue to make huge investments in the "ACES" — autonomy, connected car, electrification, and shared vehicles. Bob Cretel

Claims Group Manager The testing and development of technologies to support a fully autonomous vehicle for personal Auto Club of California PREPARE TO | TIME use, last-mile delivery, and numerous other scenarios continues to ramp up, despite timelines for mass-market delivery pushed out by at least a decade. In 2020, Waymo will expand its pilots of autonomous vehicles across numerous cities, and expand its testing of Waymo One, its commercial driverless taxi fleet. In China, Baidu’s Apollo Robotaxi goes live with its fleet of robo-taxis operating at Level 4 autonomy in Changsha, Hunan province and in North China’s Hebei province. As China looks to become a global leader in autonomous and connected car technology, it has ramped up autonomous vehicle trials on dedicated roads and test centers.220 What do you think the biggest evolution In 2019, global private AI investment was over $70 billion, with investment in AI startups growing from $1.3 billion in CY 2010 to over $40.4 billion in 2018.221 The Artificial Intelligence Index Report in the industry will be? 2019 showed the largest share of global investment in 2019 went to Autonomous Vehicles ($7.7 billion or 9.9 percent of the total), followed by Drug, Cancer and Therapy ($4.7 billion, 6.1 percent), Facial Recognition ($4.7 billion, 6.0 percent), Video Content ($3.6 billion, 4.5 percent) and Fraud Detection and Finance ($3.1 billion, 3.9 percent).222 Having spent almost four decades in the personal lines insurance industry, I’ve had the benefit of seeing many The autonomous vehicle of the future will be a platform integrating numerous types of technology changes during my career. Artificial Intelligence (AI) adapted and software from numerous providers. With people increasingly removed from the task of driving, for insurance claims will be the biggest evolutionary change, the variables used to determine risk and appropriate coverage will change dramatically. Access to the information generated by the vehicle, by infrastructure, by other vehicles, and more will perhaps even revolutionary change, in our industry. Much of be critical to inform risk and liability. Who has access to that information may also change who is the claims resolution process involves a consistent process underwriting that risk in the future. “ of gathering key bits of data to determine coverage, liability and damages. Artificial Intelligence tools have shown to As automakers, technology companies, insurers and others look to the future of transportation, be easily adapted to assist with the process. I do not know anticipating which technologies will be most impactful is perhaps the biggest challenge. Creating an environment where new technology can be understood, studied, and tested for various how it works, but I know how it helps. From completing applications will be central to an organization’s readiness for the future. rudimentary tasks to damage assessment to automatic payments, AI is streamlining the claims process as never before. Artificial Intelligence applied to facts, damages, and GPS quadrants have shown to be able to replicate real events removing some of the guess work of claim resolution.

How insurance carriers adapt to AI will determine their ability to compete in the coming years both in service and financially. The key will be for AI to augment the claims professional’s skills rather than replace the claims professional. We have become a very connected society. Gathering, filtering and effectively using all this data for the benefit of consumers will require confidence and a little blind faith." © 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 161 CRASH COURSE 2020 COURSE CRASH 1. We are passionate and committed to Key Fundamental technological advancement as a way to deliver services to our customers that Pillars of Innovation benefit from these new technologies. To be successful, we must understand and at CCC explore numerous hypotheses related to the embodiment of these technologies The success of any innovation facility is measured ultimately by one major factor: into our products and services. PREPARE TO | TIME How quickly and successfully have we 2. It is essential that we work closely with delivered operational solutions to our customers as we navigate together this customers that substantially improve their maze of new technologies. To do so, ABOUT THE AUTHOR service and operations? Focused on success we need dedicated facilities (our lab), and value delivery, we’ve organized around Reza Rooholamini resources, and processes. five core pillars: Vice President, Chief 3. To evolve and deliver the best of the Architect, CCC • Talent: we have staffed the operation best solutions and services, we need to with highly technical people with diverse be agile, proactive, and focused on the technical foundations. Over 90 percent core business and evolving peripheral of our technical staff have advanced markets. degrees (Master and Ph.D.) in a variety of disciplines such as computer vision, 4. We must invest in our most valuable computer science, operation research, assets — our people, who need to mathematics, statistics, and quantum continue to evolve their knowledge of physics. This enables our innovators to technology and the market. AN INCUBATION analyze problems from many different perspectives. FACILITY WITHIN R More than 60 CCC: WHY? insurance“ customers have embraced our by Reza Rooholamini mobile platform and The computer and communication industry is facing the we’re moving toward largest technology inflection point seen in the last 30 years. a similar tipping point Technological advances like artificial intelligence (AI), advances in mobile computing, Natural Language Processing/ TALENT CONTINUOUS DATA STRONG DIRECT with AI-powered Understanding (NLP), Internet of Things (IOT), and cloud RESEARCH & CONNECTION CUSTOMER solutions, achieving a computing are greatly influencing and driving change within EXPERIMENT TO THE CORE INVOLVEMENT every sector, including the industries served by CCC and our level of maturity that customers. supports streamlined incorporation into In order to proactively address and embrace these and other advanced and enabling technologies, CCC has been hard at the workflow and work in our own research and solutions incubation facility. delivering immediate The main driver for such an initiative? There are several: benefits." 163 CRASH COURSE 2020 COURSE CRASH

models that can detect with high level Continued of accuracy the likely damaged parts Ahead in 2020 from a few photo images. The system also recommends replacement or repair The solutions delivered from our work in of the damage parts based on the the Innovation Foundry have been vetted in 2019 and have demonstrated substantial AN INCUBATION FACILITY inputs provided. This technology has

been integrated with CCC’s estimating gains towards smart, digital, and efficient PREPARE TO | TIME platform. Other examples of use of these claim processing. More than 60 insurance WITHIN CCC: WHY? models include a predicted range of customers have embraced our mobile total cost of repair from photos, or given platform and we’re moving toward a similar an estimate of the repair, models flag tipping point with AI-powered solutions, estimates that need to be re-examined. achieving a level of maturity that supports We have also developed models to streamlined incorporation into the workflow • Continuous experimentation: At any What have we examine the quality of the input images and delivering immediate benefits. given time, our staff is conducting from a damaged car and detecting the dozens of parallel experiments. These accomplished so far? angle from which the image was taken; In 2020, CCC’s Innovation Foundry will experiments are designed with defined for example, left-front, right-front, left- continue to deliver on its promise and scope and timelines. The results are In the three years since opening the doors rear, etc. charter. The activities will continue to support either followed into a full developed to the CCC Innovation Foundry we have digital APD and claim processing. Our mobile product or valued for their contributions produced and introduced significant • Telematics based Accident Detection: platform will continue to evolve both from a as learning opportunities. solutions and industry advancements, of From the signals received from the car, computational perspective and as a platform which we are extremely proud, including: for information ingestion. Our research also • Data: We have developed a data via intelligent analysis we can detect remains focused on further digitization and analytics and science platform to host occurrences of several events including automation of auto casualty claim processing. and enable transformation of data in • Mobile Platform-based APD Claim an accident. This capability enables a Telematics-based data captured at the time potentially new ways. Processing: The end-to-end digital claim series of potential proactive services processing starts with the smart mobile including informing emergency services of the accident provides opportunity to • Tight coupling with the core business: device. We have developed and delivered and insurance company engagement. further enhance Straight Through Processing The innovation foundry operates within a consumer-led experience capturing (STP) of casualty claims. the financial and operational framework damage related images via a mobile that governs the core and growth phone. The ingested data feeds the businesses of CCC, making the activities insurance appraisers’ mobile platform of the innovation foundry are planned work area for assessing the damage. This and correlated. has dramatically increased accuracy and efficiency of estimation and overall claim • Direct customer involvement: Customer processing. feedback in the innovation process has resulted in an extremely practical and purposeful environment. The foundry • Incorporation of Artificial Intelligence often has several experiments and in CCC’s Estimating Platform and research being conducted by a CCC team Workflow: We have developed hundreds soliciting feedback from customers of AI models for integration within the customer’s workflows. For example, we have developed an ensemble of

Earlier in Crash Course 2020, we discussed the impact of ride-hailing, car-sharing, vehicle connectivity, and incorporation of the building blocks to full vehicle autonomy on today’s vehicle repair. As the number of connected vehicles grows, capabilities to capture and analyze data collected from the growing number of sensors and software modules will help to avoid accidents altogether, or where unavoidable, provide critical information regarding damages to both the vehicle and its occupants.

That information could be used to identify which parts are damaged, the extent of the damage, pre-order the parts, and route to the repairer that best meets the unique criteria set for that consumer. Repairers must respond to a new environment of ultra-personalization and specialization. As automakers and technology companies work to craft their own ‘special sauce’ for the "ACES," the challenges to keep up with training and tooling will grow. Having an effective plan in place to train and retain staff will be critical. Segmenting market opportunities based on potential profit, volume, business partners, and opportunity to provide world class service will become more challenging but increasingly important. How a shop is chosen for a certain repair is likely to change, THE but how a world-class experience is delivered remains dependent on high quality of repair and ability to provide transparency and service throughout.

At the start of a new decade, we stand poised for immense change and opportunity.

Over the last 10 years, technology like mobile, IoT, and AI has brought dramatic change to the who, how, why and when of customer engagement. Companies across all industries have worked over the last several years to develop digital platforms that provide a unified platform experience that some now expect, where parties are connected, to ultimately deliver both transparency and saved time. And perhaps most important for our industry, these technologies have freed up talent to provide a personalized, high-touch experience, providing emotional support to those in a highly ROAD AHEAD traumatic situation.

Looking forward, as new mobility options such as car-sharing, ride-sharing, bike-sharing, and autonomous vehicle technology mature, the one-vehicle to one-person model is expected to change. Traditional roles played by automakers, insurers, repairers, and other players in our ecosystem will continue to blur, and collaboration will be paramount to provide a seamless and streamlined experience. Personalized experiences are the expected norm.

While there remains much uncertainty for the future, and the ability to predict the future is perhaps more difficult than ever, understanding how to prepare for the future is paramount. IT'S TIME.

37. U.S. DOT FHWA Policy & Government Affairs Highway Policy Ratemaking, Product and Modeling Seminar. Information Highway Statistics, 1995 to 2016. “Licensed Drivers, Vehicle Registrations and Resident Population (in Millions).” 60. 2017 National Household Travel Survey, p.8 and p.13, Tables 1a and https://www.fhwa.dot.gov/policyinformation/statistics/2016/ 3b. https://nhts.ornl.gov/assets/2017_nhts_summary_travel_ dv1c.cfm. trends.pdf.

38. https://www.bloomberg.com/news/articles/2020-01-13/ 61. Conway, Salon, and King. “Trends in Taxi Use and the Advent of millennials-could-end-up-being-a-boon-to-the-u-s-auto-market. Ridehailing 1995-2017: Evidence form the U.S. Household Travel Survey.” Urban Science, 2018, 2, 79. 39. https://medium.com/@sjpeterson/are-car-free-and-car-lite-

lifestyles-on-the-rebound-58f67f470a09 62. Schaller Consulting, “The New Automobility: Lyft, Uber and the Future of American Cities.” July 25, 2018. | 40. Ibid. SOURCES 63. Ibid. 41. Kurz, Christopher, Geng Li, and Daniel J. Vine (2018). “Are Millennials Different?” Finance and Economics Discussion Series 64. Steven Vance and John Greenfield. “The Most Common Chicago SOURCES 2018-080. Washington: Board of Governors of the Federal Reserve Ride-Hailing Trip Is a 1-Mile Hop From River North to the Loop.” System, https://doi.org/10.17016/FEDS.2018.080. https://chi.streetsblog.org/2019/04/18/the-most-common- chicago-ride-hailing-trip-is-a-1-mile-hop-from-river-north-to- 42. Christopher R. Knittel and Elizabeth Murphy. “Generational Trends loop/. in Vehicle Ownership and Use: Are Millennials Any Different?” MIT Center for Energy and Environmental Policy Research. CEEPR WP 65. https://www.seattletimes.com/seattle-news/transportation/how- 2019-006. April 2019 popular-are-uber-and-lyft-in-seattle-ridership-numbers-kept- secret-until-recently-give-us-a-clue/?amp=1. 43. https://www.wsj.com/articles/u-s-existing-home-sales-decreased- 1-7-in-november-11576767884. 66. John M. Barrios, Yael V. Hochberg and Livia Hanyi Yi. Stigler Center for the Study of the Economy and the State. 44. III Quarterly PC Industry Snapshot, Published April 12, 2019. University of Chicago Booth School of Business. “The Cost 1. www.autonews.com, January 6, 2019. 20. https://www.autoconnectedcar.com/2019/12/evs-more-popular- of Convenience: Ridesharing and Traffic Fatalities.” October for-chinese-french-car-buyers-than-americans/ 45. U.S. Census Bureau, 2017 American Community Survey 2018. https://research.chicagobooth.edu/-/media/research/ 2. J.D. Power. “Used Car and Light Truck Guidelines: October 2019 1-Year Estimates. https://factfinder.census.gov/faces/ stigler/pdfs/workingpapers/27thecostofconvenience. Industry Review.” www.nada.com/b2b. P. 2. 21. Ibid. tableservices/jsf/pages/productview.xhtml?pid=ACS_17_1YR_ pdf?la=en&hash=A15B1513F98D7A17B9 B08141&prodType=table. B08141 “Means of Transportation to E37F78DD2EBDC4C6338BFA. 3. Ibid, P. 4. 22. https://store.frost.com/future-of-carsharing-market-to-2025.html Work by Vehicles Available. Universe: Workers 16 Years and over in households.” 67. https://www.nber.org/papers/w22627.pdf. https://www. strategyanalytics.com/strategy-analytics/ blogs/ automotive/ 4. IHS Automotive Registered Vehicle and Scrappage Report through 23. https://www.usatoday.com/story/news/world/2013/06/13/un- connected-mobility/automotive-connected-mobility/ 2019/08/20/ January 1, 2019. world-population-81-billion-2025/2420989/ 46. Global Workplace Analytics and Flexjobs. “2017 State of Telecommuting in the U.S. Employee Workforce.” P. 7. alto-taps-uber’s-consumer-surplus. 5. http://energyfuse.org/are-americans-driving-older-cars-or-just- 24. https://www.futuremind.com/blog/car-sharing-and- 68. Michael Graehler, Jr. Richard Alexander Mucci. Gregory D. leaving-them-in-the-driveway/ transportation-trends. 47. https://www.wardsauto.com/dealers/auto-dealers-get-personal- digital-retailing. Erhardt (corresponding author). Department of Civil Engineering, University of Kentucky. “Understanding the Recent Transit 6. Naughton, Keith. “Ford Unveils Electric Mustang SUV to Challenge 25. Intellias Automotive. “Five Boldest Automotive Tech Trends in Ridership Decline in Major U.S. Cities: Service Cuts or Emerging Tesla Dominance.” www.bloomberg.com, November 18, 2019. 2019.” www.intellias.com, November 13, 2018. 48. https://www.wsj.com/articles/in-cities-across-america- Modes?” 98th Annual Meeting of the Transportation Research the-fight-for-curb-space-heats-up-during-the-holidays- Board. Submitted: August 1, 2018 and Revised and re-submitted 11576238403?mod=djemlogistics_h 7. https://www.wsj.com/articles/car-makers-rushing-to-copy-tesla- 26. Boudette, Neal E. “Despite High Hopes, Self-Driving Cars Are ‘Way November 14, 2018. lack-one-thing-buyers-11575647982 in the Future’.” www.nytimes.com, July 17, 2019. 49. Trucking by the Numbers 2019: Overview. www.fleetowner.com 69. “After Uber arrives, heavy drinking increases.” https://www. 8. Temple, James. “Why the electric-car revolution may take a lot 27. Ibid. economist.com/graphic-detail/2019/11/19/after-uber-arrives- longer than expected.” MIT Technology Review, November 19, 50. U.S. EIA. 2019 Annual Energy Outlook, www.eia.gov/aeo, January heavy-drinking-increases, November 19, 2019. 24, 2019. P. 120. 2019. 28. https://www.iihs.org/news/detail/new-studies-highlight-driver- confusion-about-automated-systems 70. Meera Joshi, Nicholas Cowan, Olivia Limone, Kelsey McGuinness, 9. Ibid. 51. Ibid. and Rohan Rao. NYU Wagner Rudin Center for Transportation. “E-Hail Regulation in Global Cities.” November 2019. https:// 29. https://www.autonews.com/mobility-report/ntsb-scrutinizes- wagner.nyu.edu/files/faculty/publications/RUDIN_EHAIL_ 10. Naughton, Keith. “Ford Unveils Electric Mustang SUV to Challenge automation-complacency-after-uber-crash. 52. https://usa.streetsblog.org/2016/03/10/its-true-the-typical-car- REPORT.pdf. Tesla Dominance.” www.bloomberg.com, November 18, 2019. is-parked-95-percent-of-the-time/. 30. Kim, W., Kelley-Baker, T., Sener, I., Zmud, J., Graham, M. & Kolek, 71. National Association of City Transportation Officials. Guidelines 11. https://www.wsj.com/articles/gm-lg-to-spend-2-3- S. (2019). Users’ Understanding of Automated Vehicles and 53. INRIX defines inner city last-mile speed as “the speed at which for Regulating Shared Micromobility: Version 2. www.nacto.org, billion-on-venture-to-make-electric-car-batteries- Perception to Improve Traffic Safety –Results from a National a driver can expect to travel one mile into the central business September 2019. 11575554432?mod=djemlogistics_h Survey. AAA Foundation for Traffic Safety. district during peak hours.” “INRIX 2018 Global Traffic Scorecard,” INRIX, February 2019, http://inrix.com/scorecard/. 72. Weinberg, Cory. “Lime’s Loss to Top $300 Million in 2019.” www. 12. Ibid. 31. Els, Peter. “5 Setbacks to the future of mobility following the fatal theinformation.com, October 21, 2019. Uber accident.” www.Automotive-iQ.com, 04/03/2018. 54. Texas A&M Transportation Institute and ATRI. Urban Mobility Report 2019. August 2019. P. 5. 13. Lawson, Philippa“The Connected Car: Who is in the driver’s seat? 73. Ibid. A study on privacy and onboard vehicle telematics technology.” 32. Kolodny, Lora. “Self-driving cars were supposed to be here already Office of the Privacy Commissioner of Canada. -here’s why they aren’t and when they should arrive.” www.cnbc. 55. American Trucking Research Institute. “The Nation’s Top Truck com, Sat Nov 30, 2019. Bottlenecks 2019.” www.truckingresearch.org. 74. American Association of Motor Vehicle Administrators. “Electric Dockless Scooters Whitepaper.” www.aamva.org, 2019. 14. Charniga, Jackie. “Ford CEO: Decontenting boosts vehicle affordability.” www.autonews.com, December 4, 2019. 33. “Road-safety regulator rapped over regulating autonomous cars.” 56. Andrew Owen, Brendan Murphy, Accessability Observatory, Claims Journal, November 22, 2019. https://www.claimsjournal. Center for Transportation Studies, University of Minnesota. David 75. APCIA Bulletin. “Electric Scooter Insurance Model Act Proposed at Levinson, School of Civil Engineering, University of Sydney. Access NCOIL.” Thursday November 14, 2019. 15. Ibid. com/news/national/2019/11/22/294242.htm. Across America: Transit 2016. CTS 17-07. 76. https://medium.com/radical-urbanist/dont-be-fooled-uber-doesn- 16. Els, Peter. “Part One: Shared mobility is a game changer, but is it 34. Martinez, Michael. “Automakers exploring autonomous tech find 57. https://www.un.org/development/desa/en/news/ t-care-about-transit-e89880d375b0. viable.” www.automotiveiq.com, 10/29/2019. safety in numbers.” www.autonews.com, March 2, 2019. population/2018-world-urbanization-prospects.html. 77. Cox, David. “Smart road technology could turn highways into 17. Ibid. 35. CB Insights. What’s Next in AI? Artificial Intelligence Trends 2019. www.cbinsights.com, p. 72. 58. Trucking by the Numbers 2019: Traffic. www.fleetowner.com crash-sensing ‘touchpads’.” June 25, 2018. https://www.nbcnews. com/mach/science/can-smart-roads-save-lives-new-test-may- 18. https://justauto.nridigital.com/just-auto_magazine_dec19/ 59. Auto Loss Cost Trend Report, January 2018 (Casualty Actuarial provide-answer-ncna885816. shared_mobility_and_mobility_as_a_service_maas. 36. Annmarie Geddes Baribeau. “Driverless Utopia.” Casualty Actuarial Society Review, May 4, 2018. Society, Property Casualty Insurers Association of America, Society of Actuaries). Insurance Information Institute, Loss 78. https://insideevs.com/news/382664/gotland-installation- 19. Ibid. Cost Presentation given 3/21/2018 Casualty Actuarial Society’s dynamic-wireless-charging/

79. “Automakers Seek Control in Connected Car Software.” https:// september-2019-10. 134. Vincent Romans. “Advancing Our Insights Into The Collision Repair 162. https://injuryfacts.nsc.org/motor-vehicle/overview/introduction/ tu-auto.com, 8th October 2019. Marketplace.” MSO Symposium presentation, Las Vegas, NV, 108. https://www.fleetowner.com/equipment/economic-forecast- November 4, 2019 163. https://consumer.healthday.com/public-health-information-30/ 80. Statista, March 2019. (https://www.statista.com/outlook/320/109/ trucking-gloomy-chance-recession marijuana-news-759/12-million-americans-drove-while-stoned- connected-car/united-states?currency=usd#market-arpu). 135. https://www.autonews.com/fixed-ops-journal/dealers-battle- last-year-753236.html. 109. Zaremski, Michael and Lederer, Charles. Credit Suisse. “Property behind-body-shop-work. 81. https://www.strategyanalytics.com/strategy-analytics/ & Casualty Insurance. Reserves Update: Moderating Loss-Expense 164. Ibid. blogs/automotive/infotainment-telematics/infotainment- Inflation on Newly Written Comm’l Auto Biz.” 23 December 2019. 136. Lajdziak, Robert M. “Customer experience: The auto insurance telematics/2018/10/25/connected-cars-high-tech-low-demand. x-factor.” Aug 19, 2019. https://www.propertycasualty360.com. 165. https://www.iihs.org/topics/alcohol-and-drugs. 110. ISS Fast Track Plus™ Personal Auto, As of June 30, 2019.

82. https://www.strategyanalytics.com/strategy-analytics/blogs/ 137. Lajdziak, Robert M. “High Customer Satisfaction Leads to New 166. National Safety Council. The State of Safety: A State-by-State | automotive/connected-mobility/automotive-connected- 111. Ibid. Challenges for Insurers.” September 18, 2018. https://www. Report. www.nsc.org/stateofsafety, pp. 10-11. mobility/2019/07/08/onstar-rolls-back-free-trials. carriermanagement.com/features/2018/09/18/184296.htm. SOURCES 112. Ibid. 167. Ibid, p. 7. 83. Ibid. 138. https://www.zendesk.com/resources/searching-for-self-service/

113. Ibid. 168. Based on IIHS/HLDI Studies completed by April 3, 2016, IIHS 84. “Stout Automotive Defect & Recall Report.” www.stout.com. 139. Lajdziak, Robert M. “Customer experience: The auto insurance presentation at Lifesavers Conference 2017. x-factor.” Aug 19, 2019. https://www.propertycasualty360.com. 114. USDOT NHTSA. “Police-Reported Motor Vehicle Traffic Crashes in 85. Ibid. 2018.” DOT HTS 812 860, November 2019. 169. https://www.aaa.com/AAA/common/aar/files/Research-Report- 140. Auto Insurance Report, Vol.27#7/1256. November 4, 2019. Pedestrian-Detection.pdf 86. Singh, Sarwant. Frost & Sullivan’s Global Automotive Outlook, 115. Ibid, p. 8. 2019. “Top Automotive Trends in 2019: A Year of Wows and Woes.” 141. Ibid. 170. https://www.autoblog.com/2019/12/12/iihs-top-safety-pick- 116. “What’s Happening and What Lies Ahead for P/C Insurance pedestrian-crash-avoidance-headlights/ 87. “Among Bad Driving Behaviors, Speeding is Strongest Predictor of Pricing: Moody’s.” https://www.insurancejournal.com/ 142. CB Insights. “The Future of Insurance Keynote.” www.cbinsights. Crashes.” www.insurancejournal.com, August 22, 2019. news/2018/10/30/506028.htm. com. 171. Micah Hartman, Anne B. Martin, Joseph Benson, Aaron Catlin and The National Health Expenditure Accounts Team. “National 88. Governors Highway Safety Association. Speeding Away from Zero: 117. Auto Insurance Report, Vol.26#38/1239. June 24, 2019. 143. National Safety Council. Injury Facts 2017. Health Care Spending in 2018: Growth Driven By Accelerations Rethinking a Forgotten Traffic Safety Challenge. www.ghsa.org, in Medicare and Private Insurance Spending.” DOI: 10.1377/ January 2019. P. 20. 144. Auto Care Association. Auto Care Factbook 2019, p. 57. IMR Inc. hlthaff.2019.01451 HEALTH AFFAIRS 39, NO. 1 (2020): –©2020 118. Moore, Alan, P.E. “Tesla Autopilot, Self-Driving Vehicles and Driver Project HOPE—The People-to-People HealthFoundation, Inc. Assistance Systems; Who Caused The Accident?” Presentation at (www.AutomotiveResearch.com) Continuing Consumer Auto 89. Ibid. the Symposium on Traffic Safety, May 21-24, 2018. Maintenance Tracking Study. Copyright 2018, IMR Inc. 172. Ibid. 90. Ibid. 119. Reagan, Ian J. “Effects of an Aftermarket Crash Avoidance System 145. “Total customer lifetime value for dealers estimated at more than on Warning Rates and Driver Acceptance in Urban and Rural $50K.” www.autoremarketing.co, Dec. 4, 2019. 173. Ibid. 91. U.S.DOT NHTSA Traffic Safety Facts. “Critical Reasons for Crashes Environments.” Springer International Publishing AG, part of Investigated in the National Motor Vehicle Crash Causation Springer Nature 2019 N. Stanton (Ed.): AHFE 2018, AISC 786, pp. 146. Huetter, John. “Pega: Independent repairers part of ‘moments that 174. Ibid. Survey.” DOT HS 812 115, February 2015. 776–787, 2019. https://doi.org/10.1007/978-3-319-93885-1_71. matter’ for OEM brand.” www.repairerdrivennews.com, January 2, 2019. 175. Ibid. 92. Annmarie Geddes Baribeau. “Driverless Utopia.” Casualty Actuarial 120. Jessica Cicchino. “Autobrake is good, but it could be better.” IIHS/ Society Review, May 4, 2018. HDLI Status Report, Vol. 54, No. 2. February 21, 2019. www.iihs.org. 147. Gerardi, Ryan. “Trends Affecting Car Dealership Fixed Ops.” www. 176. The Institutes and Insurance Research Council. Trends in Auto wardsauto.com, Jan 02, 2020. Injury Claims, 2019 Edition. February 2019, p. 7. 93. USDOT NHTSA Traffic Safety Facts. “Distracted Driving in Fatal 121. CREF I-CAR Snapshot of the Collision Repair Industry: Executive Crashes, 2017.” DOT HS 812 700, April 2019. Summary of the 2019 Survey. 148. Finlay, Steven. “Most Polled Auto Dealers Say They Fluctuate Parts 177. Ibid., p. 2. Pricing.” www.wardsauto.com, Dec 04, 2019.

94. https://www.wsj.com/articles/car-makers-wager-on-bigger- 122. Finlay, Steve. “Car Dealers Amp Up Efforts to Recruit Young 178. Auto Insurance Report, Vol.26#38/1239. June 17, 2019. screens-to-lure-buyers-11574677804?mod=hp_lead_pos11. Mechanics.” www.wardsauto.com, Oct 24, 2019. 149. NADA DATA 2018. “Average Light-Vehicle Dealership Profile”.

179. Ibid. 95. Ibid. 123. “Repair myths and misinformation.” 03/23/2016 -13:24. www. 150. NADA DATA Mid-Year 2019. “Average Light-Vehicle Dealership searchautoparts.com. Profile”. 180. “IRC: Auto Injury Claims Costs Outpace Inflation, Lawyer 96. Ibid. Involvement Increases.” April 23, 2018. https://www.claimsjournal. 124. http://www.bodyshopbusiness.com/cover-story-the-importance- 151. Overby, Joe. “Parts & service, along with used cars, helps com/news/national/2018/04/23/284266.htm. 97. Chirag Ramesh Kalose. “Unveiled: the most common scoring of-oem-data. push franchised dealer profits, optimistic valuations.” www. autoremarketing.com, December 1-14, 2019, p. 26. factors for Usage-based Insurance.” Thursday, July 25, 2019. 181. Ibid. https://www.ptolemus.com/insights. 125. Becker, Mitch. “Electronic blueprinting a necessary step in repairing today’s vehicles.” www.bodyshopbusiness.com, 152. McNamara, K. “E-commerce is shaping the aftermarket supply December 2, 2019. chain.” www.searchautoparts.com, 04/09/2019. 182. https://www.wsj.com/articles/who-wins-in-a-personal-injury- 98. Ibid. lawsuit-it-can-be-the-doctor-11578479400. 126. Huetter, John. “With new technology in cars, calibration becomes 153. Ibid. 99. CreditSuisse Equity Research, Property & Casualty Insurance. 22 183. Credit Suisse Equity Research, 14 January 2020. “Social Inflation July 2019. vital with changes.” www.repairerdrivennews.com, November 17, 2015. Is Getting Personal – Tort Inflation Isn’t Only Impacting Comm’l 154. Ibid. Lines.” 100. Webinar held by Bryce Wilson, Teletrac Navaman. 2019 Telematics Benchmark Report – U.S. Edition. 127. Yoswick, John. “ADAS Calibrations Latest Challenge Facing 155. NADA DATA 2018. “Average Light-Vehicle Dealership Profile”. Collision Repairers.” www.autobodynews.com. 184. Auto Insurance Report, Vol.26#6/1255. October 28, 2019.

101. Clinton, Paul. “Telematics Market Expected to See Explosive Near- 156. Ibid. 128. Ibid. 185. Zender, Mathew. “The aging workforce and workers’ Term Growth.” www.automotive-fleet.com, January 23, 2019. compensation.” www.propertycasualty360.com, October 15, 2019. 157. Experian. “State of the Automotive Finance Market Q3 2019 and 129. “Preparing for advanced material repairs with certification, 102. https://www.automotive-fleet.com/303123/fleet-safety-metrics- Q3 2015.” 186. U.S. Census Bureau. “Census Bureau News: American Community reverse-negatively-accidents-increase training, and equipment.” www.searchautoparts.com, 12/15/2016 – 08:49. Survey 2014-2018 5-Year Estimates Now Available.” Release CB19- 158. Ibid. 196. Dec. 19, 2019. 103. Insights from USI National Practice Leaders. Commercial Property 130. Ibid. & Casualty Market Outlook Q4 2019-2020 Update. P.7. 159. AnnaMaria Andriotis and Ben Eisen. “A $45,000 Loan for a 187. Zender, Mathew. “The aging workforce and workers’ $27,000 Ride: More Borrowers Are Going Underwater on Car compensation.” www.propertycasualty360.com, October 15, 2019. 104. https://www.wsj.com/articles/surging-truck-insurance-rates-hit- 131. Auto Care Association. Auto Care Factbook 2020. “U.S. Shops Loans.” www.wsj.com, November 9, 2019. Specializing in Collision Repair CY 2003-CY 2018. freight-operators-11578934834?mod=djemlogistics_h. 188. https://news.gallup.com/poll/269003/households-report- 160. U.S. DOT NHTSA Traffic Safety Facts. “Early Estimate of Motor preexisting-conditions.aspx?version=print 105. Ibid. P.2. 132. Auto Care Association. “Auto Care Factbook 2020.” Page 55. Vehicle Traffic Fatalities for the First 9 Months (Jan-Sep) of 2019.” Bethesda, MD: 2020. DOT HS 812 874, December 2019. 189. Ibid. 106. https://www.automotive-fleet.com/303123/fleet-safety-metrics- 133. Ibid. reverse-negatively-accidents-increase 161. U.S. DOT NHTSA Traffic Safety Facts. “2018 Fatal Motor Vehicle 190. Insurance Research Council. “Affordability in Auto Injury Insurance: Crashes: Overview.” DOT HS 812 826, October 2018. Cost Drivers in Twelve Jurisdictions.” June 21, 2016. http:// 107. https://www.businessinsider.com/trucking-truck-drivers-job-loss-

© 2020 CCC INFORMATION SERVICES INC. ALL RIGHTS RESERVED. 171 www.insurance-research.org/sites/default/files/downloads/ 208. Lyons, Daniel. “Insurance 4.0: Six winning strategies in the fourth NRcostdrivers2016.pdf. industrial revolution.” www.dig-in.com, November 14, 2019.

191. Insurance Research Council. “The Affordable Care Act and 209. Capgemini and Efma. World Insurance Report 2019. www. Property-Casualty Insurance.” Copyright 2014. www.theinstitutes. worldinsurancereport.com, P. 18. org, p. 5. 210. https://www.artemis.bm/news/as-exposure-rises-insurance- 192. Yural Rosenberg and Michael Rainey. “Why So Many Americans capital-central-to-global-recovery-research/. Can’t Afford Health Care.” www.thefiscaltimes.com, May 2, 2019. 211. Ibid. 193. Ibid. 212. “Insurer Estimates $2.5 Trillion Price Tag for 10-Year Fight 194. Rachel Garfield and Kendal Orgera, Kaiser Family Foundation. The Against Climate Change.” December 4, 2019. https://www. Uninsured and the ACA: A Primer, January 2019, p. 5-6. insurancejournal.com/news/national/2019/12/04/550266.htm.

195. “How Trump is Remaking Health Insurance Without Repealing 213. www.artemis.bm/news/u-s-severe-storms-in-july-cost-2bn- Obamacare.” www.insurancejournal.com, June 17, 2019. insurers-to-take-majority-aon./ August 8, 2019.

196. Ibid. 214. Aon plc 2019. Global Catastrophe Recap: September 2019.

197. https://www.ozy.com/presidential-daily-brief/pdb-258080/ 215. Aon plc 2019. Global Catastrophe Recap: First Half of 2019. July stricken-258601/. 2019.

198. Rachel Garfield and Kendal Orgera, Kaiser Family Foundation. The 216. Taylor, Brittany. “More than 10,000 vehicles damaged during Uninsured and the ACA: A Primer, January 2019, p. 7. Tropical Storm Imelda, insurance companies say.” September 23, 2019. https://www.click2houston.com/news. 199. Schramski, Tom. “7 Ways TrumpCare Will Change the Healthcare Industry in 2017.” www.axial.net, January 11, 2017. 217. www.Artemis.bm/news/California-wildfire-damage-said-25bn- spread-continues-fire-weather-forecast-worsens/. October 29, 200. Morgan Stanley, “The Second Machine Age Hits the Tipping Point.” 2019. August 5, 2019. 218. www.Artemis.bm/news/dallas-tornado-insured-loss-said-2bn- 201. https://www.pewresearch.org/fact-tank/2019/12/20/10-tech- costliest-ever-texas-outbreak/. October 25, 2019. related-trends-that-shaped-the-decade/. 219. Wayne Allen, CEO. “Some Hope in the Face of the Wildfire Threat.” 202. Autonomous Research LLP, “Autonomous Next: #Machine Blog.insurancethoughtleadership.com, April 30, 2019. Intelligence & Augmented Finance.” http://next.autonomous.com, April 2018. 220. https://justauto.nridigital.com/just-auto_magazine_dec19/ adas_to_avs_-_more_progress_in_2020_but_with_added_ 203. https://www.propertycasualty360.com/2019/2019/01/16/four- reality_check PREPARED BY trends-shaping-insurancein-2019/ 221. Raymond Perrault, Yoav Shoham, Erik Brynjolfsson, Jack Clark, 204. Morgan Stanley, “The Second Machine Age Hits the Tipping Point.” John Etchemendy, Barbara Grosz, Terah Lyons, James Manyika, August 5, 2019. Saurabh Mishra, and Juan Carlos Niebles, “The AI Index 2019 Annual Report”, AI Index Steering Committee, Human-Centered AI Institute, Stanford University, Stanford, CA, December 2019. P. 6. 205. Accenture. “The Post-Digital Era Is Upon Us: Get Ready for What’s Next in Insurance.” Technology Vision for Insurance 2019, www. SUSANNA GOTSCH accenture.com, p. 20. 222. Ibid. DIRECTOR, INDUSTRY ANALYST, CCC 206. https://www.technologyreview.com/s/613432/how-a-google- street-view-image-of-your-house-predicts-your-risk-of-a-car- accident/. Susanna Gotsch is Director, Industry Analyst at CCC Information Services Inc. 207. Ref: arxiv.org/abs/1904.05270: Google Street View Image of a House Predicts Car Accident Risk of Its Resident. She has been with CCC since July of 1992.

She has authored Crash Course, CCC’s annual publication on trends impacting collision repair and total loss costs since 1995. This publication has become a key resource for the industry in understanding how broader trends within the economy, new and used vehicle market places, and collision industry are impacting auto claim frequency and costs. Ms. Gotsch also consults with the on-going development of CCC’s industry-leading data warehouse and reporting products.

Ms. Gotsch has presented her research at numerous industry meetings and symposiums. In 2011, Ms. Gotsch was awarded the annual industry honorarium Most Influential Women in the Collision Repair Industry established by AkzoNobel Automotive & Aerospace Coatings America (A&AC).