Interest rate ceilings and the role of security and collection remedies in loan contracts.

Manage, Neela D.

NEELA D. MANAGE

Collateral or other security for personal loans and restrictions on creditor remedies for the collection of debts have varying effects on the price and quantity of credit which depend in turn on the level of interest rate ceilings. We report here on reduced-form equations of a supply-demand model estimated for five states with different interest rate restrictions. Interest rate ceilings limit how far lenders can raise loan rates to compensate for expected default losses but restrictions on collection remedies are generally associated with a higher interest rate.

I. INTRODUCTION

The role of collateral in loan contracts has been the focus of many recent studies on credit markets. This paper examines the role of collateral in personal loan contracts when government regulations, such as usury laws and restrictions on late charges and wage garnishment limit the responses of creditors to increases in default risk. Special attention is given to the use of the borrower's future income as a form of collateral in personal loan markets and the effect of laws governing garnishment and wage assignment which limit the lender's claim to the borrower's income.

Theoretical studies by Barro [1976], Benjamin [1978], and Harris [1978] discuss the effects of collateral on the interest rate, loan size and other features of the loan contract, particularly when the collateral has a different value to the borrower and to the lender. More recent studies by Rea [1984], Schwartz [1981; 1984], Hess [1985], Eaton [1986], and Hess and Knoeber [1987] further analyze the role of collateral in loan contracts. Hess [1985], in particular, evaluates the earlier work done by Benjamin [1978] and develops an excellent analysis of the relation between collateral and other terms of the loan such as loan rates and loan size.

The regulations governing personal loan markets that have been most frequently analyzed are interest rate ceilings and restrictions on the remedies available to creditors attempting to collect on delinquent or defaulted accounts. Empirical studies by Shay [19701, Benston [1973; 1977], Greer [1973; 1975], Aho et al. [1979], Barth and Yezer [1977], Barth et al. [1983], Manage [1983), and Peterson [1979] analyze several personal loan market relationships. These studies differ widely in terms of scope, type of data used (with respect to both geographical coverage and level of aggregation), model specification, estimation technique, and the specific hypothesis being tested.

This paper demonstrates that the effect of interest rate ceilings depends on the restrictions imposed by other financial regulations and visa versa, and that the effect of regulation, in turn, depends on the specifics of the loan transaction. Both interest rate restrictions and a number of other factors affecting borrowing and lending decisions vary systematically by state. First, regulations governing loan contracts differ from state to state. Interest rate ceilings are set by each state and different laws govern the use of garnishment and wage assignment, so that the amount of future income that may serve as collateral also varies. Creditors in different states thus operate under different rules. Second, there are differences in the risk characteristics of borrowers (due to demographic, financial and institutional factors) which affect the supply decisions of lenders. In order to hold these factors constant, this paper analyzes cross-sections of loans made within a state, rather than use a cross-section of states. This approach differs from previous studies in that it allows regression coefficients to vary across states.

II. COLLATERAL AND GOVERNMENT REGULATIONS

Collateral is used as an instrument to enforce loan contracts. It provides an incentive to repay the loan because the borrower loses the collateral in the event of default. Theoretical studies generally suggest that a decrease in collateral will increase the costs to the lender of a default and increase the interest rate on the loan.

Barro [1976] points out that transactions costs, collection costs, and any costs involved in selling the collateral asset cause the lender to value the collateral less than the borrower does. When default occurs, deadweight losses are imposed on the economy, which are then incorporated into interest rates. Barro's [1976] analysis shows that the explicit interest rate on a loan rises as the ratio between the value of the loan and the value of the collateral increases. Harris [1978] demonstrates that an increase in the value of the collateral lowers the deadweight costs associated with default. It also causes a decrease in the equilibrium probability of default and a decrease in the equilibrium interest rate on loans. Benjamin's [19781 approach is quite similar to that of Barro, except that he makes a distinction between tradable and non-tradable collateral assets. According to Benjamin, collateral functions in two ways: it either increases the cost to the debtor or lowers the cost to the creditor of a default. Any factor which increases the probability of default increases the interest rate at which lenders are willing to supply a given loan amount.

Hess [1985) recently modified Benjamin's analysis by assuming the competitive lender is a price-taking profit maximizer. Under these conditions the lender's supply curve cannot be the locus of loan amounts and interest rates along which expected profits are zero. Hess argues that in the short run the lender is likely to supply credit so that profits are maximized. In the long run, however, expected profits are driven to zero.

Based upon these assumptions, Hess derives the relationship between credit supply and its determinants (cost of funds, loan rate, probability of default, and value of collateral). In the Hess model, an increase in the cost of funds decreases the supply of credit. Unlike Benjamin's model, that derived by Hess implies that the backward bending segment of the supply curve may be economically relevant, at least in the short run, and that an increase in the probability of default does not necessarily shift the supply curve to the left. Finally, when collateral increases in value, the creditor makes larger loan offers at any given interest rate (i.e. credit supply increases).

In sum, although there are important differences between the underlying assumptions of the models surveyed, all the studies cited here suggest that a decrease in the value of collateral raises loan interest rates. It is important to note that when the long-run equilibrium condition, that the maximum expected profit per customer is zero, is imposed in the Hess model, the value of collateral adjusts until long-run expected profits are driven to zero. Schwartz [1981; 1984] also considers the issue of why changes in collateral are a substitute for changes in interest rates and Rea [1984] has investigated the endogeneity of arm breaking in loan contracts. Recent literature therefore suggests that, in the context of long-run equilibrium, collateral is endogenous in the loan model.

Consider the credit demand and supply relationships existing at the level of an individual loan transaction. Briefly, the individual consumer's demand for a personal loan is assumed to be a function of the interest rate, the borrower's current and future income, non-price loan terms, and any personal losses that may be incurred by the borrower if he defaults. The individual lender's supply of a personal loan is assumed to be a function of the interest rate, the borrower's current and future income, the opportunity cost of capital, collection costs incurred by the lender if the borrower defaults on the loan, claims on the borrower's income or assets, and, more generally, the overall default-risk of the borrower. The borrower's default risk is not directly measurable but is assessed by the lender on the basis of information about the borrower, such as his current and future income, outstanding debts, age, marital status, and previous credit performance. These basic demand and supply relationships may be summarized as follows:

(1) Demand: D = D(r, CY, FY, PL)

(2) Supply: S = S(r, CY, FY, O, CC, BC)

where D= loan demand, S= loan supply, r= interest rate on the loan, CY = borrower's current income, FY = borrower's future income, PL = personal losses of the borrower in default, O = opportunity cost of capital, CC = collection costs incurred by the lender and BC = lender's claims on borrower's income or assets in the event of default.

An important feature of the market for personal loans made by finance companies is that it is heavily regulated, and borrowing and lending decisions are likely to be significantly affected by these regulations. Usury laws governing the maximum interest rates lenders may charge and restrictions on the methods of collection available to lenders are the two most important sets of regulations and are thus considered in this paper.

As interest rate ceilings become more restrictive, lenders are willing to assume only a lower degree of risk. They are likely to respond by making smaller loans to any one individual and rejecting loan applications of high-risk individuals. Lower ceilings thus eliminate higher risk-and hence higher interest rate-loans, thus making observed loan rates lower.

Regulations that restrict, and in some cases prohibit, the use of creditor remedies may also be conveniently analyzed within the supply and demand framework outlined above. More specifically, creditor remedies are legal provisions that generally give the lender a claim on part of the borrower's income or impose additional charges on the borrower. One form of remedy in personal loan markets which is regulated is the amount of future income that may be garnished or assigned to the lender in the event of default. To the extent garnishment and wage assignment increase the cost to the debtor and lower the cost to the creditor, they perform a function quite similar to other forms of collateral. In contrast, the inclusion of late charges or attorney's fees in the loan contract primarily lower collection costs in the case of default.

Restrictions on creditor remedies are likely to affect both the demand for and the supply of loans. Such restrictions either limit the ability of lenders to collect the principal and interest payments on defaulted loans or increase the collection costs incurred by lenders when trying to collect a loan. This increases default risk and is likely to decrease the supply of credit (i.e. lenders would be willing to supply a smaller loan amount at a given interest rate or a given loan amount at a higher interest rate), The influence of restrictions on the demand for credit, on the other hand, results from their lowering the costs of default to the borrower. Such restrictions cause an increase in the demand for credit because the borrower can anticipate a reduction in his personal losses, or a lower liability, in case he defaults. Borrowers may thus be willing to borrow a larger amount at a given interest rate or pay a higher interest rate for a given loan amount to avoid the inconvenience of legal collection procedures.

In sum, restrictions on creditor remedies tend to increase the demand for credit while decreasing the supply of credit, and so tend to increase the equilibrium interest rate. The impact on the equilibrium loan amount, however, depends on the relative magnitude of the changes in supply and demand. If borrowers are not well informed about the extent of creditor remedies available to lenders, their demand for credit will not be significantly affected by such restrictions and regulations restricting creditor remedies would primarily affect supply, resulting in a reduction in the equilibrium loan amount.

It was pointed out above that restrictions on creditor remedies tend to decrease the supply of credit. However, when low interest rate ceilings are in effect, lenders will be unable to raise interest rates (beyond the ceiling rates) to compensate for expected default losses. In such a situation lenders may respond to restrictions on garnishment, wage assignment, and late charges by offering stringent non-price terms of credit, thereby increasing the effective supply price of credit. On the other hand, if rate ceilings are high and market rates are substantially below the ceiling, then restrictions on these creditor remedies may exert substantial upward pressure on loan rates. In fact they would have the potential of increasing loan rates all the way up to the level of the ceiling. Therefore, the hypothesis to be tested in this paper is that the effect of security (e.g., a loan cosigner or collateral) and assigned or garnished income on loan interest rates varies with the upper limit on legally allowable interest rates applicable to the loan. As regards the loan size, the impact of such restrictions depends on the relative magnitudes of the shifts in supply and demand.

III. DATA DESCRIPTION, METHODOLOGY, AND EMPIRICAL SPECIFICATION

Data Description

The data set was compiled by the National Consumer Finance Association (NCFA) and covers the period January 1975 through August 1977. It consists of a random sample of active personal loan accounts from the master files of nine major finance companies in different states. This data set provides information on various borrower characteristics, size and maturity of the loan, interest rate charged, and the type of collateral. Personal loan transactions constitute a large portion of consumer finance activity.

Methodology

Section II described the possible effects of state regulations governing interest rate ceilings and the exercise of creditor remedies. The degree of restrictiveness of these regulations was the basis used for categorizing the states. The number of states suitable for individual estimation was substantially narrowed by excluding those which did not have a sufficiently large number of sample observations for individual estimation. Furthermore, states which did not reveal sufficient variation in the explanatory variables (e.g. some of the 0-1 dummy variables took the same value for all observations) were not considered. States were selected to represent: (1) states with a restrictive interest rate ceiling structure (2) states where interest rate ceilings are moderate, and (3) states where interest rate ceilings are not restrictive.

An unambiguous ranking of states by interest rate ceilings is not possible since rate ceilings typically vary by loan size and maturity. It was decided to initially compare the ceiling rates on loans ranging from $1000 to $2000 since the average loan size observed in the sample was in this range. Five states-Wisconsin, Louisiana, Illinois, Massachusetts, and New Jersey-were selected to represent states with varying restrictions on interest rate ceilings and creditor remedies.

For these five states the rate ceilings were further compared for various loan sizes between $200 and $2000 since over 70 percent of the loans in these states were in this range. Wisconsin was selected to represent states with a severely restrictive interest rate ceiling structure. Louisiana represents those states where interest rate ceilings are not restrictive. Illinois also has high interest rate ceilings, but when compared to Louisiana has less restrictive creditor remedies, so it too was included. Massachusetts was included as another low ceiling state. Finally, New Jersey was chosen to represent those states with a medium interest rate ceiling structure. The average loan size in each of these five states was between $1000 and $2000, and the average maturity of the loans was approximately thirty-six months. The interest rate ceilings in these states for various loan sizes ranging from $200 to $2000 and a maturity equal to thirty-six months. Empirical Specification

Reduced form equations based on the supply and demand relationships given by equations (1) and (2) were estimated by ordinary least squares for each of the five states. The reduced form equations were estimated with both the interest rate and loan amount as the dependent variable. The interest rate is measured by the annual percentage rate appearing in each contract and the loan amount is measured by the amount financed in the contract. It is possible that the amount of collateral on the loan is a contractual term like the interest rate. If so, the presence of an endogenous variable in the equation might introduce some simultaneity bias in the coefficient estimates.

The regressors in the reduced form equations measure the exogenous determinants of loan demand and supply. These are divided into three groups: (a) borrower characteristics, (b) legal/regulatory variables, and (c) general economic conditions.

Borrower Characteristics. The following variables are used to measure current income, CY, and future income, FY: current disposable income, the borrower's outstanding debt, the age of the borrower, marital status of the borrower, and borrower class which is identified by two 0-1 dummy variables-for former borrowers and for current borrowers. Both linear and squared terms for the current income, debt, and age variables are included in the regression equations to allow for possible non-linear effects (based upon a life-cycle hypothesis) of these regressors on the dependent variables.

The data set provides partial measures of the extent of claims lenders had on borrowers' income and of the potential collection costs in the event of default. Two 0-1 dummy variables were included to measure the security on the loan. These variables indicate the presence of a cosigner and whether or not the loan was secured. In those cases where the information was available, three 0-1 dummy variables are used to indicate the type of collateral: an automobile, household goods, or a major consumer durable including a mobile home or real property. The collateral and cosigner variables are expected to be associated with additional claims on the borrower's assets, BC, or lower collection costs for the tender, CC, and with higher potential losses for the borrower in the case of default, PL.

Legal/Regulatory Variables. Legal provisions for debt collection also influence creditor costs and borrower liability in the event of a default. Possible creditor remedies are garnishment, wage assignment, late charges, and attorney's fees. The inclusion of these legal variables is based on legal research, relying mainly on the Consumer Credit Guide [1976], and the Cost of Personal Borrowing in the United States [1977]. Three continuous economic variables are calculated for each loan to quantify the effects of three creditor remedies-garnishment, wage assignment, and late charges. Such continuous variables are appropriate because the quantitative impact of these legal provisions for debt collection on borrower liability and credit supply differs for each and every loan. For example, the maximum amount of a person's income that may be garnished varies with the earnings of the borrower, and the maximum amount that may be assigned to the lender is a complex function of various characteristics of the borrower, including his income,

Two variables are calculated for each loan to measure the maximum amount of the borrower's monthly income that may be garnished or assigned to the lender. Based on these two continuous variables, a third variable, defined as the larger of the garnishment and wage assignment variables, was calculated for each loan. This variable is the maximum monthly payment for which the borrower is liable under garnishment or wage assignment. The economic effects of late charges, which increase the amount that may be collected in case of delinquency or default, are measured by a continuous variable equal to the legal maximum late charge collectable if the first loan payment were twenty days late. The value of the late charge variable also differs for each loan. The impact of attorney's fees is measured by a 0-1 dummy variable indicating a provision for attorney's fees in the contract. Finally, the interest rate ceiling associated with each loan was determined on the basis of the loan size, term-to-maturity, and the type of loan law governing the transaction.

Fewer restrictions on the use of creditor remedies will increase the maximum amount of monthly income that may be garnished or assigned to the lender. This implies higher personal losses for the borrower, PL, in the event of default and by increasing the lender's claim on the borrower's income, adds security to the loan. The maximum amount of income that may be garnished or assigned to the lender thus provides a partial measure of PL and BC in equations (1) and (2). Provisions for late charges and attorney's fees in the contract lower collection costs, CC, to the lender in the event of delinquency or default.

General Economic Conditions. The opportunity cost of capital variable, 0, is measured by the finance company paper rate for the month the loan was made. Two measures of statewide economic conditions, per capita income and new unemployment insurance claims, are included on the grounds that these conditions may be indicative of a borrower's future income and thus influence the supply of credit.

IV. EMPIRICAL RESULTS

The reduced-form equations are estimated for each of the five states, and both the interest rate and loan amount are used as the dependent variables. For each equation three different specifications are employed. The first model includes only the economic variables that influence credit demand and supply. The second specification adds the legal variables-interest rate ceilings, late charges, presence of attorney's fees in the contract, and the maximum amount of monthly income that may be garnished or assigned to the lender-to the economic variables. The final model adds (to the economic and legal variables) two cross-product terms which capture the interaction between interest rate ceilings and late charges and between interest rate ceilings and garnishment or wage assignment. These cross-product terms are included since the impact of these creditor remedies may vary with the level of the interest rate ceiling. The test of the null hypothesis that the creditor remedy has no significant impact on the interest rate and loan amount involves the use of a F-test to determine whether the coefficient of the creditor remedy variable and the coefficient of the cross-product (of the rate ceiling and creditor remedy variable) are both equal to zero.

The discussion of the estimated equations will focus on the legal and regulatory variables since these are most important for testing the hypothesis. Furthermore, some of these regulatory variables determine the amount of the borrower's future income that may serve as collateral on the loan. Collateral is measured by a series of 0-1 dummy variables which indicate the presence or type of collateral goods and other security on the loan and by the garnishment variables specifying the amount of future income that may be garnished or assigned to the lender in case of default.

The impact of collateral and collection remedies on the interest rate and loan amount in high, medium, and low ceiling states. It also reports the F-statistics and t-statistics for tests of the impact of garnishment, late charges, attorney's fees, a cosigner and collateral on loan contracts. The empirical estimates of the reduced-form coefficients of the determinants of loan interest rates and loan amounts for Wisconsin.

Summary of Basic Findings

The basic findings are summarized here in order to serve as a guide to a more detailed discussion of the empirical results.

1. Low ceiling states: Decreases in collateral or collection costs had no significant influence on interest rates. However, dollar loan size increased with collateral and legal provision for lowering collection costs, such as attorney's fees or late charges.

2. Medium ceiling states: Decreases in collateral had a significant influence on interest rates but not on the amount financed. On the other hand, the provision for late charges influenced the loan amount but had no effect on the loan interest rate.

3. High ceiling states: Decreases in collateral or collection costs generally had a significant effect on interest rates. The presence of collateral and legal provision for lowering collection costs influenced the dollar loan size.

Determinants of the Interest Rate

The empirical results suggest that the impact of security and collection remedies on personal loan interest rates varies with the restrictiveness of the legal interest rate ceiling. In the case of Wisconsin and Massachusetts, the low ceiling states, collateral variables (including income subject to garnishment) as a group do not have a statistically significant impact on loan interest rates. Late charges and attorney's fees, which lower collection costs, also do not reveal any impact on loan interest rates in these states. The empirical results suggest that lenders in Wisconsin and Massachusetts are unable to raise interest rates above the ceiling to compensate for expected default losses associated with restrictions on garnishment, wage assignment, late charges, and attorney's fees. These results strongly support the hypothesis that a decrease in collateral or restrictions on creditor remedies do not affect interest rates when low interest rate ceilings are in effect.

As interest rate ceilings rise, however, such restrictions are generally associated with a higher equilibrium interest rate. In the case of New Jersey, the amount of income that may be garnished or assigned to the tender is an important determinant of the equilibrium interest rate. However, restrictions on late charges do not influence interest rates in New Jersey. Collateral in the form of durable goods or real property has a negative impact on interest rates and the collateral variables as a group have a significant impact on the interest rate in New Jersey.

In spite of the fact that both Illinois and Louisiana are high ceiling states, there are dramatic differences in the effect of state regulations affecting collateral and collection costs on interest rates. Collateral and collection costs are important determinants of loan interest rates in Illinois but have no effect on loan rates in Louisiana. In Illinois, collateral in the form of durable goods or real property and the presence of a cosigner lowers the interest rate. The estimated coefficients also suggest that a decrease in the maximum amount of income that may be garnished or assigned to the lender increases the loan interest rate as long as the interest rate ceiling is above 21 percent. This seems plausible in the case of Illinois where interest rate ceilings are high. The results for Illinois are consistent with prior expectations that the use of collateral is effective when interest rate ceilings are not restrictive. On the other hand, although an unsecured loan implies upward pressure on interest rates in Louisiana (for the specification without the legal variables), collateral variables as a group do not affect interest rates. The ineffectiveness of garnishment, wage assignment, late charges, and attorney's fees in influencing interest rates in Louisiana implies that factors other than rate ceilings have an important influence on borrowing and lending activity. The results also suggest the possibility that some other regulations might interfere with the use of collateral in loan contracts.

Borrower characteristics such as income, age, marital status, outstanding debts, and previous credit performance affect the risk of default and therefore credit supply and interest rates. These variables reveal varying degrees of importance across states. In Wisconsin, for example, lenders consider married individuals less risky than single individuals thus increasing the supply of credit. Former borrowers who have successfully repaid loans tend to be associated with lower risk and therefore lower interest rates in Massachusetts. If one assumes that lenders consider present borrowers no less risky than former borrowers, since they may be having difficulties and merely consolidating outstanding debt, they would be associated with higher interest rates. Such an effect is observed for interest rates in New Jersey. The coefficients for the outstanding debt of the borrower (and its square term) indicate that in the relevant range of the sample (i.e., at the mean value of this variable in the state), debt has a significant negative impact on interest rates in Massachusetts, Illinois and New Jersey. An increase in the borrower's age lowers the interest rate over the loan sizes in New Jersey. The implication is that the non-linear effects of the borrower's age are consistent with the life-cycle hypothesis of consumption and saving.

Determinants of the Loan Size

The discussion in Section II implied that a decrease in collateral decreases credit supply and that state regulations which restrict the use of creditor remedies by lenders decrease credit supply and increase credit demand. The equilibrium loan amount will decrease, however, if such regulations primarily affect supply and if borrower's demand is not sufficiently affected by such regulations.

In the basic model, which includes no legal variables, unsecured loans tend to lower the equilibrium loan amount in Wisconsin, Massachusetts, Illinois, and Louisiana, and collateral in the form of automobiles or household goods tends to increase the loan amount in New Jersey, When the legal variables are added to the economic variables, the collateral variables as a group have a statistically significant impact on the loan amount only for Wisconsin, Illinois and Louisiana. For these three states the coefficients on garnishment and its crossproduct term imply that an increase in the maximum amount of income that may be garnished or assigned to the lender increases the equilibrium loan amount. Such a result implies that restrictions on garnishment and wage assignment have a stronger impact on credit supply than credit demand.

Collection costs are important in explaining the equilibrium loan size in all five states. The coefficients indicate that increases in late charges generally increase the amount financed in all five states. The attorney's fees variable was not included in Louisiana and New Jersey, but the provision for attorney's fees in loan contracts had a significant impact on the dollar loan size for Illinois and Massachusetts.

Among the borrower characteristics, the marital status has a significant impact on the loan amount in the case of Massachusetts, Illinois, and Wisconsin. The borrower's age and outstanding debts are significant explanatory variables in the case of Illinois and New Jersey , and former borrowers imply lower risk and increase the equilibrium loan amount in New Jersey.

V. SUMMARY AND CONCLUSIONS

This paper examines the impact of selected state regulations governing the use of collateral and collection remedies on personal loan markets in five states having different interest rate restrictions. The effect of these regulations was analyzed within a simple framework of supply and demand for personal loans. Reduced-form equations with interest rate and loan amount as dependent variables were estimated as a function of key economic and legal variables.

The empirical results suggest that the effect of security for loans and restrictions on creditor remedies on the price and quantity of credit varies with the height of the legal interest rate ceiling. When interest rate ceilings are low, lenders are unable to raise loan rates above the ceiling to compensate for expected default losses associated with legal restrictions on garnishment, wage assignment, late charges, and attorney's fees. As interest rate ceilings rise, however, such restrictions are generally associated with a higher equilibrium interest rate (New Jersey and Illinois). There are cases, however, in which the impact of restrictions on creditor remedies is mixed. Specifically, restrictions on garnishment, wage assignment, and late charges do not have any significant impact on loan interest rates in Louisiana (a high ceiling state). It seems plausible that such restrictions may have different impacts in the two high ceiling states because of factors other than interest rate ceilings affecting borrowing and lending activity. First, there may be differences in the social or economic conditions of borrowers which influence the overall risk of default. Second, regulations not included in the model may also interact with interest rate ceilings and creditor remedies and thus have important effects on the equilibrium price and quantity of credit. To the extent such factors are excluded from this as well as previous studies, a more complete analysis of the interaction between all regulations governing personal loan markets remains to be performed.

Regulations governing the use of a variety of creditor remedies and the amount of future income of the borrower that may be garnished or assigned to the tender have different impacts on credit activity in states with different interest rate ceilings. The empirical results for the five states, which represent alternative regulatory regimes, reveal that there do exist dramatic differences in the role of security and collection remedies in personal loan contracts across states. Further work on identifying and quantifying regulations other than creditor remedies and interest rate ceilings should be a valuable aid in assessing the joint impact of regulations in each state.

REFERENCES

Aho, Michael, James R. Barth, Joseph J. Cordes, and Anthony M. Yezer. "Impact of Regulation on the Cost and Availability of Credit." Journal of Consumer Credit Management, Winter 1979, 66-79.

Barro, Robert J. "The Loan Market, Collateral, and Rates of interest." Journal of Money, Credit, and Banking, November 1976, 440-56.

Barth, James R. and Anthony Yezer. "The Economic Impact of the Federal Trade Commission Proposals for Credit Contract Regulations on the Cost and Availability of Consumer Credit." Bureau of Social Science Research, 1977.

Barth, James R., Padma Gotur, Neela Manage, and Anthony M. J. Yezer. The Effect of Government Regulations on Personal Loan Markets: A Tobit Estimation of a Microeconomic Model." Journal of Finance, September 1983, 1233-5 1.

Benjamin, Daniel K. "The Use of Collateral to Enforce Loan Contracts." Economic Inquiry, July 1978, 333-59.

Benston, George J. "Risk on Consumer Finance Company Personal Loans." Journal of Finance, May 1977, 593-607.

The Cost to Consumer Finance Companies of Extending Consumer Credit." National Commission on Consumer Finance, Technical Studies Volume 2, 1973, 1-158.

Board of Governors of the Federal Reserve System. Federal Reserve Bulletin, various issues, 1975-1977.

Commerce Clearing House, Inc. Consumer Credit Guide, Chicago 1976.

Eaton, Jonathan. "Lending with Costly Enforcement of Repayment and Potential Fraud." Journal of Banking and Finance, June 1986, 281-93.

Financial Publishing Company. Cost of Personal Borrowing in the United States, Boston, 1977 and 1978.

Goudzwaard, Maurice B. "Price Ceilings and Credit Rationing." Journal of Finance, March 1968, 177-85.

Greer, Douglas F. "An Econometric Analysis of the Personal Loan Market." National Commission on Consumer Finance Technical Studies Volume IV, 1973, 37-126.

Rate Ceilings and Loan Turndowns." Journal of Finance, December 1975, 1376-83.

Harris, Richard. "The Consequences of Costly Default." Economic Inquiry, October 1978, 447-96.

Hess, James D. and Charles R. Knoeber. "Security and Penalty in Debt Contracts." Journal of institutional and Theoretical Economics 143(1), 1987, 149-67.

Hess, James. The Use of Collateral to Enforce Debt: Profit Maximization." Economic Inquiry, April 1985, 349-56.

Manage, Neela D. "Further Evidence on Estimating Personal Loan Market Relationships." Quarterly Review of Economics and Business, Winter 1983, 66-80.

Peterson, Richard L. "Effects of Creditor Remedies and Rate Restrictions," in The Regulation of Financial Institutions, Federal Reserve Bank of Boston Conference Series, Boston, 1979, 24-43.

Rea, Samuel A. "Arm Breaking, Consumer Credit and Personal Bankruptcy." Economic Inquiry, April 1984, 188-208.

Schwartz, Alan. "Security Interests and Bankruptcy Priorities: A Review of Current Theories." Journal of Legal Studies, January 1981, 1-37.

"The Continuing Puzzle of Secured Debt." Vanderbilt Law Review, October 1984, 1051-69.

Shay, Robert P. "Factors Affecting the Price, Volume, and Credit Risk in the Consumer Finance Industry." Journal of Finance, May 1970, 503-15.

Villegas, Daniel J. "An Analysis of the Impact of Interest Rate Ceilings." Journal of Finance, September 1982, 941-54.