law and economics of consumer finance

7/31/2019 Law and Economics of Consumer Finance

1/52


2/52

The Law an d Econom ics of Consum er Finan ce

Richard H ynes1 & Eric A. Posner2

February 20, 2001

Abstract: This survey of the law and economics of consum er finance d iscusses economic mod els

of consumer lend ing, and evaluates the major consumer finance laws in light of them . We focus

on usury law s, restrictions on creditor rem edies such as the ban on expansive security interests,

bankruptcy law, limitations on third -party defenses such as the holder in du e course doctrine,

information disclosure rules including the Truth in Lending Act, and anti-discrimination law.

We also discuss the em pirical literatur e.

Introduction

The law regulates consum er credit transactions mu ch more h eavily than n on-

credit transactions like the cash sale of a computer . Nearly anyone can sell computer s

to the pu blic, but the creditor bank , finan ce company, pawnshop, credit card issuer 3

is heavily regu lated by federal and state agencies: licensed, inspected, and less so now

than in the recent past circum scribed by geograp hic, market, and p roduct restrictions.

The computer seller may offer any cash contract acceptable to the market, subject to

some light restrictions imposed by federal and state law. The cred itor may not choose a

price that exceeds the relevant usu ry ceiling, or remed ial terms tha t are considered too

burdensome by the law . Many kind s of attractive collateral like household good s

will not be used, because the security interest wou ld not be enforceable. The law d oes

not requ ire the comp uter seller to explain what RAM is, but it does require the creditor

1 Assistant Professor of Law, College of William and Mary. Contact: rmh yne@wm .edu .2 Professor of Law, University of Chicago. Contact: eric_posn [email protected] .edu . Thanks to Douglas

Baird, Peter Alces and George Triantis for their helpful commen ts, and Steve Aase, Nick Patterson and

Scott Hessell for their valuable research assistance. Posner than ks The Sarah Scaife Foun da tion Fund and

The Lynde an d Harry Bradley Found ation Fund for generous financial support.3 Or an or dina ry seller of good s, but to the extent that the seller offers the good s on credit, she is treated

like any specialized creditor, and sellers of goods often subcontract to such sp ecialists.


3/52

2

to explain wha t a finance charge is, and additionally to present information abou t credit

terms in a stylized w ay that is sup posed to ease comparison of the terms offered by

different companies.

In this survey of the law and economics of consu mer finance, we d escribe and

evalua te the main pattern s of consum er finance regulation in the United States. We

examine the state and federal laws that regulate loans to consum ers, includ ing cash

loans and loans that finan ce the pu rchase of real estate and consum er goods. We focus

on (1) price controls (usury laws); (2) restrictions on creditor remedies; (3) bankruptcy

law; (4) limitations on the use of third-party defenses; (5) information d isclosure ru les;

and (6) anti-discrimination law. We do n ot d iscuss general doctrines of contract lawthat are app lied to cash sales and credit transactions alike, includ ing the

unconscionability doctrine; statu tes and regulations that ap ply to all consum er

transactions, not just consumer cred it transactions, such as laws that regu late

advertising or warran ties; and laws that regu late the m arket as a w hole, includ ing

licensing requirements for creditors, geograph ic and activity restrictions, and antitru st

enforcement.4

The literatu re on th e regulation of consum er credit is not as lively as it once was.

Most contributions w ere written in the 1970s and early 1980s, and there h as been little

work in the 1990s other than w ork on p ersonal bankrup tcy. Yet consum er credit

remains a significant topic of pu blic policy, and it continu es to p ose d ifficult questions.

For poorly un derstood reasons, the individual bankrup tcy filing rate has risen rapidly

and steadily since the 1970s, and bankruptcy reform is a recurrent issue in Congress,

genera ting significant attention in the media. The credit card industry has attracted a

great deal of criticism for its aggressive m arketing efforts, confusing credit terms, and

high interest rates. Major retailers such as Sears are criticized for their efforts to

4 We also do n ot d iscuss p ublic choice app roaches to the law of consumer finance; see, e.g., Boyes (1982),

Ekelun d , Heber t and Tollison (1989), Letsou (1995), Buckley and Brinig (1996), Posner (1997).


4/52

3

persuad e customers to reaffirm debts in bankr up tcy. And controversy has swirled

arou nd the sale of credit insu rance to low-income borrowers, a practice that has

genera ted considerable profits for creditors. These and similar issues deserve more

attention from scholars than they have received.

I. Model

A. Lend ing in a Comp etitive Market

An ind ividu al, Debtor, seeks to borrow money in order to sm ooth consump tion

over time. A firm, Creditor, offers to lend mon ey at a certain rate of interest. In a

competitive m arket th e interest rate w ill reflect the time value of money, inflation, and

the risk of default. Debtor accepts the offer if the benefit, that is, the transformation offuture wealth into cur rent consum ption, exceeds the interest rate. If Debtor defau lts on

the loan, he is legally required to pay Cred itor. If in fact Debtor does pay d amages as a

result of a lawsu it, or forfeits collateral of sufficient value, there is no default in an

economic sense, as Cred itor is fu lly compensated . The problem for Creditor is that

Debtor may be jud gment proof as a resu lt of both legal and non legal factors. The legal

factors, to be d iscussed mor e extensively below, include restrictions on the ability of

Cred itor to seize assets or fu ture income in ord er to satisfy a jud gment. Nonlegal

factors include the d ifficulty of tracing Debtor if he flees the jurisdiction or goes into

hiding and collecting from Debtor if he simp ly does not ever earn enough money to pay

off the d ebt.

Default usually occurs in a bad state of the w orld in w hich Debtor loses his job, his

health, or a valuable asset. Risk-averse debtors wan t insu rance against such bad states,

and in add ition to the usu al forms of insuran ce, such as au tomobile and health, Debtor

may pu rchase credit insurance, which w ould repay h is debt to Creditor if he und erwent

certain hard ships such as unem ploym ent, illness, d isability, or destru ction of the

collateral gran ted to Creditor. Debtor may also obtain insu rance from Creditor himself

in the form of a comm itment from Cred itor to forgive missed paym ents if certain even ts


5/52

4

occur.5 Nonrecourse loans also reflect this interest in insu rance. Debtor allows Creditor

to seize certain collateral up on d efau lt but Cred itor gives up the right to seek

repaym ent from Debtors other assets.

Consumer loans take many d ifferent forms. Simplest is an unsecured loan of

cash. Open-end loans like credit card transactions roll over from period to period;

closed-end loans terminate after a specified n umber of payments of principal and

interest. Cred itors also issue loans secured by goods, contract rights, and oth er

collateral. Both ord inary creditors and retailers often lend money necessary to purchase

a particular good, and retain a p urchase money security interest in that good . Banks

and other creditors issue hom e equity loans; these are loans secured by real estate.There is a debate about w hy secured credit exists.6 Creditors shou ld be

indifferent between issu ing a risky un secured loan w ith a high interest rate, and a

relatively safe secured loan w ith a lower interest rate. Debtors should be indifferent

between an ad ditional claim on their assets and a h igher interest rate. Therefore,

because issuing secured rather th an un secured credit involves add itional ad ministrative

costs greater than zero, secur ed cred it shou ld not exist. Two simple nonefficiency

explanations for the existence of secured credit are (1) that security interests are used for

transferring risk to tort and other non adjusting un secured creditors, and (2) that in the

consum er finance context, security interests may be used to circumvent p roperty

exemption laws (White, 1984). Efficiency explanations for secured credit are beyon d the

5 It is likely that som e lenders informa lly comm it to forgive loans or at least missed p aymen ts through

their rep utation. For example, Caplovitz (1967) describes a practice of many cred it sellers of abstaining

from legal action after missed p aymen ts after using social netw orks to verify that their low-income

consum ers are unable to repay. A commitmen t to forgive the loan upon the occur rence of certain events

is identical to credit insurance underwritten by Creditor.6 For early ar ticles pr esenting m ost of the basic argu men ts, see Scott (1977;1979), Smith and Warner

(1979), Jackson and Kronman (1979) and Schwar tz (1981). For more recent treatm ent of the topic, see

Hu dson (1995), Bebchuck and Fried (1996) or see Volum e 80, No. 8, of the Virginia Law Review.


6/52

5

scope of this pap er, although we n ote below w here they are relevant to the law of

consumer finance.7

B. Mon opoly Power

Abundant evidence suggests that the credit market is generally comp etitive

(DeMuth, 1986; Pierce, 1991; Elliehausen & Wolken, 1990), bu t there m ay be local

monopolies in certain areas of the country, perhap s poor n eighborhood s, perhaps the

result of regu lations that raise the cost of entering the credit market. In add ition, many

laws to be d iscussed below are defend ed on the grou nd that th ey correct inefficiencies

created by creditors market p ower, so it is useful to examine the possibility that

creditors do have market power.In an environment with full or symmetric information, a creditor with m onopoly

power will charge an interest rate that is greater than that available in a comp etitive

market bu t w ill generally supp ly the same n onprice terms as a creditor in a comp etitive

market (Schwartz, 1977). Even if these nonp rice term s (such as harsh collection

procedu res) are otherw ise beneficial to the creditor, they will redu ce the willingness of

well-informed debtors to pay for credit. As long as these terms are inefficient, the

creditor wou ld p refer to use its market power to force the debtor to p ay a higher

interest rate. The creditor wou ld gain more by charging a h igher interest rate than by

obtaining consent to a contract with inefficient nonpr ice terms.

The conclusion that the monopolist will use the sam e terms as a competitive

lender requ ires some techn ical assum ptions. Otherwise, because the mon opolist lends

less in equilibrium , the optimal terms of the contract may change. Still, there is no

reason to believe that the contract terms in the m onopolized market w ould be harsher

than the contract terms in the competitive market. And there is no reason to believe

7 Some of the other efficiency explanations clearly have relevance in consum er finance. For example,

these argumen ts focus on the ability of secured cred it to assign p arties to monitor a firms assets.


7/52

6

that forcing monop olists to sup ply some of the terms that w ould prevail in a

competitive market would p rodu ce a gain. Because the monop oly power remains,

further d istortions wou ld occur in the un regulated terms. (Schwartz, 1977)

Monop oly power can have other effects as well, bu t these require asymm etric

information and thu s will be discussed below.

C. Asymmetric Information: Consum er Ignorance

Even if there are num erous lend ers in a m arket, each lender may h ave some

degree of market power because of the inability of consu mers to costlessly comp are

pr ices and term s. Depending on the source of the information failure, this may result in

either an abnorm ally high price or abnormally harsh term s. Some creditors will lendonly to those consumers wh o are u nable to comp are the (price or non price) terms of the

loan offered w ith the terms available elsewhere in the market.

The problem requ ires that a large enough num ber of consum ers find it difficult

to shop arou nd . The comp etitive outcome wou ld occur if a significant subset of the

consum ers become informed and if creditors are unable to d iscriminate between th ese

creditors and the uninformed by, for examp le, offering loans with d ifferent terms and

interest rates with only the informed consum ers able to determine the m ost desirable

loans. (Schwartz and Wilde, 1979; 1983). That is, if enough consu mers compare loans

before borrowing, no lend er could make a profit by lending only to those who d id not

compare. However, by shopping the informed consumers effectively confer a positive

externality on th e un informed and thus consum ers may h ave too little incentive to

acquire information.

Creditors would seem to have every incentive to distingu ish them selves from

their competitors if they offered credit on more attractive terms. However, they cannot

overcome consu mer ignorance (possibly resulting from m isleading claims mad e by

Likewise, a mortgage may ensure that some creditor m onitors the homeow ners insurance pu rchased by

the consumer.


8/52

7

rivals) when th at ignorance is severe enough. There is a limit to how mu ch explaining a

creditor can do before losing the attention of its custom ers. Further, in a comp etitive

market each creditor has insufficient incentive to educate consumers because of that

creditors inability to internalize all of the ga in from that information. This problem is

lessened somewhat if lend ers have market p ower because they capture more of the

returns from the information and thus have an incentive to provide information abou t

the entire product, not just the brand . How ever, a creditor with market power may

have an incentive to provide too little information in order to aid in pr ice discrimination

(Beales, Crasw ell, and Salop, 1981a). Fur thermore, creditors will have insufficient

incentive to explain the economics of the credit market, and the meaning of contractterms, because they cannot prevent peop le who have benefited from their expectations

from seeking loan s elsewhere. (Beales, Crasw ell, and Salop, 1981b).

It is of course possible for th ird p arties such as trad e associations or even

independ ent group s such as Consumers Union to prov ide comp arisons or stand ards for

comparison. How ever, each of these solutions has its own problems. An independen t

group such as Consum ers Union m ight supp ly too little information because it would

have difficulty p reventing consumers from sharing the information w ith others who d o

not pay Consumers Union for it. Trade associations may have an incentive to create

standard s or report information that favors those w ithin the a ssociation over other

competitors, or, conversely, to avoid creating stand ard s for fear of d raw ing the

attention of an titrust regu lators. (Beales, Craswell, and Salop, 1981b; Schwartz an d

Wilde, 1977).

D. Asymm etric Inform ation: Creditor Ignoran ce

The simplest form of information asymm etry occurs when Debtor kn ows his

willingn ess to pay for credit while Creditor does not. If Creditor has a mon opoly, he has

an incentive to d iscover Debtors valua tion so that he can price d iscriminate. It is

possible that Creditor can separate higher and lower valuation d ebtors by offering


9/52

8

contracts with inefficient terms. For example, Creditor might offer a loan w ith a h igh

interest rate and a loan with a collateral requ irement bu t a lower interest rate if this

wou ld help him distingu ish between those w ho are particularly sensitive to the interest

rate and those who are not. The efficiency implications of this practice are obscure. As

long as the m onop oly remains intact, a law that p rohibits the inefficient term w ill both

eliminate the cost associated with the term and red uce value by interfering w ith price

discrimination. Creditor will offer an average interest rate that drives low valuation

debtors ou t of the m arket (Craswell, 1995).

Another form of asymm etric information occurs wh en Debtor knows the

probability of default and Creditor does not. Assume that because of personalcharacteristics unobservable by creditors, some debtor s have a h igh probability of

default (bad d ebtors) and others have a low p robability of default (good debtors).

Harsh remed ial terms are more costly for bad d ebtors than for good d ebtors because the

bad d ebtors are more likely to default and thu s to become subject to the terms. If

creditors believe tha t any d ebtor who fails to grant a security interest (or agree to some

other harsh rem edial term such as a cognovit clause) is a bad d ebtor, creditors may offer

two contracts: a secured loan with a low interest rate and an u nsecured loan with a h igh

interest rate.8 The good debtors effectively signal their type by choosing the secured

loan with the low interest rate wh ile the bad debtors choose the un secured loan. The

creditors beliefs are va lida ted in th is separa ting equ ilibrium . This wou ld be tru e

regard less of wheth er the m arket is competitive or mon opolistic (Rea, 1984; Aghion

and Hermalin, 1990).9

8 Creditor might also be able to d etermine the type of the d ebtor through the size of the loan requested

(Freixas and Laffont, 1990).9 If one imp oses stronger assump tions, one can show that a monopo list will behave differently than a

lender in a comp etitive mar ket. For examp le, Besanko and Thakor (1987) show that un der certain

conditions a m onop olist is more likely to prefer credit rationing over collateral.


10/52

9

A rule banning security interests and other harsh rem edial terms would be

efficient if the total costs of the signaling exceed th e total gains. If there is no cred it

rationing an d no effect on the d ebtors efforts to avoid d efault (we d iscuss both these

assump tions below), the redu ced interest rate charged to the good debtors should be

roughly offset by the increased interest rate charged to the bad d ebtors. In fact, it is

even possible that banning such signaling would benefit the good debtor s. The reason

is that the good debtors might prefer a contract w ith n o collateral and with an interest

rate that reflected the average p robability of default in the p opu lation, compared to a

contract with collateral and a lower interest rate. In the absence of a legal ban on

secur ity interests Creditor would not offer the efficient p ooling contract because of hisbelief in equilibrium that good debtors issue security interests and bad debtors refuse to

issue secur ity interests.

The fact that secur ity interests and other consensu al creditor remed ies can be

used to signal information abou t debtors does not necessarily mean th at they shou ld be

banned because this signaling m ay p lay a role in red ucing a related p roblem caused by

asym metric inform ation, cred it rationing (Betser, 1985; 1987). Creditor sets the interest

rate to reflect the average p robability of default in his portfolio. Assum e that good

debtors are less willing to p ay a higher interest rate because they are more likely to

repay the loan.10 If Creditor cann ot distingu ish am ong d ebtors, the expected p rofit from

any p articu lar loan w ill decline as the interest rate rises beyond some p oint, because as

the interest rate increases the good d ebtors drop out of the market. Therefore, creditors

(monop olistic or competitive) will not r aise interest rat es above th is point and credit

will be rationed : the demand by bad debtors for (even h igh-interest) loans w ill be

unm et (Stiglitz and Weiss, 1981). If there are too m any bad debtors in the market, their

pr obability of defau lt is sufficiently high, and the d ivergence in the p robability of

10 The assum ption that the good debtors are more likely to d rop ou t of the market as the interest rate rises

is stand ard , but not un iversal. For an article assum ing the contrary, see Besanko and Thakor (1987).


11/52

10

default is too large, the market u nravels leaving only the bad debtors w illing to borrow,

but creditors un willing to lend to them. This is the phenom enon ofadverse selection

(Akerlof, 1970). Security interests and related terms may red uce adverse selection by

enabling the creditor to distinguish am ong good and bad debtors. Security interests

and similar terms can serve as signals because they are cheap er for debtors who are less

likely to d efault.

Credit rationing can also result if there is asymm etric information abou t wh ether

or not the d ebtor can repay a loan (Jaffee and Russell, 1976). That is, debtors may

have an incentive to claim d estitution in order to avoid rep aymen t and it may be

difficult for creditors or courts to verify that th is is correct. In an extreme case, the onlymechanism that the creditor may use to force repaym ent is to deny futu re credit (Allen,

1983). Collateral with personal value to the debtor and other forms of creditor remed ies

ensure that a d efaulting d ebtor cannot in fact repay if the d ebtor wou ld rather repay the

loan th an end ure the p un ishm ent of repossession (Rea, 1984; Scott, 1989).

Another kind of asymm etric information p roblem arises when Debtor has

private information abou t the care with w hich he avoids default. Care can m ean a lot

of things: (i) working hard , so that he is not fired and dep rived of an income to repay

the loan; (ii) protecting assets or collateral so that they may be liquidated in case of

default; (iii) avoiding p hysical risks that might resu lt in injury; or (iv) avoiding risky

investmen ts. If Creditor cann ot observe Debtors level of care and pen alize Debtor if he

takes insufficient care; and if Debtor d oes not expect to repay the d ebt in full because of

the legal and non legal factors mentioned above; then Debtor will take a subop timal

level of care. This is the problem ofmoral hazard.

One response to this moral hazard is to proh ibit, by contract, behav ior that

increases risk. An examp le is the covenan t against using residential property for

commercial pu rposes. But this response really assum es away m oral hazard by

sup posing tha t cond uct is observable: when cond uct is un observable, it cannot be


12/52

11

prohibited by contract. The second response to moral hazard is to requ ire Debtor to

bear some of the cost of default, thu s converting a debtor who m ight otherw ise be fully

judgm en t-p roof in to on e w ho is part ly jud gmen t-p roof. For example, requirin g that

per sonally valuable prop erty be collateral redu ces the probability that Debtor w ill be

able to protect it at the time of default throu gh jud icial process. Alternatively, Creditor

might seek to destroy Debtors reputa tion by publicizing the default; to cause psychic

harm by liqu idating a gu aran tee from a loved on e; or, in the case of loan sharks, to

break bones. Even though these actions prov ide no direct benefits to Creditor while

conferring costs on Debtor, they m ay be efficient because they red uce moral hazard

(Rea, 1984).

II. Law

A. Price Restrictions: Usury Laws

Description. Every state has laws restricting the interest rate that can be charged

for consumer loans. How ever, although the interest rate ceilings in some states are qu ite

low, their effect on th e credit market is likely to be limited . There are many reasons for

this. First, federal law p reemp ts state usury laws in a variety of cases, the most

importan t being hom e equity loans, for which there is no federal interest rate ceiling.11

Further, since the late 1970s, federal law h as permitted federally insu red state

institutions to exp ort the high interest rate ceilings of the states in wh ich they are

located, thus perm itting them to lend at high interest rates to debtors wh o reside in

states with low ceilings. 12 Second, state usury ceilings have long been ridd led by

exceptions for, among other th ings, small loans, retail installment loans, and loans

issued by favored institutions like cred it unions. Third, interest rate ceilings often

11 This was actually an incomplete preemption as the states were given the right to op t out and fourteen

states did so. (Alperin and Chase 1986)12Marquette National Bank of Minneapolis v. First of Omaha Service Corporation, 439 U.S. 299 (1978).


13/52

12

un derstate their effective limits, the r esult of special rules for calculating interest rates

when lend ers comp ound , charge fees, give discounts, and calculate balances in different

ways. Fourth , remed ies for violation of usu ry laws are frequently nar row (Alperin and

Chase, 1986). Fifth, usu ry ceilings may be evaded in many w ays, for examp le, by

disguising interest as part of the p rice of the good if sold on credit with a d iscount for

cash transactions, or by d isguising a secured transaction as a lease with h igh rental

payments and a low bu y-out price (Peterson, 1983). Sixth, man y usury ceilings are set

at fixed interest rates thereby lessening their imp ortance in p eriods of low inflation such

as the p resent time. Still, usury ceilings have theoretical interest and historical

significance, and th ey continue to influence many ord inary lend ing practices.Effects. Usury laws are simp ly price controls, and can be pred icted to have many

of the same effects: queuing, unsatisfied deman d, and an illegal market, loansharking.

Unlike stand ard price controls, however, it is dou btful that usu ry laws lower the p rice

of a loan (the interest rate) paid by an y particu lar borrower. Because there are m any

alternative uses of capital, a p rohibition of high interest rates will simply lead creditors

to refuse to lend to h igh-risk debtors and instead lend to lower-risk debtors at legal

rates or to seek other investment op tions. To the extent that high interest rates are the

result of market p ower enjoyed by lenders, either as a result of monop oly power or

search costs (Ordover and Weiss, 1981), usury laws m ight be able to lower the rate

charged to borrowers. But there is little evidence that creditors have market pow er or

that consu mers lack good information abou t interest rates, especially after the

implementation of the Tru th in Lend ing Act described below (Schwartz and Wilde,

1979). Even if lenders did have some mon opoly pow er and the usu ry ceiling red uced

the rates pa id by some debtors, these ceilings would cause higher risk debtors to be

denied credit as cred itors would be un able to charge them higher rates, thus offsetting

much, if not all, of the welfare gains. Ausubel (1991) raises the possibility that interest

rates on cred it card s are artificially high because of the irrationality of consumers. He


14/52

13

argues that low-risk credit card users intend never to borrow and therefore do n ot

consider the interest rate when choosing among credit card s; wh ereas high-risk credit

card users do consider the interest rate. Because cred itors cannot d istinguish between

low-risk and high-risk debtors, no creditor wou ld lower its interest rate as it wou ld

d isproportionately attract high-risk debtors. A limit on interest rates could th erefore be

welfare imp roving. However, Ausubels thesis is controver sial. More recent stud ies

have found that consum ers are sensitive to interest rates (Gross and Souleles, 2000) and

economists remain nearly unanimous in condemn ing u sury laws.

Some of the early emp irical literatu re on usu ry d id find that states with usury

laws had lower average interest rates.

13

But m ost of the literature found that u sury lawsresult in a significant red uction in the access to credit for high-risk debtors.14 In fact,

Villegas (1982; 1989) find s that the en tire decline in th e average interest rate is

attributable to the exclusion of these debtors from th e market; the usury laws do not

redu ce the interest rate paid by an individual debtor. This result is unsurp rising: the

sup ply of loans should not be inelastic if capital can be used for other p rojects or in

other jur isd ictions. The only surprising thing abou t these findings is that because usury

laws are so easy to circum vent, it is d ifficult to believe that th ey have any impact on

behav ior in a m odern econom y with efficient capital markets.

Usury laws have a long and significant history, are still imp ortant in many

jurisd ictions, especially Islamic cou ntr ies, an d con tinue to r eson ate w ith th e m oral

13 See, for exam ple, Greer (1973), Peterson (1979), Peterson and Ginsberg (1981), Shay (1973) and Wolkin

and Navratil (1981).14 See, for exam ple, Boyes and Roberts (1981), Dunkelberg and DeMag istris (1979), Greer (1975), Kawaja

(1969) and Shay (1970). For stud ies find ing no credit rationing , see Eisenbeis and Mur ph y (1974),

Goud zwaard (1968; 1969) and Peterson (1983). This is consistent with stud ies of the mortgage cred it

market w hich typ ically find that restrictions on usur y redu ce the number of building p ermits due to a

red uction in hom e finan cing. See Au stin and Lindsley (1976), Boyes and Roberts (1981), Robins (1974),

Ostas (1976), and Crafton (1980). But see Rolnick, Graham , and Dah l (1975) and McNulty (1980) find ing

no significant effect on bu ilding p ermits bu t finding a significant effect either on n on-price terms or on

loan volume.


15/52

14

intuitions of many p eople. This has led scholars to suggest possible benign

explanations for their pop ularity. First, a usury law may be a crud e form of social

insurance in a jur isd iction that has poorly developed capital markets closed to the

outside world and an inefficient or nonexistent welfare system (Glaeser and

Scheinkman, 1998). If usu ry ceilings dep ress the p rice of cred it, the p oor w ould be able

to borrow mor e cheaply and this may be efficient if the p oor have a su fficiently higher

marginal utility of money than th e rich. From an ex ante perspective, an ind ividual

benefits from usur y laws if his lower retu rn w hen h e has capital to spare in some futur e

state of the w orld is offset by his lower borrow ing costs when h e needs to borrow in

some alternative futu re state of the world. This argu men t is inconsistent with themobility of mod ern capital, and so has no application to mod ern cond itions;

significantly, usury laws hav e been repealed in every ind ustrialized n ation except the

United States, Belgium , and France (Alper in and Chase, 1986), though a fairly restrictive

usury law w as enacted in Italy in 1996.

Second , welfare laws create a moral hazard and usu ry laws m ay therefore be

need ed precisely because they restrict access to credit. Because welfare laws reduce the

consequences of default for the debtor by providing him w ith a m inimum stand ard of

living after his creditor employs all available remedies, the debtor w ill be willing to

borrow to un dertake riskier ventu res (Posner, 1995).15 Usury ceilings prevent these

high-risk loans and therefore redu ce the negative consequences of the m oral hazard .

This argu men t assum es that peop le benefit from welfare laws, and , unlike the first

argu men t, that an effective w elfare system is in p lace.

There is little statistical evidence for these theories; they are intend ed to

rationalize historical practice and rely mainly on an ecdotal historical evidence.

15 A related argu ment p osits that usury laws prevent low-income d ebtors with a high discount rate from

borrowing against future welfare paym ents and that this credit rationing p ermits a society committed to

pr ovided a minimu m per period w elfare to do so at a lower cost. Avio (1973).


16/52

15

B. Restrictions on Creditor Remed ies

Description. A confusing array of federal and state laws restrict many of the tools

that creditors have trad itionally used to force repaym ent, includ ing the rep orting of

past consumer behav ior16 and nonlegal mechanisms such as contacting the debtor and

third parties to requ est repaym ent. Self-help can be effective; debtors repay loans in

ord er to avoid unp leasant phone calls, threaten ing letters, hu miliation in front of

friends, emp loyers, and family members, and damage to their credit repu tation.17

Federal and state laws restrict the ability of credit reporting agencies to record

information that creditors may find relevant. The Fair Debt Collections Act requires

certain kinds of creditors to (1) ver ify the debt if the consum er challenges it; (2) refrainfrom threa ts and h arassment; (3) refrain from p ublishing the nam es of defau lting

debtors; and (4) refrain from misrepresentation of their legal rights, the consequ ences of

non payment, and so forth (Alperin & Chase, 1986). While the federal act does not

d irectly app ly to the creditor that originated th e loan, its restrictions may ap ply to the

creditors law yers18 and some states app ly similar regulations to the original creditors as

well. Accord ingly, we d iscuss these rules in th is section rather th an in the section,

below, on th ird party defenses.

When self-help fails, creditors often sue and obta in repaymen t throu gh

prejud gment and p ostjud gmen t remedies. Prior to jud gmen t a creditor may be able to

obtain a lien on the d ebtors assets and to garnish the d ebtors wages, and these powers

are usually sufficient to obtain repayment. How ever, prejud gment attachment an d

garnishment are now regulated in various ways by the state and federal government;

16 The Fair Cred it Reporting Act, 15 U.S.C. 1681 (2000), limits the rep orting o f bankru ptcies by consumer

repor ting agencies to ten years and limits the repo rting of most other adverse information to seven years.17 For examp le, early in the mod ern history of consumer cred it small lenders relied on the pro fessional

services of the bawlerou t, a female employee who w as assigned the job of trapp ing the d elinquent

borrower before co-workers and family in order to browbeat him publicly for being a sorry deadbeat.

Calder (1999); Rea (1984).18Heintz v . Jenkins, 514 U.S. 291 (1995).


17/52

16

they are also subject to constitutional du e process limitations. Garnishment, both

prejud gment and postjud gment, is heavily restricted by federal law (roughly to 25

percent of wages, but with many exceptions) and some states have even m ore restrictive

limits or prohibit garn ishm ent altogether. There are fewer restrictions on post-

judgm en t rem ed ies; these u su ally involve the sheriff seiz ing and au ctioning off

property, or (again) garnishmen t of wages, which remains heavily restricted even as a

postjud gment remed y. Postjud gment seizu re of property is significantly curtailed by

state (and federal) exemp tion laws, wh ich limit the kind and amou nt of prop erty (home

equ ity, clothing, furniture, pensions and so forth) that can be seized in ord er to satisfy

unp aid d ebts.A creditor can imp rove h is ability to collect by bargaining in ad vance for certain

rights. For example, a cognovit note, in w hich the debtor essentially bind s himself to

confess jud gment if he defaults, relieves the cred itor of the trouble of proving h is case in

court. However, cognov it notes are illegal in many contexts (Alperin an d Ch ase, 1986).

By obtaining a security interest a creditor gains priority over un secur ed cred itors and , if

the security interest is perfected , cred itors with later in time security interests in the

same p roperty. In the context of consum er finance, the power of the secur ity interest to

alter the creditors rights w ith respect to the debtor is perhap s even more significant.

Because a secured creditor can seize mu ch of the prop erty that w ould otherwise be

exempt under state or federal law, debtors and creditors can use security interests to

effectively waive many of the exemptions. In add ition, the secur ity interest may a lso

allow the cred itor to skip som e of the steps in the jud icial pr ocess, and even skip it

altogether if he can rep ossess the collateral without br eaching the p eace. At one time,

creditors wou ld obtain secur ity interests in all the debtors household good s, even those

that w ere not purchased from the cred itors or with the creditors money.

Today, however, secured consum er cred it is heav ily regulated . FTC regulations

and some state laws forbid creditors to obtain nonpossessory non pu rchase money


18/52

17

secur ity interests in household goods, although there are some exceptions. The

bankru ptcy code also permits debtors to nullify nonpurchase money liens in many of

these same hou sehold goods. Some states provide the d ebtor (and sometimes even

junior lien -hold ers) a right to red eem the collatera l for up to a year even if the collatera l

has been sold to a third p arty, and require the creditor to obtain a court jud gmen t before

repossessing collateral. Finally, a foreclosure on collateral will sometimes p reclude the

creditor from seeking the remainder of the amou nt ow ed through a d eficiency

judgm en t. Common law and state statu tory ru les granting a r igh t of redem ption an d

prohibiting d eficiency jud gments are impor tant forms of regulation of the hom e

mortgage market.Many of these restrictions are available in ban kru ptcy, but w e discuss

bankru ptcy separately, below.

Effects. Critics argu e that the strong contractua l rights to repossess consum er

good s are inefficient because the repossessed p roperty has minimal resale value for the

creditor but considerable person al value for the debtor; these remed ies are used in

ord er to coerce (Leff, 1970; Whitford, 1986). While there is some evidence tha t fire sales

exist (assets are sometimes sold for less than their wholesale book value),19 critics argue

that the p erception that value is lost is based on a misun derstand ing of the op eration of

markets; it is unlikely that value wou ld be destroyed g iven the characteristics of the

debtors and creditors and the ability to renegotiate (Schwartz, 1983) and as long as

creditors believe tha t a rep utation for aggressive collection techn iques might scare off

debtors (Peterson, 1986). As we saw above, how ever, even collection mechan isms that

are inefficient at the time of collection m ay be efficient ex ante precisely because they are

coercive. They can, in theory, redu ce moral hazard by increasing the cost to the

19 See Schuchman (1969), White (1982), Note (1971), Note (1975). Grau & Whitford (1978) show that

repossessions declined after Wisconsin enacted a statu te that required creditors to obtain a jud gmen t


19/52

18

debtor from defaulting (Rea, 1984), and ad verse selection by enabling the cred itor to

distinguish among debtors by risk level (Bester, 1985). (See generally Epstein, 1975;

Scott, 1989.) Regard less, the rest rictions on creditor collections generate costs for

creditors and creditors should p ass these costs on to debtors in the form of higher

interest rates or else deny access to credit, particularly for high risk debtors.

Restrictions on coercive creditor rem edies in general,20 and exemp tions in p articular, 21

are associated with higher interest rates and increased probabilities of denial of credit.

The effects are more pronou nced for low-income or low-asset debtor s.22

As noted above, exemp tions do n ot d irectly affect the ability of the secured

creditor to foreclose on the most valuable assets of the d ebtor (such as the hom e) andtherefore only affect the supp ly of cred it by raising tran saction costs, that is, by

requ iring p arties to go throu gh the formality of obtaining a secur ity interest in order to

make assets available for collection in case of default. In contrast w ith the resu lts for

genera l credit foun d by Grop p, Scholz and White (1997), Berkowitz and Hynes (1999)

find that mortgage lend ers do not increase the rate of denials or the interest rate in the

face of larger exemptions. There is actually a small decrease in these variables. This

may occur because as exemptions increase, debtors have more un encumbered wealth

with which to pay off secured creditors, and costly foreclosures are less likely to be

necessary.

before seizing collateral from a d efaulting debtor. This result is entirely pr edictable, and , as they app ear

to acknowledge, they do not show that debtors are mad e better off by the law in an ex ante sense.20 See, for example, Greer (1974) and Barth, Gotur, Manage and Yezer (1983).21 See Grop p, Scholz and White (1997) (examining th e effect of the exemptions on credit m arkets

generally) and Berkow itz and White, (2000) (examining th e effects of the exemptions on the m arket forsmall business loans). We note that Grop p, Scholz and Wh ite (1997) use the same d ata set used by

Villegas (1990) to investigate the effects of usu ry law s and restrictions on creditor collections other than

exemptions. A further stud y disentang ling the effects of each of these restrictions wou ld be useful.22 That exemp tion laws have a pr onou nced effect on debtors with few assets is somew hat of a puzzle as

these debtors can exempt all of their assets in almost any regim e. For examp le, Gropp , Scholz and White

(1997) find a significant red uction in th e access to credit for d ebtors with assets of less than $7,885 when


20/52

19

Limitations on creditor remed ies do provide some benefits to the debtor. These

limits provide some insuran ce by protecting the debtors income and assets when h e is

least well off. As noted above, they m ay also prevent socially wasteful debt collection

practices. But a defense of these laws has two predicates, both of them difficult to

establish. First, the law should restrict remed ies only if a market failure p revents

creditor from supplying remed ial terms that d ebtors would be willing to pay for and

prevents the d ebtor from using alternative form of protection, such as credit insurance.

23 The usual market failure argum ents can be m ade, of course. Perhaps, adverse

selection exp lains w hy credit contracts rarely limit the creditors remed ial rights. But if

the market did fail in this way, it is hard to understand wh y there is such a robustmarket in credit insurance.

Second, the d efense assum es that the law d oes reflect debtor s preferences. But

the variation of the law across states is too extreme to reflect plausible d ifferences in

debtors preferences. For examp le, an individual can exemp t only a few thou sand

dollars worth of assets in Alabama but a potentially unlimited am oun t of home equ ity

in Florida. The variance is too high to reflect risk preferences across the two states or

other d ifferences in the states economies; and indeed a stud y of exemp tion laws in all

fifty states over a twen ty-two year p eriod reveals no correlation between the generosity

of exemp tions and deman d for insuran ce (Posner, Hyn es, and Malani, 2001).

The exemptions and the bankruptcy right to a discharge may add ress another

concern, that of creating a class of people who d o not w ork because they cannot keep

their income or the assets they pu rchase with it; this explanation is also consistent w ith

limitations on the ability of creditors to contact (and ann oy) a debtor s emp loyer.

Although creditors and debtors have incentives to renegotiate ex post, as the history of

the exemp tions move from the mer ely large (exemp tions between $25,400 and $70,400) to the un limited

exemptions.


21/52

20

debtor p risons shows renegotiations will occasionally fail because creditors want to

maintain a repu tation for toughness or hope to flush out d ebtors who have concealed

their assets. A class of peop le imm obilized or even impr isoned for debt sits un easily

with mainstream p olitical commitmen ts. Of cour se, this argum ent fails to the extent

that d ebtors can w aive exemp tions un der state law or throu gh the u se of security

interests. However, statutory waivers of exemptions are not effective in bankruptcy;

nor are nonpurchase money security interests in m any forms of personal property. In

addition, a d ebtor cannot waive his right to a d ischarge.

To the extent that d ebtors can w aive exemp tions and other limitations on

creditor remed ies, these laws merely change the default rule for collections up ondefault. Rather than contracting for protection throu gh credit insuran ce, nonrecourse

loans and other m eans, the debtor w aives protections throu gh security interests,

cognovit notes, etc. A comp arison of the merits of the two d efault rules would r equire a

deeper analysis of the preferences of debtors, the costs of contracting, the enforcemen t

of limitations on default planning and other factors that are beyond th e scope of this

paper.

Villegas (1990), Peterson (1986), and Barth , Cord es and Yezer (1986) try to

determine if restrictions on creditor rem edies provide a net benefit or a net cost. Their

logic is that if the r estrictions are ben eficial, the increase in the interest ra te d eman ded

by the creditors should be more than offset by the increased w illingn ess to pay by the

debtors. One shou ld be able to verify this d irectly by separately estima ting supp ly and

demand or indirectly by observing the total quan tity borrowed. The results of these

stud ies are mixed. Barth, Cord es and Yezer (1986) found that w hile statutes limiting

deficiency jud gments may p rovide a net benefit, legal restrictions on confessions of

judgment clauses, rest rict ion s on garnishment, and restrict ion s on security interests in

23 Supra note [__] and accomp anying text. We acknowled ge the criticisms of this market, where p rofits

app ear to be unu sually high.


22/52

21

real prop erty created a net cost. Villegas (1990) found that r estrictions on security

interests in p ersonal property and restrictions on wage garn ishm ent prov ided a net

benefit but that proh ibitions on w age assignment created a net cost. Although h e did

not d iscuss this norm ative ana lysis, Greer (1974) and Peterson an d Frew (1977) both

foun d that p rohibitions against attorneys fees and garnishment red uced the total

borrow ings. Gropp , Scholz and White (1997) also do n ot conduct an explicit comparison

of the costs and benefits of the exemp tions. However, they do examine the effect of the

exemptions on the total quan tity of credit and find that the exemp tions increase total

borrow ings by high-asset debtors but decrease total borrowing by low-asset debtors.

Therefore, following the logic of Villegas (1990), larger exemptions seem to provide anet benefit for high-asset debtors but provide net costs for low-asset debtors.25

While these results are interesting, the tests are imp erfect. The comparisons

assume that lenders and borrowers (or at least some borrow ers) are aware of the legal

restrictions, can correctly predict their imp lications a t the time of borrowing, and can

ad just the contract in light of these factors. This assum ption is questionable if the

market failure justifying governm ent intervention is that d ebtors und erestimate the

probability of defau lt or that d ebtors lack information abou t the consequences of

default. Moreover, even if debtors are fully informed , a find ing that total credit

increases is not a necessary cond ition for d etermining that the laws are beneficial.

Debtors may be w illing to accept lower borrow ing levels as a p rice for increased

insurance.26 In add ition, involuntary creditors such as tort claimants cannot ad just to

the laws by charging a h igh interest rate. Finally, the limits on creditor remed ies may

25 Schill (1991) finds that the righ t of redem ption an d an ti-deficiancy jud gmen t rule in the m ortgage

market raise mor tgage interest rates by only a small amoun t, and argu es that this cost may be

outw eighed by other benefits. He does not examine the effect of these rules on access to credit, how ever,

and he does not empirically evaluate the benefits in add ition to the costs.26 For an example of this, see App end ix.


23/52

22

play a role similar to those of usu ry laws in discouraging high-risk loans und ertaken as

a resu lt of the moral hazard created by social welfare law s (Jackson, 1985).

C. Bankruptcy

Description. By filing for bankrup tcy un der Chapter 7 of the federal bankru ptcy

code, a debtor can p rotect all of his future income from his creditors, retain exemp t

prop erty, preserve certain kinds of trust fun ds includ ing pensions even when they are

not exempt u nd er nonbankrup tcy law, and d elay the seizure of other assets through the

au tomatic stay. Debtors may not file for bankruptcy seriatim, bu t the bankru ptcy

discharge rem ains a pow erful w eapon for the d efaulting debtor. There is an extensive

literature on the structure and effects of this law , and w e will not rep roduce it here.

27

However, a brief overview of some of the emp irical literature on ban kru ptcy is

necessary for a p roper un derstand ing of the results discussed in Section B.

Effects. While several stud ies cited above find that restrictions on cred itor

remedies, includ ing exemp tions that apply in bankrup tcy, affect the decision to borrow ,

there is little evidence that these same restrictions affect the decision w hether or n ot to

repay. This is surp rising because debtors in finan cial d istress shou ld be more aware of

the law of collections than d ebtors applying for a loan, particularly if they h ave retained

an attorney. Likewise, creditors shou ld not change their lend ing behavior in response

to exemptions un less the exemp tions have a real effect on their expected losses.

Unfortunately, good d ata on d efau lt and collections are not available by state.28 The

existing evidence, based on ban krup tcy d ata, suggests that exemp tions d o not

27

For a recent survey of the consum er bankru ptcy literature, see Kowalewski (2000). The law isfrequ ently criticized for being too generou s and inflexible. See, e.g., White (1998a; 1998b), Wang and

White (2000), Adler, Polak, and Schwartz (2000).28 Empirical stud ies of the effects of garnishmen t restrictions on the filing rate h ighlight the shor tcomings

of focusing on bankru ptcy filings. These stud ies generally find that states with laws that are more

restrictive of the ability of a creditor to garnish a debtor s wages have h igher filing r ates. See, for example,

Ap ilado, Dau ten, and Smith (1978), Ellis (1998a) and Heck (1981). While this effect could be due to higher

repayment rates, it is more plau sibly du e to the ability of defau lting debtors to pro tect their futur e

incomes without filing for bankruptcy.


24/52

23

significantly a ffect the filing rate (see below) an d it is unlikely th at the exemp tions

substantially affect repaym ent rates in bankru ptcy given the m inimal repayments that

unsecured creditors actually receive (White 1987).

Arguing that larger exemp tions should make ban krup tcy more attractive to

debtors, many scholars pred icted that larger exemptions should increase bankrup tcy

filings. While White (1987) found a p ositive and statistically significant effect, the effect

that she found was small and virtually all other pu blished p apers have found either no

statistically significant effect or even an effect with the wrong sign.29 This result has

been repeated in m ore recent stud ies reexamining the p roblem w ith p anel data or

quasi-experiments.

30

Because the literature w as forced to compare th e exemptions and the bankrup tcy

filing rate, its failure to find a strong p ositive correlation is less surp rising than it

appears. The majority of exemptions available in bankru ptcy are also available to a

debtor d efaulting u nd er state law and therefore while the exemp tions should m ake

default relatively more attractive than repaym ent, they d o not necessarily make

bankru ptcy relatively more attractive than defaulting u nd er state law. One can

reestablish a link betw een the exemp tions and the incentive to file for bankru ptcy based

on transactions costs, thou gh these theories are more tenuou s. In add ition, if larger

exemp tions lead to more d efaults they shou ld also lead to more bankru ptcy filings as

long as bankruptcy filings remain a fairly constant fraction of all defaults. (Hynes,

1998).

The failure to find a correlation between exemption levels and bankruptcy filings

may also be due to an inappropriate use of aggregate data w hen testing a hypothesis

29 See, for examp le, Apilado, Dau ten, and Smith (1978) (finding m ixed results when testing for a link

between exemptions an d the filing rate p rior to the en actment of the Bankrup tcy Reform Act of 1978);

Peterson an d Aoki (1984); and Shiers and Williamson (1987).30 See Buckley an d Brinig (1998), Weiss, Bhan dari and Robins (1996). But see Pomykala (1997) and Hynes

(1998) finding sign ificant p ositive effects.


25/52

24

about ind ividu al behavior.31 The exemp tions may h ave little effect on aggregate filing

rates because they are relatively generou s compared to the assets of most Am ericans,

and the redu ction in access to credit may mean that d ebtors in states with large

exemptions may be less likely to end up in financial d istress.32 While current working

papers u se ind ividua l level data to examine the filing d ecision, their results cann ot

read ily be interp reted as a test of the impact of exemption levels. These papers test

wh ether d ebtors respond to the financial incentives of bankru ptcy m ore generally,

including th e d ischarge, rather than just the exemp tions and therefore examine the

effect of the d ebtors benefit from filing. Benefit is d efined as the d ebt that can be

discharged less any assets above the exemp tion that the debtor would lose by filing(Fay, Hurst and White, 1998; Chakravarty an d Rhee, 1999). Even if the exemp tions have

no effect on the filing decision, the coefficient on benefit m ay still be significant

because households w ith more d ebt file in ord er to obtain the d ischarge.

Several studies investigate w hether the Bankruptcy Reform Act of 1978 increased

the filing r ate and related actions.33 That statute instituted several reforms that could

have mad e bankru ptcy more attractive (Domowitz and Eovaldi, 1993),and the

bankru ptcy filing rate increased marked ly in the years that followed . Using time series

econometr ics techniques, scholars have tried to d isentangle the effects of this act from

the significant macroeconomic effects of this time period. The majority of the early

stud ies addressing th is question d id, in fact, estimate that the code played a significant

31 Early scholars attributed this failur e to possible simultaneity bias; legislatur es might ad opt sm aller

exemptions in r esponse to higher filing rates. Peterson and Aoki (1984), Shiers and Williamson (1987).

However , it is unclear why this same bias wou ld no t have a significant effect on the stud ies of the creditmarket. While we lack a good explanation for a states choice of exemp tions, one might be able to test

this theory by using historical exemp tions as an instrum ental variable.32 It is possible to collect data on loans mad e by lend ing institutions in each state. How ever, the

importance of national lenders in the mortgage and credit card ind ustry makes it un likely that such a

variable wou ld be highly correlated with th e debt issued by residen ts of each state.33 While we w ill discuss those articles d iscussing the decision to file, the interested r eader m ay wish to

consult those articles discussing the effect of the act on the choice between Chap ters 7 and 13. See, for

example, Domowitz and Sartain (1999a; 1999b).


26/52

25

role in increasing the bankru ptcy filing r ate.34 One d ifficulty w ith this literatu re,

how ever, is that it requires a controversial assum ption regard ing the treatment of

married coup les filing jointly, which was not permitted prior to the Bankruptcy Reform

Act of 1978. Domowitz and Eovaldi (1993) examine su mm ary statistics presented in

stud ies of actual filings to determine a range of values for the proper ad justm ent to the

postact filing rate. When they use the lowest value of this range, they estimate tha t the

Bankruptcy Reform Act of 1978 increased the filing rate by twen ty-two p ercent.

However, the standard errors in their regression are so large that th is estimate is not

statistically significant; one does not find a statistically significant resu lt un til one uses a

value near the u pp er end of this range. One solution to the p roblem h ighlighted byDomowitz and Eovaldi (1993) would be to measure the effect on d efau lts (as measured

by loans charged off by banks) rather than bankru ptcies.

Anoth er d ifficulty with examining th e effect of the Bankruptcy Reform Act of

1978 is that there were th ree other major legal chan ges that occurred at abou t the same

time. In 1977 the Sup reme Cou rt ruled that restrictions on advertisements by lawyers

are an un constitutional restriction of free speech,35 thus increasing the spread of

information about the advan tages of filing for bankruptcy. In 1978 the Sup reme Cour t

ruled th at the interest rate paid by a borrow er on a loan from an out of state bank

wou ld be governed by the u sury ceiling of the state in w hich that bank was located.36

This reduced the ability of a state to set effective interest ra te ceilings and increased the

nu mber of high-interest, high-risk loans.37 Both of these events could have stimulated

bankruptcy filings indep end ent of the effect of the 1978 Act. Finally, the Fair Debt

34 See, for examp le, Shep ard (1984), Peterson an d Aoki (1984) and Boyes and Faith (1986). But see

Bhand ari and Weiss (1993).35Bates v. State Bar of Arizona, 434 U.S. 881 (1977).36Marquette Natl Bank of Minneapolis v. First of Omaha Serv. Corp., 439 U.S. 299 (1978).37 Ellis (1998b) does d iscuss the relative impor tance of interest rate ceilings and the Bankrup tcy Reform

Act of 1978. However , by using state usu ry rates prior to 1978, a more rigorou s attempt at d isentangling

the effects migh t be p ossible.


27/52

26

Collections Act, discussed above, was passed in 1977. While this act may have

increased the d efault rate, it should have d ecreased the bankru ptcy rate by enh ancing

the ability of debtors to avoid rep aym ent without filing for bankru ptcy. The fact that

the Fair Debt Collections Act may have had an effect on the ban kruptcy rate that w ould

have conflicted with the presum ed effects of the Bankruptcy Reform Act and th e other

laws is yet another reason to examine the d efault rate rather than the bankru ptcy rate.

D. Third-Party Defenses

Description. When retailers sell prod ucts on credit, they frequen tly resell the debt

to a third pa rty creditor. After this sale, the buyer is obligated to make paym ents

directly to the third pa rty creditor. Historically, this was true even if the contractbetween the buyer and the retail seller was vu lnerable to legal challenge. If, for

example, the buyer p urchases defective goods from a subsequently jud gment-proof

seller, the buyer would not be able to use the sellers breach as a d efense against the

third-party creditors claim for repaymen t of the loan, and wou ld have no rem edy

against the original seller. This outcome was compelled by the holder in due course

doctrine w hen the bu yer signed a negotiable instrum ent, but could easily be obtained

contractually by adding a waiver-of-defense clause to a nonnegotiable instrum ent. The

usefulness of these doctrines for th ird-par ty creditors is now severely restricted by

federal and state law. (Alperin and Chase, 1986)

Effects. The division of labor betw een seller and th ird-par ty creditor clearly has

ad van tages. Each par ty can specialize in developing expertise in its own market. The

third -party d octrines also enhan ce the ability of creditors to reduce region or seller

specific risk by reselling the d ebt, sometimes in large pools as secur itized assets. The

existence of these advan tages is sup por ted by studies show ing a redu ction in the ability

of retailers to obtain financing and in the ability of consumers to obtain cred it in

jurisd ictions that were the first to ban the third -par ty doctrin es (Rohner, 1975).


28/52

27

Opp onents of the holder in due course and negotiability d octrines argue that the

deep-pocketed finan cier can more cheaply bear the risk of breach by the seller than the

buyer can, and further that it can m ore cheaply m onitor sellers and prevent them from

breaching in the first place. When financiers have a continu ing relationship with the

seller, these conditions may be met. But if these conditions are met and the market is

competitive, then a ll three p arties will voluntarily p lace the risk on the finan cier. It is

not necessary for the law to p rohibit the par ties from choosing alternative relationships,

and indeed such a p rohibition w ould redu ce social welfare.

The regulations appear to be based on the assump tion tha t the market fails,

perhap s because of pervasive consumer ignorance, and th at the regulations compel theoutcome that the parties wou ld wan t. This argument assumes that consum ers

irrationally fail to upd ate their beliefs abou t credit practices, even though they app ear to

do so in cash sale contexts, wh ere sellers sup ply w arran ties (for examp le) in order to

attract buyers. Although th is is possible, it seems just as likely that consum ers take

ad van tage of the cost savings perm itted by specialization an d d iversification.

E. Information disclosure

Description. The Truth in Lend ing Act and related state and federal laws require

creditors to pr ovide credit information in a clear and consistent way. These laws app ly

not just to the cred it contract itself, bu t to all comm unications such as advertisements,

bills, respon ses to billing inqu iries, and credit reports. Although the Tru th in Lend ing

Act and the associated regulations are comp lex and impose a n um ber of obligations on

creditors, 38 we w ill discuss two of the primary elemen ts of this act. First, this law

requ ires lenders to clearly present the amou nt finan ced, finance charges, and

annual percentage rate as calculated in a standard ized m anner. Second , the law

38 For example, the Truth in Lending Act regulates the p rocess of correcting billing errors, the credit card

customers liability for unau thorized u se of the card, and so forth. For reasons of space, we d o not deal

with these and other restrictions.


29/52

28

requ ires cred itors taking a security interest in the debtors home to pr ovide an explicit

d isclosure of such security interest and the d ebtors right to rescind the contract w ithin

three d ays (this right m ay be extended to three years if certain d isclosure requ irements

are not met). The Tru th in Lend ing Act provides for enforcement both by regu latory

agencies and by borrowers w ho are given a private right of action (Alperin an d Chase,

1986).

Effects. The stated goals of the Truth in Lend ing Act are to increase econom ic

stability, to enhance the ability of consum ers to shop for attractive loan term s, and to

prevent inaccurate and unfair billing. The first of these goals cann ot be evaluated

emp irically and the last of these goals is similar to the prevention of fraud and hencebeyond the scope of this paper. The second goal is largely consistent with the

discussion of information failure presented above. The stand ard ized calculations

requ ired by the act, the amount finan ced, the finance charge, and the interest rate, are

classic examples of scoring systems and there is some evidence that the Truth in

Lending Act increased consumer aw areness of the terms covered by the act, pa rticularly

the annu al percentage rate.39

Unfortunately, there is also some evidence tha t the beneficial effects of these laws

in enabling consu mers to better shop for attractive loans may h ave been limited to well-

edu cated, affluent borrowers.40 Moreover , a problem comm on to all scoring systems is

that firms are d riven to em phasize the m easured attribute at the expense of hard -to-

measu re attribu tes (Beales, Crasw ell, and Salop, 1981a; 1981b). If consu mers focus

disproportionately on the interest rate, lenders have an incentive to compete over this

term an d p rovide less attractive collection term s or cut back on custom er service. There

is some evidence of this phenomenon: borrower awareness of terms not covered by the

39 Mandell (1971), Brandt and Day (1974), Day and Brandt (1973), Shay and Schrober (1973).40Brandt and Day (1974), Day and Brandt (1973), Deutcher (1973), Mandell (1971), Shay and Schober

(1973), White and Munger (1971).


30/52

29

Truth in Lending Act, such as the d ollar am oun t of the finance charges, actua lly fell

after its passage (Brandt an d Day, 1973).

The required d isclosure of the scores created by the Tru th-in-Lending Act is

more controver sial. These scores are bran d specific information and creditors should

have sufficient incentive to d isclose this information in ord er to ga in a comp etitive

ad van tage. Governmen t regulation may overcome a collective action problem if no

single creditor would have the incentive to invest the resou rces to establish a cred ible

standard . While a period of mandatory d isclosure may be helpful in establishing the

govern men t spon sored scoring system ((Beales, Craswell and Salop 1981b), any furth er

per iod of mandatory d isclosure wou ld seem u nnecessary as typical stories of collectiveaction p roblems stemm ing from brand sp ecific informa tion are inapp licable.42 Of

course, we noted above that creditors with m arket pow er may wish to conceal private

information in ord er to engage in price d iscrimination, bu t there is little evidence that

they have such market power.

The requiremen t that creditors provide special disclosure (accompanied by a

right of rescission) of any security interest taken in the home is a better examp le of

mand ated d isclosure. Creditor obviously has no incentive to inform Debtor of the legal

consequences of a security interest and to d isclose the right of rescission, and his

competitors may have insu fficient incentive to disclose them as well (see above). We

note, however, that the traditional argum ent for mand ated d isclosure wou ld seem to

encomp ass much broad er d isclosure of the legal consequences of failing to pay a debt

that w hat is required by the Act. If debtors do not know abou t the effects of security

interests, they are not likely to know about the holder in d ue course d octrine or the

41 Beales, Craswell and Salop (1981b) do concede th at such requ ired d isclosure m ay be necessary for an

introductory period.42 Securities law, for example, requ ires issuers of securities to reveal a great d eal of financial and business

information. A pop ular explanation for this requiremen t is that issuers fear that if they prov ided


31/52

30

right of redemption. The difficu lty is that too much d isclosure of techn ical information

may overw helm d ebtors and cause them to ignore it (cf. Beales, Craswell, and Salop,

1981b).

The most forceful complaints abou t the Truth in Lending Act have centered on

the cost of compliance. In addition to the adm inistrative costs of compliance, the Tru th

in Lending Act may have red uced the ability of creditors to collect on bad loans since a

determined debtor can almost certainly find some fault with the disclosure by the

creditor (Rubin, 1991). The act has been amen ded severa l times, most recently in 1995,

in part to add ress this complaint. While there is limited su rvey eviden ce that the

difficulty in comp liance red uced creditors willingness to advertise and risk violation(Angell, 1971), we know of no p apers assessing the effect of this law on creditors

willingn ess to lend .

F. Antid iscrimin ation laws

Description. The Fair Hou sing Act, as amend ed (FHA ), forbids creditors to

discriminate against ap plicants for hom e mor tgage loans on the basis of race, color,

religion, sex, national origin or han dicap or family status. The Equal Credit

Opp ortunity Act, as amended (ECOA ), forbids them to d iscriminate against applicants

for credit generally on similar, though not identical grounds. Although not technically

an antidiscrimination statute, the Home Mortgage Disclosure Act, as amended,

enhances the ability to test for d iscriminat ion by requ iring finan cial institu tions to

report da ta on all of their app licants for hom e mortgages, including the race of the

applicant.

Perhap s the m ost significant an tidiscrimination statu te is, by its express terms,

not an antidiscrimination statute at all. The Commun ity Reinvestm ent Act (CRA )

adequate information to their investors, this information would also be revealed to their competitors, but

all investors and issuers wou ld be better off if adequate inform ation wer e revealed (Mahoney, 2001).


32/52

31

requ ires the ap propriate federal banking regu lators to "encourage . . . institutions to

help m eet the credit needs of the local comm unities in w hich they are chartered

consistent w ith the safe and sound operation of such institutions. H owever, the CRA

was largely justified on the groun ds of perceived d iscriminat ion and interpretations by

regu latory agencies refer to minority grou ps (Hy lton and Rougeau , 1996).

Effects. There is an extensive literature on th e role of d iscrimination in lend ing

markets and a full review is beyond the scope of this paper.44 There is clear ev idence of

historical discrimination in lending m arkets, often sup ported by overt governm ent

policy, but there is no consensus as to whether d iscrimination still plays a significant

role in credit markets, wheth er it p lays a role in some credit markets like the mortgagemarket bu t not others,45 and wh ether such d iscrimination that exists is based on animus

or the use of race as a statistical proxy for credit risk (Hylton and Rougeau , 1996; 1999;

Swire, 1995). To und erstand the d ifficulty of evaluating the law s against d iscrimination,

suppose that d iscrimination is due to the use of proxies. On the one hand , a prohibition

of the use of statistical discrimination m ay force creditors to expend resour ces to try to

distingu ish between d ebtors and may exacerbate asymmetric information problems.

On th e other hand , statistical d iscrimination m ay cause minorities to und erinvest in

hu man capital and th e development of a credit history, in anticipation of being den ied

credit on accoun t of their race (Hy lton and Rougeau , 1996).

There have been few successful suits brough t und er either the FHA or theECOA

(Swire, 1995), and therefore there has not been mu ch academ ic debate concerning these

laws. By contrast, the CRA has been controversial: many h ave argu ed th at it is costly,

possibly self-defeating , and at best ineffective. The CRA generates significant

compliance costs only for those banks that have branches in low income areas and thus

44 Good su rveys can be found in Hylton and Rougeau (1996) and Swire (1995).45 Partly because of the data generated by the Hom e Mortgage Discrimination Act, emp irical studies of

discriminatory lending focus on the mortgage m arket rather than other segments of the credit market.


33/52

32

may d iscourage large banks from serving low-income areas (Macey and Miller, 1995)

and discourage the d evelopm ent of small banks to serve low-income d ebtors (Hylton

and Rougeau , 1999). Hylton and Rougeau (1996; 1999) argue tha t the curren t

enforcement approach encourages hollow comp liance in the form of loans to wealthy

developers op erating in low-income neighborhood s or agreements d esigned solely to

appease p oliticians and political activists, and rent seeking beh avior by p oliticians,

interest group s, and even rival banks trying to block bank mergers. Finally, Schill and

Wachter (1995) argu e that by targeting the location of the investmen t, the CRA an d

related laws may encourage concentration of poverty in u rban a reas.

In the end , there are plausible arguments for and against the CRA and its neteffect remains un resolved. Commentators agree that the CRA needs substantial reform,

but they d isagree strongly as to the d irection this reform shou ld take with some calling

for safe harbor provisions or a switch to a subsidy system an d oth ers calling for more

vigorous enforcement.

Conclusion

Regulation of the market for consu mer cred it provides a nu mber of benefits to

consum ers. It gives them information about the terms and consequ ences of the credit

transaction, it provides them insuran ce against shocks, and it protects them from

discrimination. But a prop er defense of consum er credit regu lation mu st explain why

the market wou ld not supply these benefits if consum ers are willing to p ay for them.

The availability of credit insu rance, the many ways in which typical credit transactions

trad e off between interest rate and risk, and the existence of information intermediaries

all suggest that the market does respond to some degree to consum er demand for credit

protections.

Models that incorporate information asymm etry and m arket power have

ambiguou s implications for consumer credit regulation. Information problems do

prevent m arkets from achieving the first best, and laws regu lating the cred it market can


34/52

33

in theory increase social welfare. But it is difficult to determ ine wh ether the prem ises of

the models are met in reality. Complicating the analysis, it is not clear how sensitive

consum ers and cred itors are to the law, whether because of irrationality or rational

ignorance. And it is not clear how mu ch the law would influence the behavior of even

a rational, well-informed consumer , given the many loopholes, the limited p enalty

structures, and the man y ways in w hich creditors can evade the law an d creditors and

debtors can contract around it.

APPENDIX

This appen dix sets forth a simp le examp le of how a law that solves a failure of thecredit market could, in theory, result in a d ecline in total borrow ing.

Assum e that there is a d ebtor w ith per p eriod u tility U() where U>0 and U


35/52

34

)))1(

)((()1()()(

p

ELpBHUpEpUBU

++

or, the d ebtor w ill set

))

)1(

)(((')('

p

ELpBHUBU

=

In this very simple example, the amou nt borrow ed is always decreasing in the exemp tion. The reason is

that the debtor seeks two things: low cost credit, and insu rance. The lower exemp tion wh ich in this

example is set by contract rather th an by statu te reduces the cost of credit but also reduces the amou nt

of insurance. If the latter effect dominates (as in this example), an op timal exemption resu lts in less

bargaining than a less generous exemption. This example is deliberately contrived ; it assum es that the

probability of default is ind ependent of the amount borrow ed, and only considers how the exemp tions

affect borrow ing throu gh a change in the interest rate. If a debtor is reluctant to borrow a certain amou nt

because he m ay end up in a very p ainful d efault, the exemptions could increase borrowing by lessening

that fear. This effect is not present h ere because a marginal change in borr owing has no effect on theprobability of default and never redu ces consumption w hen the m arginal utility of consump tion is higher

than it is in p eriod one. A more general mod el would show that if both factors are considered, an

increase in total borrowing is sufficient to show that a law is efficient but is not necessary.


36/52

35

References

Adler, Barry, Ben Polak, and Alan Schw artz. 2000. Regulating Consumer Bankruptcy:

A Theoretical Inqu iry, 29Journal of Legal Studies 585-614.

Aghion, Phillippe, and Benjamin H ermalin. 1990. Legal Restrictions on Private

Contracts Can Enhance Efficiency, 6Journal of Law, Economics, & Organization 381-409.

Akerlof, George A. 1970. The Market for Lemons: Quality Uncertainty and the Market

Mechanism, 84 Quarterly Journal of Economics 488-500.

Allen, Franklin. 1983. Credit Rationing and Paym ent Incentives, 50Review of Economic

Studies 639-46.

Alperin, How ard J., and Roland F. Chase. 1986 and 2000 Sup plement. Consumer Law:Sales Practice and Credit Regulation. St. Pau l, MN: West Publishing Comp any.

Angell, Fran k J. 1971. Some Effects of the Truth-In-Lend ing Legislation, 44Journal of

Business 78-85.

Apilado, Vincent P., Joel J. Dauten, and Dou glas E. Smith. 1978. Personal

Bankruptcies, 7Journal of Legal Studies 371-92.

Austin, Dou glas V., and David A. Lind sley. 1976. Ohio Usu ry Ceiling and ResidentialReal Estate Development, 4(1)American Real Estate and Urban Economics Association

Journal 83-95.

Ausubel, Lawrence M. 1991. The Failure of Competition in the Credit Card Market, 81

American Economic Review 50-81.

Avio, Kenneth L. 1973. An Economic Rationale for Statu tory Interest Rate Ceilings,

13(3) Quarterly Review of Economics and Business 61-72.

Barth , James R., Padma Gotur , Neela Man age, and Anth ony M. J. Yezer. 1983. The

Effects of Governm ent Regu lations on Personal Loan Markets: A Tobit Estimation of a

Microeconomic Model, 38Journal of Finance 1233-51.

Barth , Jam es R., Joseph J. Cord es, and Anthony M. J. Yezer. 1986. Benefits and Costs of

Legal Restrictions on Personal Loan Markets, 29Journal of Law and Economics 357-80.


37/52

36

Beales, Howard , Richard Craswell, and Steven C. Salop. 1981a. Informa tion Remed ies

for Consum er Protection, 71American Economic Review 410-13.

Beales, Howard , Richard Craswell, and Steven Salop. 1981b. The Efficient Regu lation

of Consum er Information, 24Journal of Law and Economics 491-539.

Bebchu ck, Lucian Arye, and Jesse M. Fried. 1996. The Uneasy Case for the Pr iority of

Secured Claims in Bankruptcy, 105 Yale Law Journal 857-934.

Berkow itz, Jeremy, and Richard Hynes. 1999. Bankruptcy Exemptions and the Market

for Mortgage Loans, 42Journal of Law and Economics 809-30.

Berkowitz, Jerem y, and Michelle J. White. 2000. Bankru ptcy and Small Firms Access to

Cred it, Working Pap er, No. 00-005. Ann Arbor, MI: Law School, University of

Michigan.

Besanko, David and Anjan V. Thakor. 1987. Collateral and Rationing: Sorting

Equilibria in Monopolistic And Competitive Credit Markets, 28International Economic

Review 671-689.

Bester, Helmu t. 1985. Screening vs. Rationing in Credit Markets w ith Imp erfect

Information, 75American Economic Review 850-855.

Bester, Helmu t. 1987. The Role of Collateral in Credit Markets w ith Imp erfectInformation, 31European Economic Review 887-99.

Bhandari, Jagdeep S. and Lawrence A. Weiss. 1993. The Increasing Bankruptcy Filing

Rate, An H istorical Analysis, 67American Bankruptcy Law Journal 1-15.

Boyes, William J. 1982. In Defense of the Dow ntrodden, Usu ry Law s?, 39 Public Choice

269-76.

Boyes, William J. and Roger L. Faith. 1986. Some Effects of the Bankru ptcy Reform Act

of 1978, 29Journal of Law and Economics 139-49.

Boyes, William J. and Nancy Roberts. 1981. Econom ic effects of usu ry law s in

Arizona, 1981Arizona State Law Journal 27-59.


38/52

37

Brandt, William K. and George S. Day. 1974. Information Disclosure and Consumer

Behav ior: An Empirical Evaluation of Truth -in-Lend ing, 7 University of Michigan

Journal of Law Reform 297-328.

Buckley, F.H., and Margaret F. Brinig. 1996. The Market for Deadbeats, 25Journal of

Legal Studies 201-32.

Buckley, F.H., and Margaret F. Brinig. 1998. The Bankruptcy Puzzle, 27Journal of Legal

Studies 187-207.

Calder, Lendol. 1999. Financing the American Dream: A Cultu ral History of Consumer

Credit. Princeton , NJ: Princeton University Press.

Caplovitz, David. 1967. The Poor Pay More; Consumer Practices of Low-Income Families.

New York: Free Press of Glencoe.

Chakravarty, Sugato, and Eun-Young Rhee. 1999. Factors Affecting an Individuals

Bankruptcy Filing Decision, Working Paper . West Lafayette, IN: Purdue University.

Crafton , Steven.M. 1980. An Emp irical Test of the Effect of Usury Law s, 23Journal of

Law and Economics 135-46.

Craswell, Richard. 1995. Freedom of Contract, in Chicago Working Papers in Law and

Economics (Second Series). Chicago, IL: Law School, University of Chicago.

Day, George S., and William K. Brandt. 1973. A Stud y on Consumer Credit Decisions:

Implications of Present and Prospective Legislation, in 1National Commission on

Consumer Finance., Technical Studies. Washington , D.C.: U.S. Governm ent Printing

Office.

DeMuth, Christopher C. 1986. The Case Against Credit Card Interest Rate

Regulation, 3 Yale Journal on Regulation 201-42.

Deutcher, Terry. 1973. Credit Legislation Two Years Ou t: Awareness Changes and

Behavior Effects on Different Awaren ess Levels, in 1National Commission on Consumer

Finance, Technical Studies. Washington , D.C.: U.S. Government Pr inting Office.

Domowitz, Ian, and Thomas Eovaldi. 1993. The Imp act of the Bankruptcy Reform Act

of 1978 on Consum er Bankruptcy, 36Journal of Law and Economics 803-35.


39/52

38

Domowitz, Ian, and Robert L. Sartain. 1999a. Determinants of the Consum er

Bankruptcy Decision, 54Journal of Finance 403-20.

Domowitz, Ian, and Robert L. Sartain. 1999b. Incentives and Bankruptcy Chap ter

Choice: Eviden ce from the Reform Act of 197

law and economics of consumer finance

Documents