OPERATION OF AN AUDITED-MILE/YEAR AUTO INSURANCE SYSTEM UNDER PENNSYLVANIA LAW

OPERATION OF AN AUDITED-MILE/YEAR

AUTO INSURANCE SYSTEM

UNDER PENNSYLVANIA LAW

Patrick Butler

National Organization for Women, Inc.

loo0 1Slh street. W, We 700. Washinpton. CC 2@2355705

(202) 3310068

FAX (202) 7858576

Date: November 1992

TO:

Readers, The Forum, Casualty Actuarial Society

From: Patrick Butler, Ph.D.

Director, Insurance Project

Re: Practicality of the Car-Mile Exposure Unit for Auto

Insurance

The following 29-page review was prepared at the

request of Pennsylvania legislators, who are considering

rate regulatory bills that would mandate use of the car-mile

exposure unit for driving coverages.

The focus is on individual transactions because the

questions of convenience and control of odometer fraud are

generally accepted as an impracticality barrier to a "pay-

by-the-mile" method of earning premiums.

The theory behind the car-mile exposure unit is

straightforward. Since every mile traveled by a car

transfers risk to its insurer, the product of (a cents-per-

mile class rate based on the class's per-mile cost

experience) X (miles recorded on the car's odometer)

appropriately earns prepaid premium while the car is driven.

Apparently there has been no other consideration of

the impracticality issue since Paul Dorweiler's 1929 paper

on exposure units stated that "[t]he devices and records

necessary for the introduction of [the car mile] medium make

it impractical under present conditions.11

16 pcAs 319, 338;

58 PCAS 59, 78. For this reason, it is hoped that this

revFcan serve as a framework for renewed, informed

consideration of the practicality question.

308

OPERATION OFANAUDITED44ILE,‘YEAR AUTO INSURQNCE SYSTEM

UNDER PENNSYLVANIA LAW

Prepared by

National Organization for Women

June 22,1992

Executive Summary

This review examines the practical implications for insurance companies and

Pennsylvania car owners of converting premium calculation for most coverages from

dollars-per-year to cents-per-mile class rates. The purpose is to provide an

operational model for evaluating proposed legislation mandating this conversion

(SB 775 and HB 1881). Operation of a mile/year system is described through a

sequence of transactions for a hypothetical car over four policy years.

Advance payment continues to be required for keeping insurance protection in

force. Administrative expense and premium for nondriving coverages (theft, fire,

hail) at year rates are paid at policy-year renewal time. Premium for driving

coverages (liability, medical, collision) at mile rates is prepaid in mileage amounts

and at intervals chosen by the car owner. The car’s insurance ID card displays the

odometer-mile and date limits to prepaid protection.

Policy renewal is conditional on taking the car to a garage designated by the

company for the annual physical audit of its odometer. The odometer is calibrated

and read, and tamper-evident seals are applied at the initial audit. Tampering with

the odometer voids the insurance protection.

The possibility of stealing insurance protection under the mile/year system is

explored. Control measures are described using two examples: a lO,OOO-mile roll

back and stopping the odometer for 10,000 miles. (Driving with the cable

unhooked, surprisingly, does not steal insurance protection, because it usually would

be detected after an accident and tampering voids protection.)

The opposite possibility under the current year system is examined:

policyholders having to pay premiums during nondriving periods when their cars are

not consuming insurance protection and do not need it. Current “suspension of

coverage” provisions for periods of non driving appear to be cumbersome,

inadequate, and inconsistently applied. The present administrative handling of

premium refunds for non-driving periods is compared to the mile/year system’s

automatic response to non-driving periods.

The review concludes by examining compulsory-insurance enforcement and

compliance under the year and mile/year systems respectively. Attention is given to

the negative effect that year-system enforcement has on ability to pay for insurance

in comparison to the positive effect of the mile/year system on car owners’ ability to

comply with requirements.

309

OPERATION OF AN AUDITED-MILE/YEAR AUTO INSURANCE SYSTEM

UNDER PENNSYZVMU LAW

Prepared by

National Organization for Women

June 22,1992

I. Introduction

This review examines the practical implications for insurance companies and

Pennsylvania car owners of converting most automobile insurance coverages from

year to mile class rates, The purpose is to provide a framework for evaluating

legislation mandating this conversion from one-part to two-part pricing: from time

rates only to a system using both distance rates and time rates. The legislation,

which has been introduced in the Pennsylvania Senate and House and is under

consideration in other states, would add one sentence to the state’s Casualty and

Surety Rate Regulatory Act: The exuosure units for calculation of private passeneer

automobile insurance uremiums at the atmrcqrtate classification rates shall be the

mile bv audited odometer readings for driving coverws and the vear for nondriving

coveraeeLl

The review assumes that this amendment is the only action by the Pennsylvania

Legislature that would be needed to change prices for on-the-road coverages from

dollars per car year to cents per car mile.

2 The methods of conversion and operation

can be decided by the individual companies, constrained only by existing law on

insurance and motor vehicles .3 Self-interest and competition on service should

1. Identical bills--Senate Bill 77.5 and House Bill 1881--were introduced in the 1991-92

Pennsylvania General Assembly and referred to the iusurauce committees.

Premiums for driving coverages charged at mile rates according to odometer readings is a method

that works with any amount of coverage and all risk

classiicati~ns.

To calculate a premium: multiply

the rate for the car’s class (say 4 cents a mile) by the odometer miles of protection needed (say 10,OCKl

miles) over a time period (say one year). ‘Ihe resulting premium: $400. For urban residents with the

same coverage at a 9-ant rate, the premium for 10,880 miles protection would be $900.

2. Some regulatory changes would be necessary, however, because regulation of automobile

insurance transactions is specific in some areas. Changes would be needed in current specifications for

the insurance ID card contents, for example, and also regulatory review and approval would be needed

for new policy language regarding odometer fraud.

3. For example, the mile rate for each classification and coverage would still be held to the

Casualty and Surety Rate Regulatory Act standard that ‘[rlates shag not be excessive, inadequate, or

unfairly discriminatory.’ 40 P.S. Sec. 1183 (d).

NOW holds that, as an expression of public policy, this chief provision of rate regulation against

310

2 1 AUD~‘IFD-MLE/YEXR SYSTEM

I. Introduction

assure development of company systems that are efficient, convenient and credible

for consumers, and that effectively control premium fraud.

To test the operation of a mile/year system, it was judged preferable to study

one method in detail rather than attempting to anticipate a range in methods that

may be developed by individual companies. The test system is intended to be a fully

functional prototype. System specifics, such as provisions in the insurance contract

relating to odometers, are intended to help focus discussion.

As an introduction to its operation, the mile/year system selected is described

in w (page 3) through a sequence of transactions for a hypothetical car over

four policy years.

Section (page 12) examines the methods of odometer auditing and the

possibilities for fraud, in preparation for the next section on theft of insurance

protection.

m (page 14) explores the possibility under the mile/year system of theft

of insurance protection by policyholders tampering with their odometers, and

describes measures taken to prevent it.

&&RY (page 20) examines the opposite possibility under the current system:

policyholders having to pay premiums during nondriving periods when their cars are

not consuming insurance protection and do not need it. The section describes how

insurance companies now provide premium refunds for some kinds of nondriving

periods, but not for others. The present administrative handling of premium

refunds is compared to the mile/year system’s automatic response to non-driving

pXiOdS.

The final section (action Vk page 26) reviews compulsory-insurance

enforcement and compliance under the year and mile/year systems respectively.

Attention is given to the negative effect that enforcement has on ability to pay for

insurance under the year system in comparison to the positive effect of the

mile/year system on car owners’ ability to comply with requirements.

(continued)

ccst&iftiug among polieybolden requires the use of distance rates rather than the current time rates

for

driving coverages in auto insurance.

pridng. The Insurance

Commisioner, however, denied illegal

cost-sidflhg iu &u&r&

NOW

v. Q&&na, and was upheld by the Commonwealth Court,

A.2d 1162 (1988). The evidence and NOW’s response to the decisions are presented in three papers

published by the Journal of Insurance Regulation: Butler, Butler, & Wiiams, Ser Divided MiZeage,

Accidcn~ And Insurance Cost Data Show lhat Auto kwws Overcharge Most Women, 6

J. INS. REG.

243, 373 (1988); Butler, Butler, & Wiltiamq Insumnce

Depanment ‘Catxh-22’ Shields Auto Insurers

Fm Cwuumer C%ai&Jes, 7

1.1~~.

Rffi. 285

(1989);

Butler

& Butler, Driver

Ream-L

A Political Red

Hening lkzt Rewab the Bade Flaw in Automobile Insumce Pnking, 8

J.Ih’s. RED. ulo (1989).

311

AUDITED-MILE/YEAR SYSIFM

I. Introduction ( 3

A subsequent paper will treat topics, such as ratemaking for conversion from

year to mile units of exposure, that would be of direct interest to auto insurers but of

less immediate concern to most legislators.

4 Work is also continuing on other

related topics such as effect of the mile/year system on policy contract provisions

that include accidents in a rented car under coverages for an owned car, and on

arrangements

for

protection of lien holders’ security interests.

II. How a mile/year svstem ooerates: an example

Automobile mechanical breakdown insurance (“service agreement” or “extended

warranty”) uses units of distance and time (miles and years) to measure and price

insurance protection.5 It thus provides a model for the mile/year auto insurance

system. Contract language from a mechanical breakdown policy for used cars

(Exhibit A) gives

the necessary rules for measuring protection with an odometer:

WHENANDWHEREYOUAREPROTECTED

Protection included in the Plan

YOU select

is available as soon as

YOU

receive this Agreement.

****

This Agreement expires 12 months after the Effective Date or, when

YOUR

CAR registers 12,000 miles more than the Odometer Reading at Inception,

whichever comes first.

l ***

WE will not pay benefits if the odometer of the covered vehicle has stopped

or been changed.

**t*

The paper

Making Mile Rates for Automobile Insumnce

is in work. The proposal for doing

the paper was accepted by the Casualty Actuarial Society in December 1991 as a candidate for the

society’s ratemaking seminar in March 1993.

The current method of determining the

cost per claim

does not need to be modified if

appropriately applied. (Cost per claim = total cost of claims/total number of claims.) The mileage

information needed to determine the claims per mile rate for each driving coverage is not collected at

present but will be after

the conversion. (Claims per mile = number of

c&x./number of miles

driven.) The paper will examine ways in

which

the information can be determined in advance with

sufficient accuracy for making

the

conversion. The

cost per

claim multiplied by number of cfuims

per

mile equals the per-mile cost of providing protection. For example, with an average cost of $2COO per

claim and one claim per lOO,OKl miles, the mile cost for one coverage may be calculated: $2,COO/claim

x 1 cIaim/1OO,CH3O miles = $.02/miIe.

The Pennsylvania insurance department regulates the rates and policy forms for mechanical

breakdown insurance as for all other types of automobile insurance. Premiums for a policy on a used

car can exceed SMO (4 cents per mile) for a 12,CKWmile/l-year protection period.

312

4 ( AUDITED-MILE/YEAR SYSIEM

II. Lhzmpte operation

EXHIBIT A

Mechanical breakdown insurance policy

AUDWED-MILE/YEAR SYSIFM

II. C-ample operation 1 5

This Agreement may be terminated as follows:

- WE will terminate this Agreement for non payment of the Agreement

charge.

- WE will terminate this Agreement if the odometer is disconnected or

altered.

- If this Agreement is terminated, YOU may be entitled to a refund for the

cost of unused protection. Unused protection is the lesser of the unused

days or the unused miles of protection available.

- In the event YOU initiate a termination, a $10 service charge will be

deducted from the refund.

The same kind of policy conditions apply to coverages under the mile/year

system: insurance protection is strictly prepaid (as it is now); consumption of

driving protection is measured in distance units; consumption of nondriving

protection is measured in time units; and odometer tampering voids the driving

coverages.

Unlike the mechanical breakdown policy, however, the mile/year insurance

system routinely renews mileage and time protection periods. Mileage renewals are

in amounts and at intervals chosen by the policyholder, while the time period for

renewal is the policy year. Policy year renewal is conditional on complying with

company requirements (as it is now), such as providing rating information in the

renewal application. An important renewal requirement in the mile/year system is

taking the car for an annual physical audit of the odometer and its seals as directed

by the company.

From the policyholder’s perspective, the transactions that take place over a

policy year are:

Before end of previous policy year, the policyholder

Receives annual audit and renewal notice with premium bill

Obtains physical audit of car’s odometer

Pays dollars-per-year premium for nondiving coverage and fixed expense;

buys miles of driving coverage needed at current cents-per-mile rate

Receives car’s insurance ID card showing the odometer and date limits to

prepaid protection

During policy year, the policyholder

Buys additional miles needed at current cents-per-mile rate

Receives car’s insurance ID card showing the revised odometer limit to

prepaid protection

314

1 AUD~?IID-MILE!/Y~%R S~sreht

II. Example operation

To demonstrate operation of the mile/year system, the following hypothetical

example (Exhibits B through E) tracks the insurance transactions for a single car

over nearly four policy years.

Exhibit B.

The sequence of transactions, which begins with 3,000 miles on the

odometer when the car is acquired and ends with its sale at 37,000 miles, is shown in

a graph of odometer readings vs. time.

The upper, stepped plot shows miles of insurance protection bought. Vertical

segments of the “miles prepaid” line represent purchased miles of protection and are

located at the dates on which the premium payments are credited. Each horizontal

segment represents the odometer limit to prepaid protection until additional miles

are bought.

The lower plot

a “miles driven” line connecting the odometer audit points.

The line segments between audits represent average driving exposure, expressible in

miles per day or miles per year. (A plot of

the

actual miles of exposure, by which

the policyholder consumes prepaid protection and the company earns premium in

providing it, would vary in slope between horizontal for periods of no driving to

steeply positive--e.g. 500 miles per day--for

long

trips. A day-by-day plot of

odometer reading, nonetheless, would also pass

through

the odometer audit points.)

Exhibit

C. The insurance ID card, which the company is required by law to

provide for each car it insures,6 communicates the car’s insurance status to the

policyholder by prominently displaying both the mile total of prepaid protection

(expressed as an odometer reading) and a policy year renewal month.

Exhibit D (& Exhibit

B).

table of the transactions between policyholder and

company for the example car lists the premium payments and audits over the four

policy years examined.7 (Transactions are keyed by number to the odometer vs.

date graph, Exhibit B.)

The first transaction takes place on

March

15. 1991, when the new owner takes

possession of the car. Its odometer reads 3,000

miles

and the owner buys 12,000

6. 75 Pa.C.S. SW. 1782 (d) and 31 Pa. Code Sec. 61.23 et

seq.

7. The issue date of the ID Card is assumed

be 7 calendar days after the later of: 1)

oolicyholder mailiw of the renewal aoolication with any premium paid: or 2) the odometer audit.

iThi; time accoua& for internal co;pany processing; kuding iranimittal of odometer audit

information from vendors.) Receipt of the ID Card by the policyholder is taken as 4 days after the

issue date.

315

AUDITED-MILE/YEAR

SYSTEM

II. Example operation 1 I

EXHIBIT

Transactions For Example Car

50,000

40,000

30,000

dometer

?eadlng

20,000

10,000

0,000

‘91

‘92

‘93

Time (In years)

I+11

: : j

i i :

, j :

j i .

MllOS prepaId

Miles driven (avg.)

+ Audlt points

miles of insurance protection. The company provides a binder as proof of insurance

pending issuance of an ID card, which is contingent on completion of the initial

odometer audit and sealing within 30 days according to company rules.8

8. It is assumed that non-conformity to policy contract conditions would constitute valid cause

for nonrenewal at the end of the policy year and the within-HI-day cancellation period permitted for

new poIicies. 40 P.S. Sections llXl8.4 and lW8.6.

316

8 1 AUDiTFD-MILE/YITARSYS~E~~

IL Example operation

EXHIBIT C

Insurance ID Card For Example Car

Fitwncial Rwpcmlbility ldantificatim (ID) Card

InsurmcaCarplny"mm

capany natxr ,nwrancI policy I

lhlmd inwred wrers

Vehicle

Make N&l

vehicle Idew. II-r WIN)

ID Card * woteetim s&a on

it.9 L*

Josw 0 tr; adaste

,.,.:I ,&j &,*

Cdcmeter mfla Date

3,700

AFR-13-91

* IO CARD IS NOT VALID

AFTER MILE OR KIN11

LIMITS SHWN, OR IF MtllETER STWS.

If I linlt is mached, the car is NOT insured (IS rcqirsd by Pcruwylvania

Lau. DO NOT DRIVE IT

mtil

more insurmca is purchesed. If the WCMETER

STOPS, telsphom your inrursnee capany for instructions before driving.

(back of card -

raqulrcd statemnfs, 31 P&Code 67.24)

This card must be carried for productfcn qm dmml. It is suRgested that

you carry thIR card In the insured vehlcls.

"ARWING: Any omer or ragirtrat of e amtar vehiclr uho drives or prmits

e motor vehicle to ke driven In this State without the required flnsnclsl

mpaslbility may hRve hiR reglstratfm suspsndcd or revoked.

NOTE:

THIS CARD IS REWIRED UHEN:

(1) You are involved in *n auto accident.

(2) You WC cmvlcted of e traffic offar other than II perking offense

that requiraa a court eppaarbnce.

(3) Ycu we stopped for violating any prevision of 75 Pa.C.5. Wclatlng

to the Vehfcla Co&) *Id requested to prc&cc

W a pcllec offictr.

Yw must provfds a copy of this card to the Department of Trensportatim when

you reqazst restoratIon of your cperatlng privilege and/or rwistratlcm

privilege which has been

prevfcusly

susperded (lr retied.

After the first odometer audit on April 13, 1991, which shows that the car has

been driven 700 miles since purchase, the company issues the ID card, Exhibit C.

There are no further transactions until February 1992, near the end of the 1st policy

year. Along with the renewal application form, the company communicates the

1992 mile rates and bills for any year-rate charges for the 2nd policy year. The

policyholder, however, is responsible for buying sufficient miles of protection at miZe

rates for the anticipated use of the car.

317

AUDITED-MUE/YPAR SETIN

II. Example operation ( 9

EXHIBIT D

Transactions and ID Cards For Example Car

1ST POLIC" "EAR ,991

nar-13 F,\h epq,,cat,m far iluurvr. cm f., miwJ prChU.d.

#I nar-15 P.y. cm-yw.r sh.rs.8 p1w 12,000 c.r ml," Of pcotecrh et 1991 ret., vh,Ch w3d.d to th*

3,000 m admew (ml rem-dad a Ott* et trensfcr) puts a limit o&meter reedin of 15,000

nib. Oetl bin&r a proof of IIIIWUT. pmdinp .udltiq m-d r..lf~ th. odcnrter wd receipt

of th. In.w.nc. IO C.fd.

#2 Apr-13 7.k" a.? for OdQlt.r .Ldit r.q‘ir.d bv Cc.&"",' "ithin m. mmth of initlatlm Of I".w.nc. on

SW. 0dmot.r reed8 3,700 ml,"

Aqpr-2‘ R.celm th. Im.ury~c. 10 C.rd, blv., It.ud Apr.20 by CqwY

. . . . . . . . . . . . . . . . . . . . 1. . . . . . . . . . . . . .

1. . . . . . . . . . . . . . . . . . . . . .

&ts c&mtsr audit don. on

. .

. . . . . . . . . . . . . . . . . . . . . . . . . . . ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. .

2YO POLICI "EAR 1992

r.b-28 R.c.iw. www.1 yrpllc.tion. bill ulth cwrmt c.r-mlie md car-y..r ran, end notlc. the

conpml cda9et.r udit is du in Mm*.

# "ar-07 lskn c.r for cdm.t.r .dt prforn*d by Cawanq-.winr.d, lic.n..d, @lit 9~09.. Mmeter

reds 13,000 ml!.., *Iich 9.~2. rqwt. to Cnpny

war-01 SW& Colrpny praaful for cw-wr ch.rpn plw 10,000 c.? nlln of pr.t~rim. which increas..

th. mll.w. limit frm 15,OW t. 25,000, wd cm~+l.ted rmw.L ~lic.tim

114 ~a,~,‘ R.c.i~es th. inrur.w. IO Cerd, belou, isrd by Cm,mv "m-10

. . I.. . . . . . . . . .* ,. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . I

3Ro POL*n ".~ .,pbj . . . . . . . . . . . . . . . . . . ., . . . . . . . . . . . . . . . . . . . . . . . . . . .

. .

#&r-Or koceivn rmwd .plfe.tfm, bilL with currmt c.r-mile and car-y..r r.t.., wd raticc th.t

Cnpny odQlter edit Is he in L(.CCh

“m-19 Semi. corpny prriu

for e.r-y..r ch.r9.. d co"@.t.d r-1 .~lie.tim <rays for ID

xbditiaul milewe)

15 "ear-23 ,.k" 5.r for cdm.1.r check ~rf0rm.d by COI*Yn*-q,dnt.d, Ile.n..d, Palic 9.r.9.. Cdc.,.tcr

read8 21,ooo mim

Appr-03 Rsei~n th. Itw,r.w. ID C.rd, kdw, irtued "w-50

. .I... ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..I. .

. . . ,. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .a.

Jul-OS terds pajmnt for 4,000 atMit,on, car nil" Of prot.ctim .t currant lW3 csr-rail. wt..,

uh(ch iocr..... the @iI.e~. 1Wt from 25,CuYl t. zP.WO nil..

116 Jul.12 Rrc.iv.8 th. lnwrmc. IO W-d, belo*. i..ud Jut.C3

. .I... . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,. . . . . .

. . . . ., . . . . . . . . . . . . ,. :. . . . . . . . . . . . . ,. . . . .

. . . . . . . . . . . . . ..I.

Ott-05 Suds pmmnt for 4,ORO dditiciul c.r miln of prowctim .t cwrmt 1993 err-mile r.t**,

which Incr..m th.

il.w~. Iiait frm 29,000 t. IJ.wO mile.

Ott-17 I.~.iv.. the 110"r.n~. IO Urd, beLow, il.ud I*,.l,

. . . . . ., . . . . . . . . . . . . . . . . . . ,, . . . . . .I . . . . . . . . . . . . . . . . . . . . . .

21,000 IUI-25-93

47" POLIC" YEPi

. . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . .

,994

F.b-27 IHI~WS r.,w,.l qqAic.tim, bit\ with cwrmt c.r.wl. .W CW-y..r ,.t.s, and notics th.t

C- cdmmter edit is dm in IWrsh

y8 UP05 7.k.. c.r for -t.r check p.rfom.4 b" C-.w,mnt.d. 1ic.m.d. pubtIc ,,.rao.. 0,met.r

r.ada 31,ow miles

Mar-13 SW& Crrpny prrim for CW-yew ch.rgn plu P.ooO c.r mil.. of protcetfon. which for.....

th. niIo.9. limit frca 33,LWO to 42,000, .nd cq1.t.d r.nw.1 ypllc.tim

W II.r-24 Receiwa th. Irrurmc. ID C.rd, b.,w, I.su.d M.r-20

.I... . . . . . . . . . . . . . . . . . . ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . .

I10 s.p-20 err I, Iold.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. . . . . . . . . . . . . . . . .

Capy wiv.. @yric.L 0hmt.r Wdit for rrfwd v*rific.tlm bK.u* a.,* ir to

. 1ic.Pa.d d..t.r. ccqnny .a.pt. o&.wr.r redin Of 37,000 r.cor6.d on *r.n*f.r tit,..

which i. ffld dth th. .t.t., t. c.lsu1.t. r.tum pmiu f.r uxwd mil.. of in.ur.nc..

0.x-i

Receives refund far 5,000 ailn (

= 42,OOU - 37,000) at 1994 ret* plid Mar-13.

W-20-94 ml' [ml* 31,OM

IUR-05-94

318

II. Example operation

The policyholder buys more miles of protection at the beginning of the

2nd

policy year, but does not buy any at the beginning of the 3rd policy year. Twice

during the 3rd year, however, the policyholder buys more miles of protection at the

1993 mile rates in effect.

At the time of the 1994 renewal for the 4th policy year, the policyholder buys

more miles of protection in expectation of continued higher car use. The car is sold

in September, however, leaving the policyholder with a premium credit for

the

unused miles of protection This credit can be applied to premium for

another car

on the policy, used to buy miles of protection for a replacement car (at a different

mile rate if coverages change), or refunded to the policyholder.

The number of transactions, in general, differs between the mile/year system

and the current year system. Although the physical audit required by the’test system

is a mandatory annual transaction, the number of payment transactions can be

decreased or increased according to the circumstances of the policyholder. In

Exhibit E, this difference is assessed by comparing transactions for the two systems

in the 2nd and 3rd policy years for the example car. In the 2nd policy year payments

are as large and infrequent as possible under both systems to minimize the number

of transactions. During the 3rd policy year, however, smaller and more frequent

mileage purchases are compared with a typical extended-payment plan for year-rate

premiums9

Under the mile system the number and size of payments chosen by a

policyholder are constrained only by company rules and charges, such as a one

thousand mile minimum purchase and a $3 transaction fee.1°

9. This

paper

uses the current State Farm system, detimed in the company’s 1992 Pennsylvania

manual of

rates

and rules, as typical. The

State

Farm group is the laqest in Pennsylvania with nearly

1.3 million vehicles insured in 1989 (20.2% of market). PWN. INS. DEPT, I’&&

Prrssenger M&v

Vehicle Single Ctiw Shdy

for

[Philadelphia),

(1991) Exhibit 3.

10. The question may arise as to the permissibility of hoarding a large amount of prepaid

mileage

in anticipation of a sharp rise in mile rates with inflation in medicat and automobile repair

costs. The company may put rcatrictions on the amount and may also choose to limit mileage purchase

amounts during the time following fling of proposed new rates with the Insurance Commissioaer-

when increases become public--until the new rates take effect. The time between filing and

implementation is generally at least 60 days, as set by the Motor Vehicle Insurance Rate Review Act,

75 Pa.C.S. Sec. 2003.

On the other hand, insurers may simply zero everything out at policy renewal time. That is, credit

Ihe insured for the dollar amount actually spent on miles not driven, then charge the rates in effect on

the renewal date for the miles to be covered by the renewal policy.

Transaction fees charged for installment payments arc currently $2 by State Farm and $3 by

Geico.

319

ALID--MILE/YEAR SWrEh4

II. Example operation ( 11

EXHIBIT E

Mile/Year and Year Transactions Compared

lINIW ANNUAL TRANSACTIDRS

MILE/YEAR SYSTEM - 2nd Policy Year in Exhibit B ECUIVALENT IN CURRENT YEAR SYSTEM

Awl c&meter audit

(Mileege Limit/policy year ID Card issued)

I I I

I I

Returns polfcy yew reapplication with praafun )UR

Returns

policy yew reapplication

xc pa-t for w year charges plus praniun for

I I

and pays 1st semiamusl premium

mileage roeedcd at mile rates

(l/2 year ID Card issued)

SEP Paya 21-d seal-1 pramfun

(half-year ID Card fswed)

EXTENDED PAYMENTS TRANSACTIM(S*

HlLE/YEAR SYSTEM - 3rd Policy Year in Exhibit B

EWIVALENT 1N CURRENT YEAR SYSTEM'

MAR

I I

An-a&l odonrtcr audit (ID card fsswd with

I#5 odomter mile and nxnth limits to protection)

I I

MAR

Returns policy year reafqalicatfon with any

K4R Returns policy year reapplication

yew charger (Pays rm prmiun for car miles

at-d pays 50% of 1st semiewal

because sufficient uused milewe Left tram

premium

2nd policy year>

(112 year ID Card issued)

JUL Pays for additiawl miles (ID card issued with HAY Pays balance of 1st semieM premium

#6 cdmetcr mile and mmth limits to protection)

I I

OCT

Pays for additfceal miles (ID card issued with SEP

t7 c&meter mite and month limits to protection)

I I

Pays 50% of 2nd remfsnn premium

(l/2 year ID Card issued)

I I

NOV

I I

Pays balance of 2nd SemiaM premium

State Fan" IlanUL Ru(C 104 o"

Rewual of Policy

320

12 [ AUDITED-MILE/YEAR SYVEM

III. Odometer auditing

III. Q&mete

. .

r audrUllg

Regular company audits are essential to the integrity of a mile-rate auto insurance

system.ll In parallel with the mechanical breakdown insurance provisions

reproduced in Section II above, the test system’s policy provisions on odometers are:

The policyholder must submit each car covered by this policy for an

odometer inspection and reading by the Company or its contractor when

first insured by the Company, and thereafter annually prior to the end of

each policy year. The Company may cancel or refuse to renew the policy if

the odometer inspection requirements are not met.

Driving coverages for any car under this policy are automatically void and

afford no protection if the car’s odometer:

Registers more miles than the limit paid

2) Has stopped and the Company has not been notified before further

use

3) Has its Company seals broken or tampered with

4) Has been altered in any way that changes the calibrated operation.

Since the purpose of these provisions is to assure that the company receives

advance payment for all of the on-road insurance protection it provides, monitoring

compliance is the primary function of periodic odometer auditing.

An important secondary purpose of auditing is to provide accurate exposure

data for ratemaking. Through overall and class-specific aggregations of actual,

individual averagel*

car mileages over given time periods (month and year), it is

possible to relate miles of exposure to the claims aggregated in the same classes and

during the same time periods. This would produce the necessary per-mile claim

rates (claims per mile) for making cents-per-mile premium rates.U

Odometer audits are dated, independent (non-policyholder) odometer readings

(the “audit points” of Exhibit B, supra, page 7). Two kinds of audits are used in

the test system: physical and documentary.

A physical audit is performed on the car at the direction of the company by

11. ‘Audit to examine with intent to verify.” Webster’s New Collegiate Dictionary (1980).

12. The audit lines connecting the audits for individual cars (Exhibit 8) represent the average

daily mileage of the car for the interval between audits.

13. It is worth noting that claims data on a per-mile basis would for the first time allow class and

territory cost comparisons normalized to a common statistical basis that quantifks exposure.

321

AUXIED-MIL@~Y~ARSY~~~M

III. Odometer auditing 1 13

employees or vendors, and includes inspection of seals, calibration, and reading.14

Odometers are sealed for the detection of tampering and the initial audit includes

application of tamper-evident seals. (Serialized one-way adhesive seals witness the

integrity of cable connections and of the dashboard mounting of the meter itself.)15

Documentary audits are reviews of the odometer readings that are performed

through transactions between policyholders and 3rd parties, e.g., the readings made

at the transfer of ownership and attested to by both buyers and sellers on the title

certificates.16 There are a number of other transactions involving automobiles that

provide information for auditing. For example, the Pennsylvania consumer code

requires garages to record odometer readings on work orders when cars are

accepted for repairs and again when they are returned to owners.17

14. Inspection of

cars applying

for physical damage (Collision and Comprehensive) coverage is

required by law in New York and California, and is under consideration in other states. In an opinion

survey of car owners, 83% of respondents would be “very or somewhat willing” to take their cars to an

insurer for inspecting and photographing when takiig out a policy.

INS. RESEARCH COUNCIL. i-bbk

Attitude Monitor 1991,

page 12.

Technical columnist Armando Castehii, who is an insurance educator and New Jersey broker,

noted that the inspection of cars by insurance companies (“undetwriting report”) is ‘something almost

all companies have discontinued with the predictable results that (1) insureds lie about the use of their

cars and who uses them, (2) producers lie about their clients to ‘low-ball’ premiums, and (3) the

insurance companies overcharge the large majority of policyholders to make up for the sloppiness of

the system.” INS.

ADVOCATI?,

May 4 1991.

In opposition to New Jersey legislation mandating company inspections of newly covered cars for

physical damage coverage, Allstate Insurance estimated the cost at SU per car. The maximum charge

for an emissions inspection ln Pennsylvania is S8. 75 Pa.C.S. Sec. 4706 (d).

15. Tamper-evident, serial-numbered company seals made of flexible

film

face stock are

attached with strong adhesive to the ferrules and casing at either end of the odometer cable.

Any

attempt to turn a ferrule to detach the odometer cable visibly tears the face stock.

The cost of these seals in quantity would be 2

cents

to .5 cents each, according to Valley Forge

Tape and Label Co., Exton Pennsylvania, May 19!X?, for a total at of 15 cents per car at most for three

seals.

16. Title certificate odometer readings gain further validity through state law for controlling

odometer fraud in car sales:

75 Pa.C.S. Sec. 1105 (c) Title transfer odometer readmgs--The department shall compare

the odometer reading of the vehicle each time a certificate of title is transferred and

ascertain the reported mileage against the most recent previously reported mileage for the

vehicle,

Such a review presumably would screen for a decrease, or an abnormally low increase, in the car’s

reported mileage from a previous transfer.

A PennDOT pamphlet Odometer

Roll

Backs distributed by the Office of Attorney General,

Bureau of Consumer Protection gives instructions if a car buyer suspects fraud to write the Bureau of

Motor Vehicle & Licensing Information Sales for “a photocopy listing previous title holders, their

addresses, and in some cases, the car’s mileage at the time of sale if PennDOT has such information

available.” Pub. 160 (2-g.

Dealers “shah retain for four years each odometer mileage statement which he receives. He shall

also retain for four years a photostat, carbon or other facsimile copy of each odometer mileage

statement which he issues.’ 75 Pa.C.S. Sec. 7133 (a).

17. 37 Pa&de Sec. 3015. The hvo readings control unauthorized use of the car while it is in

322

1 AUDITED-MILWYSSR SYSITSI

III. Odometer auditing

As a condition of annual policy renewal, the Company requires that a physical

audit be done on each car it insures. The odometer readings and dates of these

audits are included on the insurance ID cards, Exhibit C. Policyholder convenience

is served by specifying renewal months rather than deadline dates.18 Repair garages

view the audits on behalf of insurance companies as a business opportunity.19

To secure insurance on cars just bought, a copy of the title certificate or mileage

disclosure statement from the seller provides the initial odometer reading lo the

insurance company for binding coverage until the initial company physical audit is

done, within 30 days.

IV. wrotectim

Stealing protection from auto insurance companies can be done in several ways

under the mile/year system: biasing the drive train to make the odometer read low,

resetting the odometer, and stopping the odometer. Federal and state laws against

tampering with odometers penalize these prohibited activities with fines and jaiL20

This section describes procedures used in the mile/year test system to control theft

of insurance.

(continued)

the custody of the garage.

18. With most or nearly all of premium charged at mile rates, there is relatively little per-year

premium put at risk of non payment in the mile/year system through extending the effective due date

from the anniversary within the month to the end of the month.

19. Sworn testimony by an -et of a garage licensed to do state inspections. The charge for

odometer seal@, calibration, reading, and reporting to an insurance company was estimated at less

than $10. Reproduced Record, w NOW v. Ins. D@, of Pennsvlvanig Commonwealth

Court, (No. 1276

CD. 1987 and No. 276 C.D. 1988) at 24Jla.

20. Federal Odometer Ad of 1wZ (15 U.S.C. Sections 1901, 1981-1991); Pennsylvania Vehicle

Code

(75

Pa.C.S. Sections 7131-7159).

Section 7132

of the Pennsylvania vehicle code 75 Pa.C.S. states:

Prohibited activities relating to odometers.

(a) Devices causing improper odometer reading.--No person shall advertise for sale, sell,

use or install, or cause to be

hstahd,

any dcdce which causes an odometer lo register any

mileage other than the true mileage driven which is that mileage driven by the vehicle as

registered by the odometer within the manufacturer’s designed tolerance.

@) Change of odometer reading.-No person shall disconnect, reset or alter, or caue to be

disconnected,

reset or altered, the odometer of any motor vehicle with intent to change the

number of miles indicated on the odometer.

defraud, operate a motor vehicle on any street or highway knowing that the odometer of that

vehicle is disconnected or non-functional.

Summaries

of twenty years of odometer-fraud case law, as well as federal and state statutes and

regulations, are contained in:

NAT% CONSUMER

L. CENIFR,

Odometer

Low, (19%) and

Odometer Law

Cumulative Supplement (1991).

323

ALJDSIED-~~~LE~YEAR SYSKM

Iv; Stealinginszuanceprotection ] 15

Surprisingly, keeping the odometer cable unhooked for much of the time, and

resealing it with a stolen or counterfeit seal before the next company audit is

unlikely to result in significant theft of insurance protection. According to the policy

provisions of the test system, a car operating with the odometer disconnected is

simply being driven without insurance: no mile-rate premium is being earned and no

protection is being provided.*l Even if disconnecting and reconnecting of the

odometer should go unnoticed, there is no theft of insurance protection. If the

odometer is sealed and operating, premium is being paid for the protection

provided.

Low-reading odometer.

Odometers can be made to read less than the actual

mileage driven without breaking any seals by increasing tire sizes or decreasing axle

ratios. The premium savings, however, would appear to be small compared with the

effort, the risk of severe federal and state penalties for odometer fraud, and the

adverse effect on the car’s operating characteristics.” The policyholder would have

to switch from standard to larger wheels or a higher axle ratio after the annual

audit

and calibration, and reverse the change before the next annual audit. To cut the

mileage readings about 10% from actual distance traveled would require a 2-inch

increase in tire diameter. The car’s insurer would be defrauded thereby of $100

for

every $1,000 worth of protection provided. If the use of oversized tires to steal

insurance protection should ever become a problem, however, it could be controlled

by recording the tire size on the insurance ID card with the warning that use of a

different tire size without a recalibration voids the on-the-road protection.

Switching drive axle gear ratios before and after each annual audit would require

even more effort than changing tire sizes, but can be controlled if necessary by

application of a single seal to the axle housing.

21. Protection would be received without premium payment only for accidents after which the

policyholder was able to reconnect and seal the odometer without beiig observed.

Although using a car with the odometer cable unhooked is not stealing insurance protection

because protection does not exist, use of a car in thii condition violates the compulsory insurance law,

wbicb is the subja of section VI, in&.

22. The manufactured design optimizes handling performance with given tire sizes and axle

ratios.

A “slow’ speedometer and low-reading odometer are inconveniences, at least whenever speed

limits and map distances are of interest.

1 AUDITED-MILE/YEAR SYSIEM

Iv: Stealing insurance protection

EXHIBIT F

Reset Odometer Example

35,000~-

C&S

40,000-

EXHIBIT C

Stopped Odometer Example

35,000--

C&S

nlleagc

30,000--

25.000~-

bAudit

2O.0007

WiXi;0

0.25

Pal&T

0.75 1

par

325

AUDITED-MILE/YEW SY-

N Stealing insuranceprotection 1 17

Some years ago, when odometers had metal gears that could stand high speeds,

dishonest car dealers ran odometers rapidly in reverse by unhooking the cable and

attaching a high-speed drill. For several decades, however, plastic gears have been

standard in odometers. Such gears are durable in service because even the most

rapidly revolving one-tenths digit turns very slowly--one rpm at sixty mph. These

gears heat up and break, however, when attempts are made to back up the

odometer rapidly with an electric drill.

Professional insurance thett by resetting the odometer.

Criminal experts can

turn back odometer tumblers with a thin pick. Charging $30 to $50 per two-minute

job, the criminal can remove several tens of thousands of miles from the odometers

of late model high mileage fleet or lease cars to increase resale value an average

%1500.” Given the fact that insurance rates for full coverage of relatively expensive

cars in high-rated territories can exceed $0.10 a mile, a policyholder could defraud

the insurance company of $1000 in premium each year by paying $50 to have the ten

thousands tumbler rotated back one digit.”

To counter such major odometer fraud in the mile/year system, odometer

auditors and inspectors look for telltale signs of rollback tampering that can lead to

convictions of the policyholders responsible. The following example shows how

annual audits can control and constrain this kind of theft of insurance protection.

Exhibit F.

In the initial audit, the odometer of a hypothetical insurance thiefs

car reads 20,000 miles. Although the car will be driven 20,000 miles during the

coming year, the thief plans to have the odometer rolled back 10,000 miles and

therefore only buys ten thousand miles of protection. To keep from producing an

23. Connie McNamara,

Odometer cheafs gem a Jot of mileage, officials say,

Harrisburg Sunday

Patriot News, June 28,1981.

24. Patrons of meter-tampering criminals take considerable risk. For example, when a New

York lock picker who reduced gas and electric meter readings by half was caught, he cooperated with

the prosecution of his clients--by wearing a hidden tape recorder while beiig paid--in exchange for

reduced punishment. In addition to facing criminal prosecution, hi customers are to pay restitution

for tens of thousands

dollars of stolen gas and electricity. N.Y. Ties Sep. 27, 1985, p.2 45

L.I.

Businesses Accused of Cheating UtiJiy Meters,

More recently in the Philadelphia region, criminal experts in altering electric meters to

underrecord actual consumption up to 58% were involved in tampering with more than 50 meters for

Bucks County homes and businesses. Phila. Inquirer, Jan 21, 1992, B2 73ird person faces

charges in

Bucks meter tampetings.

On the other hand, meter-tampering criminals can be caught with the help of clients hoping to

escape punishment.

56 had

7VTVboxe.r

rigged

chear,

police soy Patriot-News (Harrisburg, Pa.) May 8,

1992. Facing $303 fmes and 3 months in jail, customers who paid $35 to $100 to two criminal experts to

alter pay-TV meters, are cooperating in their apprehension.

326

1 AUDIIED-MILE/YEAR SY~XEM

IK Stealing insurance protection

odometer reading that is less than the initial audit (20,000 miles) and thus voiding

protection, however, resetting the ten-thousands digit has to be delayed until after

30,000 miles is passed at mid year. The rollback is done in the example at 9 months

(0.75 year) to reduce the reading from 35,000 miles to 25,000 miles. Apparently

unknown to the thief, however, there was no insurance protection for the last 5,000

miles before the rollback because these miles exceeded the 30,000 prepaid-miles

limit. Therefore, the amount of protection actually stolen in the example is not

10,000 miles, but only 5,000 miles, Exhibit F.

A more sophisticated policyholder than the insurance thief described above

could steal several thousand dollars of auto insurance protection annually on a fully

covered expensive car in a relatively high-rated area.5 It is significant that auto

insurance companies and their honest policyholders are not the only victims of

odometer rollbacks: providers of car warranties and mechanical breakdown

insurance, and buyers of used cars are defrauded as well. Widespread public

awareness of the seriousness of odometer tampering has developed from news

reports on a decade of prosecution of strong federal and state odometer fraud

statutes.%

Amateur insurance theft with a stopped odometer.

Odometer failure, whether

spontaneous or induced, presents an opportunity for theft of insurance protection.27

According to policy language, insurance coverage is void while the odometer is

stopped unless the company is notified. Nevertheless, to protect the validity of a

claim, a policyholder can always assert that the odometer had failed just before the

accident or was broken as a result of it. Test system procedures are designed to

control such theft.

25. Apparently some sophisticated car owners violate mandatory insttrance requirements.

State

police testimony pointed out that “It is just as probable to find a counterfeit inspection sticker or false

insurance identification card associated with a Mercedes Benz as a Chevrolet. The motivation to avoid

the law is economic, whether the owner or lessor can afford the insurance

not.”

House Insurance

Committee hearing April 25,1991, Tr. 117. Under the mile/year system, federal and state odometer

anti-tampering law provide additional strong sanctions against cheating.

26. Although the focus has been on professional thieves

who

tamper

with

tens to hundreds of

odometers, some individuals are being punished. E.g., a Baltimore pohee lieutenant recently was

indicted for theft for

having

his own odometer rolled back to increase the car’s resale value. Baltimore

Sun, Bti.$, April 22,l992.

27. An odometer cable can be caused to wear rapidly and break by pulfing its casing into an

overly tight

curve.

Society of Automotive. Engineers specifications put the minimum radius of

curvature at six inches. SAE J678 Dee 88.

In addition to the inconveniences of not having a speedometer and odometer (note 22, SUpM),

cars built after 1980 will not run efftciently with the odometer cable broken because it provides a signal

to the ignition/fuel-injection computer. R, Morse, Chief of OC:Teter Fraud Staff, USDOT, at the

327

AUDITE?D-MILE/YEAR SYm

N. Stealing inmranceprotecfion ] 19

Should an odometer stop, policyholders are instructed on the ID card to stop

driving and to telephone the Company in order to maintain coverage, Exhibit C

(supra, page 8). The company responds with a confirmation code, a time limit

for repair, and any other instructions or limitations appropriate for the situation.z

Before any seal is broken to replace the odometer or its cable, a mandatory first

step is to have a physical audit to get the reading and inspect the seals. This audit

comes between annual audits and shows the average driving before and after the

odometer failure. An unexplained and marked difference in the averages indicates

possible premium fraud, which can be further investigated. A hypothetical example

of protection theft shows how the audits are used.

Exhibit

G. Actual use of a hypothetical car is 20,000 miles during a policy year.

At the begiting of the year the odometer reads 20,000 miles and 10,000 miles of

insurance protection is bought for the coming year. The odometer cable breaks (or

is broken) at 0.25 year with a reading of 25,000 miles, but is not reported to the

Company until 10,000 miles more have been driven in the next half year. If an

accident happens during this 10,000 mile period, the policyholder can claim

coverage with the excuse that the odometer “just broke.” With the excuse sustained,

the Company has provided coverage without collecting premium for 10,000 miles.

The excuse can be challenged, however, after the end of the policy year with audit

information.

At the repair audit the odometer shows the same 25,000 mile reading it did six

months earlier when it stopped. A “miles-driven” line joining the initial and repair

audits would show 5,000 miles in 9 months, which is a rate of 6,670 miles per year,

Exhibit G. An increase of 5,000 miles on the odometer in the next three months

between the repair audit and next annual audit, however, indicates the true driving

rate is 20,000 miles per year. Even though such an apparent strong change in car

use might be insufficient to prove theft of insurance protection, it nonetheless alerts

the test system to get more evidence on the actual amount of driving done. Another

odometer failure in a following year would be even more suspicious. Stealing

insurance protection through odometer tampering is a risky way for policyholders to

try to save money.29

(continued)

11th Ann. Con& National Odometer Enforcement Association, August 12,lWl.

28. If the failure happens oo a trip, the insurance company specifies that coverage stays in effect

over the route for completion of the trip. A value for the unrecorded miles of exposure is derived from

the route distance.

29. Tbii kind of amateur theft is akin to that in which householders steal gas by running a bicycle

inner tube around the meter or steal eledricity by removing the meter and inserting spoons or forks

328

20 [ AUDITED-MILE/YEAR SYs~wl

V: Nondrivingperiods

V. Premium charees for nondrivine periods

If the possibility that policyholders may be able to steal insurance protection is

cause for concern, then the possibility that companies may be able to charge

premium for driving coverages during nondriving periods must be an equally serious

concern This section examines various administrative provisions made for periods

of nondriving and the effects they have on premiums. In the current year system,

rate manual rules allow interruptions of insurance protection and premium charges

during nondriving periods, such as:

SUSPENSION OF COVERAGE, SEASONAL USE,, WITHDRAWAL FROM USE

A. Suspension of Coverage

Coverage afforded under a policy insuring a motor vehicle may be

suspended during the time the vehicle is withdrawn from service. The

coverages suspended afford no protection under the policy during the

ptr$cJ of suspension.

The continuation of certain coverages during the period of suspension

may be desirable; e.g. comprehensive [coverage for nondriving losses].

*?.I*

Premium credits on suspended coverages will be computed pro rata for

the period of suspension.

General Rule 106 A, State Farm Mutual Auto. Ins. Co. (emphasis added)

The administrative problems of this and several other ways that the year system

currently handles nondriving periods can be compared to the automatic response of

the mile/year system to nondriving periods:

o When there is a period of “withdrawal from use” accompanied by “suspension of

coverage,” as the above manual rule provides, prepaid premium for the non-

covered period is refunded to the individual policyholder.

Example: Hypothetical car considered below in 2nd policy year.

When many policyholders in an identifiable group are not driving, a portion of

prepaid premium is refunded uniformly to all policyholders in the group.

Example: Policyholders who served in the Middle East armed forces in 1990

and 1991, were refunded prepaid premium for a fraction of the time they were

(continued)

across the connections. Control on such theft includes monitoring consumption for anomalous

pattam aiid abrupt changes. P. VALE&TIN@, On Ihe

Trail

P-Hungry Thieves,

Wash.

Post, April

6,1991 at El.

,329

AUDIZD-MUE/YIXR SYSIZM

K Nondrit6ngperiod.s 1 21

overseas, despite the fact that companies had no way of knowing whether the

cars actually were withdrawn from use or even driven less in the owner’s

absence.30

When economic recession triggers a reduction in claims because many

policyholders drive less, insurance companies provide uniform refunds

(“dividends”) to all policyholders without distinguishing whose cars actually were

driven less.3l

Example: Owing to a decrease in accident claims since the current recession

started in mid-1990, State Farm made statewide refunds in 12 states reaching 20

percent of semiannual premium.”

When individuals interrupt driving for periods of time, in most cases no refund

of prepaid premium occurs at all.

Example: Hypothetical car considered below in 3rd and 4th policy years.

Mile/v-

@ Only the miles driven by each car, as recorded on its odometer, determines

individual premium consumption and obviates the need fgr group refunds.

With miles prepaid, the car is fully insured whenever driving is resumed after a

nondriving period. Therefore, no administrative costs are incurred for

suspending and reinstating insurance.

Example: Hypothetical car considered below over four policy years.

30. Geicn Insurance Company’s ‘Desert Shield dividend consisted of a 25% premium credit for

the period on active duty in the Middle East in 1990 as certified by a superior officer. Source: Geico

forms (P-294a % P-295)

sent

to policyholders on

request.

This refund appears to assume that the estimated 16,ooO Geico-insured cars eligible were on

average driven 75

percent as much as usual while their owners were overseas. (Le.,

for every four

months that the policyholder was overseas, one month’s prepaid premium was refunded.) A Geico

official’s reply lo a reporter’s que&ion about an equivalent “Desert Storm” dividend for 1991, however,

suggests that the refunds were not based on a cost-savings estimate, but instead on the budget for

public relations: “For this year, we don’t have an idea as to whether the company is making a profit or

not, so wc can’t make a deeision yet.” NATL UNDERWRITER, Geico To

Pay Desert Shield Aura

Dividend, Feb. 4,1991,

31. Speculation that “a lot of cars are up on blocks” by the State Farm vice president for

Pennsylvania was cited by the president of the Insurance Federation of Pennsylvania to dramatize the

marked drop in claims following the July 1990 onset of the current recession. Transcript of testimony

before the House Insurance Committee November 14,1991, page 112.

32. J. Commerce, St& Farm to

Refund

Millions in Car Premiums, Dec. 18, 1991, at 9A. State

Farm’s reported explanation was that “claim costs were less than anticipated.”

At rate hearings and in

other forums, however, ofticials of State Farm and other auto insurance companies directly attributed a

drop in claims lo a rewssion-related decrease in driving and a rise in gasoline prices. In Pennsylvania

where State Farm had litigated strongly against premium reductions mandated by Act 6 of 1990, il

seems dear that the company did not take the nearly 6 percent increase in July 1991 authorized by the

law bccauae of the drop in claims caused by the recession and gasoline price rise. The effect of the

recession

on claims is evidently continuing because the State Farm Pennsylvania rate manual pages

effective May l&l992 show only moderate changes (about -6% to t 7% depending on coverage and

territory).

330

P$1200

:, 9800

mea

EXHIBIT H

Example: Effects of NonDriving Periods on Premiums

YEAR SYSTEM

MILEIYEAR SYSTEM

omcar,$soopyWrBte

$2400

Gm car. So.05 per mile rate

. . . . . . . . . . . . . . . . . . . . . . . . . __ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

_ ____._____.__.___.___.___..__.~_____.,,__._

AUDITED-MILE/YEAR SYSI’EM

V. Nondrivingperiodr ) 23

Exhibit H.

A hypothetical example involving one car compares the response of

each system to a decrease in driving from 12,000 miles during the first policy year to

8,000 miles a year for the next three years. Graphs of premium vs. time (in policy

years) for each system show the relationship between premium prepaid and

premium consumed throughout the year.

Prepaid premium is represented by a stepped plot in the graph for each system.

(Compare with the graph of odometer readings vs. time for the mile/year system in

Exhibit B, supru, page 7). The vertical segments are payments made shortly

before the start of each policy year. 33 The horizontal segments represent the

passage of time between additional premium payments.

Year system.

Premium for all coverages is earned proportionally with time, as

shown by the inclined, lower line. By the end of the first policy year, all of the $600

has been earned. At the beginning of the second policy year, $600 is again paid in

advance. During this year, however, the car is not driven for four months (1/3rd

year) and insurance coverage is suspended for this period.

The earned-premium plot is flat with time during the 1/3rd year period of

suspension because no insurance is in force. When coverage is reinstated at the end

of the period, the earned-premium plot resumes proportionality with time. Because

of the suspension of insurance, however, 1/3rd ($200) of the prepaid premium

remains unearned at the end of the 2nd policy year. This amount is credited against

the $600 premium due for the 3rd policy year so that only $400 is paid.

In the third and fourth policy years, driving remains at 8,000 miles annually, but

the periods of non driving are spread throughout the year. The periods are either

shorter than the Company allows for suspending coverage, or the car may be needed

occasionally during the longer periods and there would not be time to reinstate

insurance and get the license plate back from PennDOT.% Therefore, even though

mileage has not increased from the second policy year when $400 was paid for

insurance, the policyholder pays $600 per year because coverage cannot be

33. Payment is shown 20 days before the beginning of the policy year.

34. Given the fact that suspension of insurance benefits only the policyholder while reducing

both premium income and commissions for the company and agent with no offsetting transaction

charges, it is not surprising that few policyholders know that coverage can be snspended during periods

of nondriving. Sworn testimony by actuaries in Pennsvlvania NOW v. State Farm indicates reluctance

of companies to suspend coverage, because “someone has to make a notation on the policy and

recalculate the premium.” 6 3. INS REG. 243,282 (1988).

332

1 AUDITED-MILE/YEAR SYSI’EM

I/ Nondriving periods

suspended. The policyholder’s cost per mile of protection increases by 50 percent

from 5 cents to 7.5 cents for the final two years solely because the year system makes

it no longer practical or even possible to suspend coverage during non-driving

periods, as it was in the 2nd policy year.35

Mile/year system.

At an assumed rate of S cents a mile for coverage of the

hypothetical car, the test system produces a $600 premium for 12,000 miles in the

1st policy year and a $400 premium for 8,000 miles in the 2nd policy year, as did the

current year system. The annual odometer audits, connected by lines, show the

increases in miles recorded by the car’s odometer as the miles of premium earned.

In the 3rd and 4th policy years, however, when driving continues at 8,000 miles a

year, the mile/year system premiums continue to be $400 at mile rates, equal to the

$400 premium at the $600 per year rate when coverage was suspended for l/3 year.

Mile-rate premiums are $200 less than the year-rate premiums in the final 2 policy

years simply because, as noted above, the year-rate coverage could not be

suspended for part of the year as it was in the 2nd policy year.

Quasi-suspension OF coverage produced by the year system.

Insurance

enforcement apparently makes suspension of coverage both difficult for companies

to administer and costly for policyholders to use, judging from the special provisions

companies make for suspending required coverages. For example, State Farm has

added a Pennsylvania

section

106

C to its rule 106 A (quoted above), resulting in

what appears to be a quasi-suspension of coverage.

SUSPENSION OF COVERAGE, SEASONAL USE, WITHDRAWAL FROM USE

e* zspension of Coverage

Any coverage may be suspended unless it is required by stutute to

remain in force. (See Section C for statutory coverages.)%

***t

35. Lienholders also provide an impediment to suspension of coverage,

judge from the

intervention that Geico provided for Desert Shield policyholders who stored their cars while overseas.

The Geico letter to the service. pexson accompanying the dividend certificate states: “If your car is in

storage but you’ve been told by a lien holder that full coverage must stay in effect, let us know. We’ve

intervened succesrfully in getting some lien holders to waive this requirement for our

insweds.”

36. Despite

Rule 106’s implication here that ‘statutory coverages’ means

insurance that car

owners must buy, the coverages listed in Rule 106 C include non required Combined Benefits,

Uninsured

Motor

Vehicle, and Underinsured Motor Vehicle in addition to the Bodily Injury Liability,

Property

Damage

Liibiity, and Medical Payments coverages that are required by Pennsylvania law.

The law requires companies to make these additional coverages available, but does not require their

purchase. 75 Pa.C.S. Sec. 1791-1792

333

AUDITED-MILE/YEAR SYSIEM

I/: Nondrivingperioo!r 1 2.5

C. Premium Credit on Statutory Coverages for Vehicles Withdrawn from

Use

During the period when a motor vehicle is withdrawn from service, the

coverages indicated below may, at the option of the insured, remain in

force at 40% of the otherwise applicable premium.37 The period of

withdrawal must be for at least 30 days, and the insured must complete

a cer#icate of withdrawal.

General Rule 106 A & C, State Farm Mutual Auto. Ins. Co. (emphasis added)

This option of keeping statutory coverages in force on a car while it is not being

driven seems to be a response to the Pennsylvania requirement that the car’s

registration card and plate must be returned to PennDoT while insurance is not in

force.% The “certificate of withdrawal” specified above is a company form to be

signed by the policyholder which states: “I hereby certify that the motor vehicle

described below will be withdrawn from use for a period of not less than 30 days

from the date indicated.” The policyholder further agrees on the form that “In

recognition of the fact that my motor vehicle will not be used, the premium for the

coverages required by law will be reduced.

“3g The question is: why is the premium

not reduced to nothing if it is believed that the car will not be used?

Why is it worth

40 percent of the premium to the owner to keep the license plate?

Why does the

company need to get 40 percent of premium to keep driving coverage on a

nondriven car?& Such an arrangement argues that the company expects the car to

be driven despite its nominal “withdrawal from service.“41

37. Apparently the basis to which the 40% of fug rate applies is not just the rate for the required

minimum amounts of protection, but for any increased protection amounts--“limits”-of each coverage

chosen as wll.

38. Although California also rquires all cars to be insured Rule 106 in the State Farm manual

on file at the California insurance department (Oct. 9.1991) dces not have a section “C with its special

provisions for required or statutory coverages. The difference may be that some enforcement

provisions, including criminal penalties and proof of insurance as a registration requirement, were

allowed to sunset in California in 1990.

NATL UrroERwRITeR, Qlunge In

Rcgufucots Wonia Calif.,

Jan. 7,1991, at 4. (Rule 106 “R” provides for “Seasonally Used Farm Trucks,” and is nearly identical for

both states.)

39. State Farm Insurance Companies form G-4658-7 Rev. 07-91 “Notification that Motor

Vehide is Withdrawn From Use.”

40. What this provision means in terms of premium ccct to policyholders may be assessed from

the example of Rxhibit H, above. During the 2nd policy year the car was not driven for four months,

and the “suspension of coverage” rule was used to get a refund of $200 on a $600 prepaid premium.

Asswning that the %OO premium was only for State Farm’s ‘statutory” conrages, State Farm would

have retained 40%, or SSO, of the refund to keep enough cowsrage so that the policyholder presumably

could keep the car registered and not have to return the license plate and registration card to

PennDOT.

334

1 AUD~X~~ILEJYEAR SYSIZM

VI. Enforcement and compliance

VI. mrance enforcement and compb

This section examines enforcement provisions of mandatory insurance under the

current

year system

and under the mile/year test system.42 In each case, the analysis

assesses the effect of these provisions on both enforcement efficiency and the ability

of the car owner to achieve compliance.

With respect to the intent of the law that all cars have minimum insurance while

they are being driven, two kinds of enforcement are involved: at-car verification of

insurance when it is certain that the car is being driven; and indirect enforcement

through the registration-license plate process when tire car may or may not be being

driven.

Direct at-car enforcement.

A random sample of cars being driven receive at-car

insurance verification by police and state officials through the happenstance of

accident investigation and citations for moving violations where verification is

generally required, and in the course of roadside checks for various reasons where

insurance verification

may

be included. 43 Under both systems, the at-car

verification is determined by the policy expiration date on the car’s insurance ID

card. Under the miZe/yeor system, full verification also requires that the odometer

reads less than the ID card’s prepaid miles limit, is operating, and shows no signs of

tampering.44

Although at-car proof of insurance is required by annual safety inspections and

provides a strong incentive to get insurance, few uninsured cars are likely to be

41. Since. the 40% charge is applied to premiums rated by territory and other class categories, it

is not just an admhktrative fee, but varies presumably to reflect different driving conditions. Under

rate regulation law, State Farm is supposed to have claim-cost statistics to just@ the charges. 40 P.S.

Set 1183(c).

42. Minimum coverage required is S15,000/S30,000 bodily injury liability, SS,OOO property

damage liability, and SS,OOO medical payments. 75 Pa.C.S. Sections 1702 and 1711.

43. Accidents, and to a large extent law enforcement stops, are inherently processes of random

sampling of the cars being driven at the time. INS. INDUSTRY COMM~IZE ON MCITOR VEHICLE

ADMIN., Guidelines

for

Compulsory Liability Insumnce

Enforcement, July

192.

The sample sire for checking insurance compliance via accidents may be approximated from the

1990 PennDoT report: 234,814 reported driver involvments gives the number of vehicles sampled,

which is about 3% of the 8.7 million registered vehicles which were checked for insurance in this way.

If traftic violation checks were veritkd on approximately the same number of cars, then the proportion

of cam randomly checked for insurance this way would be less than 10% ammaffy.

fn vehicle law eaforccment checks, police officers radio in driver’s and registration iaformation to

PennDOT computers for verification. Insurance information, although entered into the PenaDOT

system, could not be checked the same way in 1991. State Poke testimony before the Hoosc Insurance

Committee, April 25,X91, Tr. 233.

33s

AUDITED-MILE/YEAR

C+iSIEM

VI. Enforcement and compliance ) 27

identified this way. Owners of uninsured cars are able to bypass this requirement

illegally, as described by the state police.45

Under the mile/year system, the company odometer audits enhance

enforcement of insurance because direct company involvement and self-interest in

verification improves control on illicit ID cards beyond what police officials and

inspection stations can do.

Furthermore, policyholder no-shows for renewal or

final audits (dropouts) indicate that the cars have been driven uninsured without

prepaid miles of protection. A gap in protected miles would be evidenced by

odometer readings or signs of tampering even if insurance is re-initiated with

another company. Enforcement sanctions could be invoked for owning a car driven

without insurance, or for odometer tampering.

Indirect enforcement. The

logic behind indirect enforcement through the

license plate issue and revocation process is that requiring insurance on a licensed

car to be kept continuously in force assures that insurance will be in force on the car

whenever it is being driven. Specifically, the law requires that a license plate may be

retained only while insurance is in

force. 47 When insurance coverage terminates,

sanctions can be avoided by not driving the car and by returning the license plate to

PermDOT within 21 days if insurance has not been reinstated or replaced.‘@

44. Owning a car that is being driven with no insurance in force is a direct violation of the law’s

intent. It is fully sanctioned by a $300 fine and three-month revocations of the license plate and the

owner’s driving license. The restoration fee at the end of the suspension periods is SM for each license.

75 Pa.CS. Sections 1786 (d), 1786 (f), & 1960.

45. I&it inspection stickers are used to avoid the inspection altogether, or i&it insurance ID

cards or other insurance documents are used to defeat the verification process at the inspection station

Testimony to the House Insurance Committee, April 2.5, 1991, by Lt. Colonel Robert Hicks, Tr. X2-

134.

Although the law requires inspection stations to report any uninsured cars seeking inspections, it

seems that stations have little to gain by reporting the few customers unaware of the insurance

requirement for inspection. 75 P&S. Sec. 4727 (d)(2).

On the other hand, at-car information for the mile/year system is collected as part of the

inspection process because mileage readings for verified-working odometers are reported to PennDOT

and help build an audit trail for each car. (Two odometer readings are reported: “present odometer

reading” and “odometer reading on old inspection sticker.” PennDOT Inspection

Record,

form TS 431

WW.

46. The company audits may or may

not

be done in conjunction with the state safety inspections

according to individual arrangements between auto insurance companies and private inspection

stations.

47. The partial sanctions applied for violation are the three-month revocation of the license plate

followed by a

550

reissuing fee. 75 Pa.C.S. Sec. 1786 (d); Sec. 1960.

The law appears to specify the 3-months suspension of registration penalty only for cars being

driven without insurance, and not just for keeping an uninsured car’s license plate. (75 Pa.C.S. 1786

(d).) According to testimony by PennDOT, however, the Imonth registration suspension periods are

beiig applied to cases where the license plate has been kept for an uninsured car, without establishing

336

28 ) AUDITED-MILE/YEAR SYSIEM

VI. Enforcement and complkmce

Whenever insurance for a car is terminated and driving coverage is not in force,

the insurance company is required by law to notify PennDOT.4g When suspended

insurance is reinstated however, it is the policyholder who must provide the

insurance information to PennDOT. PennDOT in turn sends a sampling of such

reports to the Company for verification.50 The Company may also notify

lienholders when the car securing the loan has no collision insurance, and also when

the coverage is restored.

Under the current year system there is no reliable way to establish whether or

not a car is actually being, or has been, driven without insurance beyond scattered

information that is produced by the random sampling process during at-car

enforcement, as described above. Under the mile/year system, however, the

odometer serves as a witness to prove nondriving for the policyholder, or to prove

uninsured driving for enforcement purposes.

Compliance.

Under the current year system, premium payments are fiied costs

of car ownership with inevitable due dates. A policyholder in straitened

circumstances has no legal option but to lapse insurance, surrender the license plate,

and do without the car. If a policyholder cannot meet a deadline for a premium

payment, and cannot suspend coverage to stop or lower premiums owed, it is not

surprising that there may be little real choice but to drive illegally without

insurance.51

(continued)

that the car was actually being driven uninsured. (Douglas Tobin, House Insurance Committee

hearing transcript Oct. XI, 1991, page 168.)

48. 75 Pa.C.S. Sec. 1786 (d)(l) & (g)(2). The instruction given by PennDOT’s public

information telephone (March 1992) for interrupting insurance while the car is not being driven is to

send in the license plate (not the registration card) to PennDOT with a letter of explanation.

According to PennDOT testimony before the House Insurance Committee, returned plates are

destroyed and new plates issued

with

no charge when insurance is again in-force. Douglas Tobi, Oct.

30,1991, Tr. 169.

49. 75 Pa.C.S. Sec. 1786 (e)(2): “Obligations upon termination of financial responsibility--An

insurer who has issued a contract of motor vehicle liability insurance . . . shall no@ the department in

a timely manner.*

Companies provide notices of cancellation of insurance to PennDOT on computer tape in batches

covering a week or two of activity. NOW telephone information from Chairman’s office, Erie

Insurance Group, Feb. 1,1991.

50. PennDOT’s goal is to have companies verify the insurance information provided on

registration forms of 50 percent of Philadelphia registrants and 25 percent of registrants elsewhere.

Joint State Government Commission, ‘Insure-the-Driver Program” study pursuant to Act 6 of 1990,

Section 29 (1991), page 16. (These goals would require about 2 million company verifications

annuaIly.)

51. Under the current year system, considerable administrative effort is expended in sending

biiig and nonpayment-cancellation notices to installment payers who

are

on very tight budgets. This

is especially true for assigned-risk plans in high-rated territories, An active and ongoing discussion

337

AUDITED-MILE/YEAR SYSITM

VT. Enforcementandcompliance 1 29

Under the mile/year system, in constrast, the only fixed costs for keeping

mandated insurance in force are the cost of the annual odometer audit plus any

administrative charges (about $20 total per year). Premium for driving coverages is

earned by the company only while the car is being driven. To the extent that the

mile/year system reduces mid-term lapses and eliminates suspension of coverage

transactions, there is a comparable reduction of state and company administrative

expense for recalling and reissuing license plates, with attendant insurance

verifications. The mile/year system makes required insurance an operating cost and

promotes compliance by providing the public with a means of direct individual

control over the amount and timing of premium payments.

In.surance Proiect

National O&.izatioo for Women

1000 16th Street, NW, Suite 700

Washington, DC 2Kt36 (202) 331-065

June22 1992 #486

(continued)

continues in the weekly

INS. ADVOCATE

column views of o

Sforefront

Broker by Michael Carbajal in

New York City. April 18, 1992, for example, on the size of the premium deposit needed and the

admiitrative expense to companies and agents. In Philadelphia the minimum cost assigned risk plan

premium is $696 annually, paid $211 down (30%) and S!B per month for five months, as advertised for

months in Tbe Review (Chronicle, S. Pbila.), e.~, April 16, 1992, at C14. In the past, eighty percent of

assigned risk car owners did not pay alI of the iwtallmeats, which are billed, and were canceled in their

first year. JOINT

STATE

GOVT COMM.

staff

analysis, Insure-the-Dtiver Program for Philadelphia

(1991)

at 15.

338