Personal Bankruptcy: The Credit Research Center Report on Debtors'
Ability To Pay (Letter Report, 02/09/98, GAO/GGD-98-47).

Pursuant to a congressional request, GAO reviewed the Credit Research
Center report on personal bankruptcies, focusing on the report's
research methodology and formula for estimating the income that debtors
have available to pay debts.

GAO noted that: (1) overall, the Center report represents a useful first
step in analyzing the ability of bankruptcy debtors to pay their debts;
(2) because there is little empirical basis on which to assess the
accuracy of the data used in the report's analysis, and because the data
provided by the authors showed considerable variation among the 13
locations used for analysis, the report's general findings must be
interpreted with caution; (3) GAO's review of the Center report suggests
that additional research and clarification would be needed to confirm
the accuracy of the report's conclusions regarding the proportion of
debtors who may have the ability to repay at least a portion of their
nonpriority, nonhousing debts; and (4) there were five areas of concern
with the Center's report that could affect interpretation of the
report's conclusions: (a) the report's assumption's about the
information debtors provide at the time of filing bankruptcy regarding
their income, expenses, and debts and the stability of their income and
expenses over a 5-year period were not validated; (b) the report did not
clearly define the universe of debts for which it estimated debtors'
ability to pay; (c) payments on nonhousing debts that debtors stated
they intended to reaffirm--voluntarily agree to repay--were not included
in debtor expenses in determining the net income debtors had available
to pay their nonpriority, nonhousing debts; (d) the report presented
results based on data from all 13 locations combined and provided little
discussion of the considerable variation among the 13 locations used in
the analysis; and (e) a scientific, random sampling methodology was not
used to select the 13 bankruptcy locations or the bankruptcy petitions
used in the analysis.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-98-47
     TITLE:  Personal Bankruptcy: The Credit Research Center Report on 
             Debtors' Ability To Pay
      DATE:  02/09/98
   SUBJECT:  Bankruptcy
             Personal liability (legal)
             Debt collection
             Income statistics
             Financial management
             Financial analysis
             Debt
             Statistical methods

             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

February 1998

PERSONAL BANKRUPTCY - THE CREDIT
RESEARCH CENTER REPORT ON DEBTORS'
ABILITY TO PAY

GAO/GGD-98-47

Personal Bankruptcy

(188635)


Abbreviations
=============================================================== ABBREV

  AOUSC - Administrative Office of the U.S.  Courts
  FJC - Federal Judicial Center

Letter
=============================================================== LETTER


B-278972

February 9, 1998

The Honorable Charles E.  Grassley, Chairman
The Honorable Richard J.  Durbin
Ranking Minority Member
Subcommittee on Administrative Oversight
  and the Courts
Committee on the Judiciary
United States Senate

This report provides information on the results of our review of the
Credit Research Center (the Center) report on personal bankruptcies. 
The Center report addressed a major public policy issue--the amount
of income that those who file for personal bankruptcy have available
to pay their debts.  Specifically, you requested that we evaluate the
report's research methodology and formula for estimating the income
that debtors have available to pay debts. 


   BACKGROUND
------------------------------------------------------------ Letter :1

Debtors who file personal bankruptcy petitions usually file under
chapters 7 or 13 of the bankruptcy code.  Generally, debtors who file
under chapter 7 of the bankruptcy code seek a discharge of all their
eligible dischargeable debts.\1

Debtors who file under chapter 13 submit a repayment plan, which must
be confirmed by the bankruptcy court, for paying all or a portion of
their debts over a 3-year period unless for cause the court approves
a period not to exceed 5 years.  The Center report was based on data
from 3,798\2 personal bankruptcy petitions filed principally in May
and June 1996 in 13 of the more than 180 bankruptcy court locations. 
The petitions included 2,441 chapter 7 and 1,357 chapter 13
petitions. 

The researchers collected a wide variety of information about
debtors' income, expenditures, and debts from the schedules the
debtors filed with their bankruptcy petitions.  Because the debtors'
schedules used in the report must be obtained from the case files at
each court location, obtaining the data used for the Center report
represented a considerable investment of Center time and money.  The
data are not available from the automated databases maintained by the
federal judiciary or the Executive Office of U.S.  Trustees, the two
principal sources of automated data on bankruptcy cases. 

On the basis of the Center report's assumptions and the formula used
to determine income available for repayment of nonpriority,
nonhousing debt,\3 the report estimated that about 50 percent of the
chapter 13 debtors in the 13 locations combined would have sufficient
income, after living expenses, to repay all of their nonpriority,
nonhousing debt over a 5-year period; and an additional 19 percent
could pay 60 percent or more over the same period.  The report
estimated that 5 percent of the chapter 7 debtors in the 13 locations
combined could repay all of their nonpriority, nonhousing debt over 5
years; 10 percent could repay at least 78 percent, and 25 percent
could repay at least 30 percent.\4

The Center report also estimated that about 11 percent of chapter 13
debtors and about 56 percent of chapter 7 debtors were expected to
have no income available to repay nonhousing debts. 


--------------------
\1 Eligible debts may be discharged in bankruptcy proceedings.  A
dischargeable debt is a debt for which the bankruptcy code allows the
debtor's personal liability to be eliminated. 

\2 This is the number of petitions the Center's report said were
usable for analysis. 

\3 The Center report assumed that debtors would repay their home
mortgage debt and "priority debts." As discussed later, the report
defined priority debt as "unsecured priority debt." The report did
not clearly define unsecured priority debts nor explain why the
report's analysis assumed that debtors would repay their unsecured
priority debts. 

\4 The 10 percent and 30 percent figures were cumulative.  That is,
the 10 percent figure included all debtors the report estimated would
be able to pay 78 percent or more, and the 25 percent figure included
all debtors the report estimated would be able to pay 30 percent or
more of their nonhousing debt. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

Overall, the Center report represents a useful first step in
analyzing the ability of bankruptcy debtors to pay their debts. 
Because there is little empirical basis on which to assess the
accuracy of the data used in the report's analysis, and because the
data provided by the authors showed considerable variation among the
13 locations used for analysis, the report's general findings must be
interpreted with caution.  Our review of the Center report suggests
that additional research and clarification would be needed to confirm
the accuracy of the report's conclusions regarding the proportion of
debtors who may have the ability to repay at least a portion of their
nonpriority, nonhousing debts. 

We found five areas of concern with the Center report that could
affect interpretation of the report's conclusions:  (1) the report's
assumptions about the information debtors provide at the time of
filing bankruptcy regarding their income, expenses, and debts and the
stability of their income and expenses over a 5-year period were not
validated; (2) the report did not clearly define the universe of
debts for which it estimated debtors' ability to pay; (3) payments on
nonhousing debts that debtors stated they intended to
reaffirm--voluntarily agree to repay--were not included in debtor
expenses in determining the net income debtors had available to pay
their nonpriority, nonhousing debts; (4) the report presented results
based on data from all 13 locations combined and provided little
discussion of the considerable variation among the 13 locations used
in the analysis; and (5) a scientific, random sampling methodology
was not used to select the 13 bankruptcy locations or the bankruptcy
petitions used in the analysis. 

First, the report's analysis and conclusions about the income debtors
had available for repayment of nonpriority, nonhousing debt rest on
two fundamental assumptions:  (1) the debtors' schedules of current
estimated income, current estimated monthly expenditures, and debts,
generally filed at the same time as the bankruptcy petitions,\5 were
accurate; (2) the debtors' current estimated income and living
expenses, as reported in those schedules, could be used to
satisfactorily forecast their income and living expenses for a 5 year
debt repayment period.  These assumptions have been the subject of
considerable debate, and the researchers did not test their validity. 

With regard to the accuracy of the data in the debtors' initial
schedules, the National Bankruptcy Review Commission's October 1997
report\6 noted that there has been no empirical study of the accuracy
of the financial data initially reported by bankruptcy debtors, and
it recommended random audits of such data.  The Center report
acknowledged that there were inherent uncertainties in using the data
from debtors' schedules of income and expenses, but it stated that
the assumptions used in the analysis regarding debtors' net income
available to repay nonpriority, nonhousing debts were conservative
and may actually understate net income. 

With regard to the 5-year stability in debtors' income and expenses,
the data reported in the debtors' initial schedules represent a
snapshot in time, and there is some empirical evidence that these
data may not necessarily provide a reliable foundation for
forecasting debtors' income and expenses for a 5-year period.  A 1994
study by the Administrative Office of the U.S.  Courts (AOUSC)\7
found that only about 36 percent of debtors who voluntarily entered a
3 to 5 year bankruptcy debt repayment plan under chapter 13 were able
to successfully complete their repayment plans and obtained
discharges.\8 Another 14 percent of these debtors were unable to
complete their chapter 13 plans and had their eligible debts
discharged after their cases were converted to chapter 7.  About 49
percent had their cases dismissed and did not receive a discharge of
their eligible dischargeable debts.  This suggests that the Center
report's second assumption--that debtors' income and expenses would
remain unchanged for a 5-year period--may be optimistic (at least for
a portion of debtors) and that further research may be necessary to
validate this assumption. 

Second, the report did not clearly define the universe of nonhousing
debts for which it estimated debtors' ability to pay.  There is some
evidence in the Center report that the intent of the analysis was to
estimate debtors' ability to pay their eligible dischargeable
nonhousing debts--secured and unsecured.  However, this is not
explicitly stated.  The Center report defined the net income that
debtors had available to pay nonhousing debts as the debtor's net
annual take-home pay less (1) annual living expenses (as defined in
the report) and (2) annual payments toward "unsecured priority
debt."\9 The report stated that its analysis assumed that all
unsecured priority debt would be paid in full over a 5 year repayment
period.  The report also noted that the two most common unsecured
priority debts that debtors listed were back taxes and past-due child
support.  These are debts that are generally nondischargeable in
bankruptcy proceedings.\10 However, the categories of debts listed on
the schedule of unsecured priority debts filed by the debtor\11

can, in some cases, include both debts that are dischargeable and
debts that are generally nondischargeable.\12 The Center report did
not state whether the report's assumption about the repayment of
unsecured priority debts included dischargeable as well as
nondischargeable debts.  Moreover, not all debts that are generally
nondischargeable would be appropriately listed in the debtor's
schedule of unsecured priority debts.  To the extent that the Center
report understated nondischargeable debts, it would have overstated
the net income that debtors would have available to pay dischargeable
debts.  Conversely, to the extent that the report overstated
nondischargeable debts, it would have understated debtors' net income
available to pay dischargeable debts.  In addition, to the extent
that the report assumed that all dischargeable unsecured priority
debts would be paid, it would have created a disparity in the
report's treatment of nonhousing dischargeable debts. 

Third, the Center report did not include in debtors' expenses the
payments required to repay the nonhousing debts that debtors stated
it was their intention to reaffirm.\13 In each of the report's 13
locations, debtors stated their intention to reaffirm, or repay, at
least some of their nonhousing debts.  For 12 of the 13 locations in
the Center report (Dallas reaffirmation data were incomplete), the
average amount of secured nonhousing debt that debtors indicated they
intended to reaffirm ranged from $1,362 per debtor in Los Angeles to
$6,706 per debtor in Memphis.\14 To the extent that debtors in the
Center report reaffirmed nonhousing debts and maintained payments on
those debts, these debtors would have less estimated income to pay
their eligible dischargeable debts that were not reaffirmed. 

Fourth, the Center report presented results based on data from all 13
locations combined.  However, the data provided to us by the report's
authors, but not included in the report, showed wide variation among
the report's 13 locations in debtors' estimated ability to pay.  In
the 12 locations for which the Center said its reaffirmation data
were complete, the data also showed wide variation in the percentage
of debtors who stated their intention to reaffirm at least some
nonhousing debts and the average amount of nonhousing debt to be
reaffirmed.  These variations may in part reflect the influence of
varying local bankruptcy practices.  Given these variations, we
believe it is appropriate to be cautious in making general statements
about the debtors across all 13 locations. 

Finally, the Center's researchers selected the 13 bankruptcy
locations and 3,798 personal bankruptcy petitions without using
scientific random sampling techniques.  As a result, the national
estimates presented in the report are not based on representative
probability sampling methods.  In addition, standard statistical
methods, such as the calculation of statistical error rates, cannot
be used to evaluate the likely accuracy of the results in the Center
report.  Consequently, the methods used in the Center's analysis do
not provide a sound basis for generalizing the Center report's
findings to the annual 1996 filings in each of the 13 locations nor
to the national population of personal bankruptcy filings. 


--------------------
\5 Federal bankruptcy rule 1007 provides that, among other things,
schedules and statements other than the statement of intention shall
be filed with the bankruptcy petition in a voluntary case, or if the
petition is accompanied by a list containing the names and addresses
of all the debtor's creditors, within 15 days thereafter. 

\6 Bankruptcy:  The Next Twenty Years, October 20, 1997. 

\7 Bankruptcy Statistical Trends:  Chapter 13 Dispositions, October
1994. 

\8 This total included "hardship discharges." A hardship discharge
generally may be granted to a chapter 13 debtor who fails to complete
the plan payments due to circumstances for which the debtor should
not justly be held accountable.  An AOUSC official and the Executive
Office of U.S.  Trustees said such chapter 13 discharges were rare. 

\9 The distinction between priority and nonpriority unsecured debt in
chapter 7 cases is most evident in those chapter 7 bankruptcy cases
in which money from the sale of a debtor's nonexempt assets is
available for distribution to creditors.  Such chapter 7 "asset"
cases are a very small proportion of total chapter 7 personal
bankruptcy cases.  In such asset cases, the proceeds from the sale of
the nonexempt assets are distributed to creditors in a statutorily
defined order--unsecured priority and unsecured nonpriority.  In
no-asset chapter 7 cases, the debtor's eligible dischargeable
unsecured debts--priority and nonpriority--that have not been
reaffirmed are usually discharged (with limited exceptions). 

\10 By statute, some types of debts and obligations, such as alimony,
child support, some student loans, and certain taxes, cannot
generally be discharged in bankruptcy proceedings.  The debtor
remains financially responsible for nondischargeable debts after the
close of his or her bankruptcy case. 

\11 Schedule E "Creditors Holding Unsecured Priority Claims."

\12 As a practical matter, with the exception of alimony, child
support, and taxes--debts that are generally nondischargeable in
chapter 7 cases--the priority of unsecured claims set forth in
section 507(a) of the bankruptcy code (and listed on Schedule E) does
not play a very meaningful role in no-asset chapter 7 cases.  This is
because the eligible dischargeable debts on Schedule E, if
applicable, are usually discharged in no-asset chapter 7 cases. 

\13 A reaffirmation of a debt is a formal agreement whereby the
debtor agrees to remain personally liable for a debt, despite the
bankruptcy discharge being granted by the bankruptcy court. 

\14 In no location did debtors state their intention to reaffirm more
than 1 percent of their unsecured nonhousing debt. 


   STUDY'S FUNDAMENTAL ASSUMPTIONS
   WERE NOT VALIDATED
------------------------------------------------------------ Letter :3

The Center report's analysis was based on data from the initial
schedules of current estimated monthly income, current estimated
average monthly expenditures, and debts that debtors submitted at the
time they filed for bankruptcy.  There are two reasons to question
whether broad conclusions about debtors' ability to pay nonhousing
debt can be made on the basis of the debtors' statements of estimated
income and estimated expenses at the time of filing for bankruptcy: 

  -- The accuracy of the data in the debtors' initial schedules is
     unknown, and no empirical study has been done to assess their
     accuracy.  Moreover, debtors may generally amend these schedules
     as a matter of course at any time prior to final disposition of
     the debtors' bankruptcy cases.

  -- The Center report assumed that debtors' income and living
     expenses, as reported in those schedules, could be used to
     satisfactorily forecast debtors' income and living expenses for
     a 5 year debt repayment period.  However, the report did not
     include empirical evidence to support this assumption.  There is
     some empirical evidence that this assumption may not be
     appropriate, at least for a portion of debtors who file for
     bankruptcy. 


      ACCURACY OF DATA ON DEBTORS'
      SCHEDULES OF INCOME,
      EXPENSES, AND DEBTS NOT
      KNOWN
---------------------------------------------------------- Letter :3.1

The Center report relied on debtors' self-reported data on current
estimated income, current estimated expenditures, and debts at the
time of filing and assumed that these data were accurate.  Although
the data in the various schedules are the only such information
available at the time a debtor files for bankruptcy, the National
Bankruptcy Review Commission report noted that "no study has yet been
done to test the accuracy of the data as initially reported by
debtors," and it recommended random audits of debtors' initial
schedules.  The effect of any inaccuracies in these schedules could
be that the debtor's actual net income is overstated or understated. 

The schedules that debtors complete on their current average monthly
income\15 and current average monthly expenditures\16 indicate that
debtors should estimate their income and expenditures.  The data that
debtors report in these schedules represent a snapshot in time, and
debtors may generally amend their schedules at any time prior to
final disposition of their bankruptcy cases.  Such amendments were
not included in the Center's analysis.  Amendments may be made for a
variety of reasons, but there are no readily available empirical data
on how frequently schedules are actually amended and the effect of
such amendments on the income, expenditures, and debts that debtors
report on their initial schedules. 


--------------------
\15 Schedule I "Current Income of Individual Debtor(s)." The schedule
includes such categories as monthly gross wages, salary, and
commissions; payroll deductions; and income from nonwage sources,
such as interest and dividends, alimony, and Social Security. 

\16 Schedule J "Current Expenditures of Individual Debtor(s)." The
schedule includes such expenditure categories as housing, utilities,
food, laundry, medical and dental, insurance, and transportation. 


      ASSUMPTION THAT DEBTORS'
      REPORTED INCOME AND EXPENSES
      COULD BE USED TO
      SATISFACTORILY FORECAST
      INCOME AND EXPENSES FOR 5
      YEARS
---------------------------------------------------------- Letter :3.2

The Center report's analysis assumed that the debtor's income and
expenses, as reported on the schedules filed with the bankruptcy
petition, could be used to satisfactorily forecast his or her income
and expenses during the course of a 5 year debt repayment period.  In
other words, the Center report assumed that a debtor's reported
income and expenses would remain uninterrupted and unchanged over the
5 years.  This assumption is critical to the report's estimate of the
percentage of nonhousing debt that debtors could repay over 5 years. 
However, the Center report provided no empirical support for this
assumption. 

A couple of factors raise questions about the validity of this
assumption.  First, the Center report provided evidence of
instability in debtor income in the year preceding the debtor's
bankruptcy filing.  About 77 percent of the 2,441 chapter 7 debtors
and about 85 percent of the 1,357 chapter 13 debtors in the Center's
analysis reported having some wage income at the time of filing. 
However, about 68 percent of the chapter 7 debtors and about 50
percent of the chapter 13 debtors in the Center's report also
reported they had experienced a reduction in income during the 12
months prior to filing bankruptcy.  As the Center report noted, it is
not surprising that those who file for bankruptcy have suffered a
loss of income prior to filing. 

Second, there is also some evidence that debtors may experience
fluctuating income or expenses in the 5 years following the filing of
their bankruptcy petitions.  The findings of a 1994 report by AOUSC
suggest that at least a portion of debtors could be expected to
experience deterioration in their financial circumstances during the
5 years after filing for bankruptcy.  AOUSC reviewed the outcomes of
953,180 chapter 13 cases filed between calendar years 1980 and 1988
and terminated by September 30, 1993.\17 AOUSC found that debtors
received a discharge in only about 36 percent of all chapter 13 cases
terminated.\18

A chapter 13 discharge is generally granted when a debtor
successfully completes a court-approved repayment plan.  A hardship
discharge may be granted to chapter 13 debtors who fail to complete
the plan payments due to circumstances for which the debtor should
not justly be held accountable.\19 AOUSC found that in about 14
percent of all chapter 13 cases terminated, the debtors were unable
to maintain their payments; prior to termination, their cases were
converted to chapter 7 liquidation, in which all eligible debts were
discharged.\20 The typical case that converted to chapter 7 did so
about 2 years after the case was filed.  AOUSC also found that about
49 percent of all chapter 13 cases terminated were dismissed, but
data were not available on the reasons for the dismissals.\21 The
results of the AOUSC report caution against making broad conclusions
about debtors' ability to maintain debt payments over a 5-year period
based on the data in the initial schedules alone. 


--------------------
\17 As of September 30, 1993, almost 97 percent of the 985,212
chapter 13 cases filed between calendar years 1980 and 1988 had been
terminated.  According to data in the AOUSC report, each year during
this period nonbusiness debtors accounted for at least 91 percent of
all chapter 13 petitions filed.  However, AOUSC's report on case
results did not distinguish between business and nonbusiness chapter
13 debtors, and it is possible that the results for nonbusiness
debtors could have differed from that of business debtors. 

\18 The percentage of cases for which debtors received a discharge
remained relatively stable at between 34 percent and 37 percent of
all cases terminated from 1982 to 1987.  Although the percentage of
terminated cases that resulted in discharges dropped to 31 percent in
1988, AOUSC noted that the percentage would have been expected to
increase as more 1988 cases were closed. 

\19 AOUSC did not report data on the number of chapter 13 hardship
discharges. 

\20 The percent of chapter 13 cases that converted to chapter 7 was
relatively stable at between 13 and 16 percent for cases filed each
year between 1980 and 1988. 

\21 Chapter 13 cases may be dismissed for a variety of reasons,
including failure to pay the case filing fee, the debtor's failure to
attend the required meeting with the chapter 13 trustee and the
debtor's creditors, failure to file a plan in a timely manner, denial
of confirmation of a repayment plan, failure to begin making plan
payments, or material default by the debtor with respect to a term of
a confirmed plan. 


   CENTER REPORT DID NOT CLEARLY
   DEFINE THE UNIVERSE OF
   NONHOUSING DEBTS FOR WHICH IT
   ESTIMATED DEBTORS' ABILITY TO
   PAY
------------------------------------------------------------ Letter :4

There is some evidence in the Center report that the intent of the
analysis was to estimate debtors' ability to pay their eligible
dischargeable nonhousing debts--secured and unsecured.  However, this
is not explicitly stated, and the Center report did not clearly
define the universe of nonhousing debts for which it estimated
debtors' ability to pay.  The Center report defined the net income
that debtors had available to pay nonhousing debts as the debtor's
net annual take-home pay less (1) living expenses (as defined in the
report) and (2) payments toward "unsecured priority debt."\22 As
examples of such debts, the report mentioned back taxes and past-due
child support.  These are examples of debts that are generally
nondischargeable in bankruptcy proceedings. 

Debtors report unsecured priority debts on Schedule E.\23 However,
the debts to be listed in Schedule E can, in some cases, include both
debts that are dischargeable and debts that are generally
nondischargeable.  Moreover, not all nondischargeable debts can be
found in Schedule E.\24 For example, certain student loans, debts
arising out of drunk driving, criminal restitution, and criminal
court fines are not dischargeable in bankruptcy, but such obligations
would be appropriately listed as "unsecured nonpriority debt" on
Schedule F.\25 The Center report included student loans in unsecured
nonpriority debt.  However, it is not clear if these student loans
represented only loans that were eligible for discharge.  Thus, the
Center report may not have identified all generally nondischargeable
debts for which the debtor would still be responsible following the
close of his or her bankruptcy case.\26

To the extent that the Center report understated nondischargeable
debts, it would have overstated the net income that debtors would
have available to pay dischargeable nonhousing debts.  Conversely, to
the extent that the report overstated nondischargeable debts, it
would have understated debtors' net income available to pay
dischargeable debts.  In addition, to the extent that the report
assumed that dischargeable unsecured priority debts would be paid, it
would have created a disparity in the report's treatment of
dischargeable nonhousing debts. 


--------------------
\22 The distinction between priority and nonpriority unsecured debt
in chapter 7 cases is most evident in those chapter 7 bankruptcy
cases in which money from the sale of a debtor's nonexempt assets is
available for distribution to creditors.  Such chapter 7 "asset"
cases are a very small proportion of total chapter 7 personal
bankruptcy cases.  In such asset cases, the proceeds from the sale of
the nonexempt assets are distributed to creditors in a statutorily
defined order--unsecured priority and unsecured nonpriority.  In
no-asset chapter 7 cases, the debtor's eligible dischargeable
unsecured debts--priority or nonpriority--that have not been
reaffirmed are usually discharged (with limited exceptions). 

\23 Schedule E "Creditors Holding Unsecured Priority Claims."

\24 The Center report apparently did include in debtors' living
expenses any reported alimony payments debtors listed in Schedule J. 
Alimony payments are also generally nondischargeable debts.  It
should be noted that under the bankruptcy code, nondischargeable
child support, alimony, and taxes must be paid in full in chapter 13
repayment plans.  These obligations would appropriately be listed on
Schedule E.  The Center report assumed that unsecured priority debts
would be paid in full over a 5-year period in either chapter 7 or
chapter 13 bankruptcies. 

\25 Schedule F "Creditors Holding Unsecured Nonpriority Claims."

\26 It is possible that the Center report understated total
nonhousing debts that are dischargeable.  Based on our review of data
provided by the authors of the Center report, it appeared that the
Center report did not include data from Schedule G "Executory
Contracts and Unexpired Leases" or Schedule H "Codebtors." For the
purposes of identifying debtors' total nonhousing debts that were
eligible for discharge, perhaps the most relevant data from these
schedules would be information on debtors' obligations for unexpired
leases on real or personal property, such as automobile leases. 


   CENTER REPORT'S DETERMINATION
   OF DEBTORS' NET INCOME TO PAY
   NONHOUSING DEBT EXCLUDED
   PAYMENTS ON NONHOUSING DEBTS
   THAT WERE TO BE REAFFIRMED
------------------------------------------------------------ Letter :5

A portion of personal bankruptcy debtors voluntarily agree to
reaffirm, or repay, some of their dischargeable debts by entering
into a reaffirmation agreement to remain personally liable for
reaffirmed debts.\27 According to the Executive Office of the U.S. 
Trustees, debtors tend to reaffirm secured debt, such as a home
mortgage or car loan.  By reaffirming these debts and keeping current
on the payments, the debtors retain possession of the property
secured by the debt.  To the extent that debtors maintain their
payments on reaffirmed debt, it would reduce the amount of income
debtors have to pay eligible dischargeable debts that were not
reaffirmed. 

The Center report included in debtors' living expenses the full value
of any home mortgage payments the debtors listed in Schedule J.  To
the extent that the listed home mortgage payments actually represent
the full payments required for home mortgage debt, the Center report
assumed that debtors had reaffirmed their housing debt.  However, the
Center report did not deduct from debtors' income the value of the
payments required to pay the nonhousing debts that debtors stated it
was their intention to reaffirm.\28 Data provided by the authors of
the Center report showed that for 12 of the 13 locations in the
report (Dallas reaffirmation data were incomplete), secured
nonhousing debt accounted for virtually all the average nonhousing
debt that debtors intended to reaffirm.  The average percent of total
unsecured debt that debtors indicated they intended to reaffirm did
not exceed 1 percent in any of the 12 locations.\29

The effect of deducting from chapter 7 debtors' income the payments
required to repay reaffirmed secured nonhousing debts would be
expected to vary across the 13 locations in the Center report because
of the wide variation in intended reaffirmations by location.  Using
data provided by the authors of the Center report, we show in table 1
the percentage of chapter 7 debtors in each of the report's 13
locations who stated their intent to reaffirm at least some of their
secured nonhousing debts, the average percent of total secured
nonhousing debt to be reaffirmed, and the average total dollar amount
of secured nonhousing debts to be reaffirmed.  The percentage of
chapter 7 debtors who, according to the Center's data, stated their
intent to reaffirm at least some of their secured nonhousing debt
ranged from about 23 percent in Los Angeles to about 73 percent in
Indianapolis.  The data also showed considerable differences for
those locations within the same state.  About 23 percent of chapter 7
debtors in Los Angeles reported their intent to reaffirm at least
some secured nonhousing debt compared to about 42 percent in San
Diego.  The average percentage of secured nonhousing debt that
chapter 7 debtors stated they intended to reaffirm ranged from about
23 percent in Los Angeles to about 61 percent in Memphis.  The
average amount of total debt to be reaffirmed ranged from about
$1,362 per debtor in Los Angeles to $6,706 per debtor in Memphis. 
The averages for any specific location may be based on wide variation
in the amount of debt that individual debtors stated it was their
intent to reaffirm. 



                                     Table 1
                     
                        Credit Research Center Data on the
                     Average Percent of Chapter 7 Debtors Who
                      Stated an Intent to Reaffirm at Least
                        Some Secured Nonhousing Debt, the
                         Average Percent of Total Secured
                      Nonhousing Debt to Be Reaffirmed, and
                       the Average Dollar Amount of Secured
                       Nonhousing Debt to Be Reaffirmed, by
                                  Court Location

          Chapter 7 debtors' voluntary reaffirmation of secured nonhousing debts
          ----------------------------------------------------------------------
          Percent of debtors who
             stated an intent to      Average percent of
          reaffirm at least some           total secured   Average dollar amount
Court         secured nonhousing   nonhousing debt to be   of secured nonhousing
location                    debt              reaffirmed   debt to be reaffirmed
--------  ----------------------  ----------------------  ----------------------
Atlanta,                   70.2%                   58.8%                  $5,936
 GA
Chicago,                    62.1                    54.4                   4,263
 IL
Houston,                    57.4                    46.7                   5,353
 TX
Hartford                    37.8                    36.1                   1,799
 , CT
Indianap                    72.6                    59.2                   4,655
 olis,
 IN
Kansas                      46.5                    37.8                   3,613
 City,
 MO
Los                         22.6                    22.6                   1,362
 Angeles,
 CA
San                         42.3                    38.9                   1,535
 Diego,
 CA
Memphis,                    65.2                    61.2                   6,706
 TN
Phoenix,                    66.4                    49.1                   4,680
 AZ
Pittsbur                    52.3                    42.4                   5,897
 gh, PA
Tampa,                      54.3                    46.0                   5,217
 FL
--------------------------------------------------------------------------------
\a According to the Credit Research Center, the Dallas data provided
to us on debtors' statements of intent were incomplete because the
Center did not have all the statements of intent at that time. 
Therefore, Dallas data were excluded from this table. 

Source:  GAO analysis of Credit Research Center data. 


--------------------
\27 The reaffirmation agreement must meet certain statutory
requirements. 

\28 The Center's data on intended reaffirmations are based on
information from the form, "Individual Debtor's Statement of
Intention." The debts that are ultimately reaffirmed may differ from
those included in this statement of intention. 

\29 Based on data provided by the authors of the Center report, among
the 12 locations for which complete reaffirmation data were
available, the average dollar value of the unsecured nonhousing debt
that debtors stated they intended to reaffirm did not exceed about
$300.  Given that the statement of intent form principally focuses on
property that the debtor intends to surrender or retain, it is not
surprising that debtors listed little unsecured debt on their
statement of intention. 


   LOCATIONS IN THE REPORT VARIED
   IN INDICATORS OF ABILITY TO PAY
------------------------------------------------------------ Letter :6

The Center report presented data that combined results from all 13
locations on debtors' available income to pay nonhousing debt. 
Because the Center report focused on the results from all 13
locations combined, it included little discussion of the considerable
variations among the 13 locations used in the study.  As previously
discussed, Center data not included in the report showed a wide
variation across the 12 locations with complete data for chapter 7
debtors' intended reaffirmations of secured nonhousing debt. 
Specifically, as shown in table 1, the percentage of chapter 7
debtors reaffirming at least some secured nonhousing debt ranged from
about 23 percent to 73 percent, and the average amount of total debt
to be reaffirmed ranged from about $1,362 to $6,706.  Data provided
by the report's authors showed that the percentage of chapter 7
debtors with at least some income available to pay nonpriority,
nonhousing debt ranged from about 32 percent in San Diego to about 67
percent in Dallas. 

Other studies have also concluded that there is considerable
variation among bankruptcy districts.  The National Bankruptcy Review
Commission found, for example, that chapter 13 practices "differ
dramatically from state to state, district to district, and even from
judge to judge in the same district." The Commission report noted
that divergent local interpretations of the chapter 13 system create
a situation in which expert legal advice is necessary to develop,
confirm, modify, and complete a chapter 13 plan; and debtors in very
similar circumstances encounter extremely different chapter 13
systems across the nation. 

The AOUSC report on chapter 13 cases discussed earlier found
considerable variation in case results among all bankruptcy districts
and among the 13 districts included in the Center report.\30 As shown
in table 2, the percentage of terminated chapter 13 cases that
resulted in the discharge of a successful repayment plan ranged from
about 15 percent in Central California (which includes Los Angeles)
to about 40 percent in Western Missouri (which includes Kansas City). 
The percentage of chapter 13 cases that were converted to chapter 7
liquidation prior to termination ranged from about 8 percent in
Western Tennessee (which includes Memphis) to about 43 percent in
Western Pennsylvania (which includes Pittsburgh). 



                                Table 2
                
                 Type of Disposition for All Chapter 13
                Cases Filed Between Calendar Years 1980
                and 1988 and Terminated by September 30,
                 1993, in the Districts From Which the
                 Petitions Used in the Credit Research
                      Center Report Were Selected

                                    Percent of terminated cases that
                                                 were:
                                  ------------------------------------
                                                      Converte
                           Total                          d to
District (Center        terminat  Dismisse  Discharg   chapter
location)                     ed         d        ed         7   Other
----------------------  --------  --------  --------  --------  ------
GA, Northern              41,897      55.3      34.8       9.8     0.1
 (Atlanta)
IL, Northern (Chicago)    59,088      47.6      38.6      13.6     0.2
TX, Northern              15,657      60.2      21.4      17.4     1.0
 (Dallas)
TX, Southern (Houston)    22,378      60.3      19.6      19.7     0.4
CT (Hartford)              2,106      42.2      31.3      25.0     1.5
IN, Southern               3,309      32.1      33.2      29.7     5.1
 (Indianapolis)
MO, Western                5,017      39.3      39.6      20.2     0.9
 (Kansas City)
CA, Central               72,769      77.0      14.5       8.2     0.3
 (Los Angeles)
CA, Southern              19,953      59.7      32.4       7.8     0.1
 (San Diego)
TN, Western (Memphis)     46,348      61.5      30.7       7.5     0.3
AZ (Phoenix)               6,393      40.0      32.8      24.7     2.6
PA, Western                2,883      31.9      22.9      42.9     2.3
 (Pittsburgh)
FL, Middle (Tampa)         4,728      45.7      24.9      25.2     4.3
----------------------------------------------------------------------
Note:  Cities in parentheses are court locations within each district
from which petitions were drawn for the Credit Research Center
report. 

Source:  1994 AOUSC study of chapter 13 dispositions. 

These variations among bankruptcy districts--for percentage of
debtors with at least some income to pay debts, for reaffirmations,
and for the final disposition of chapter 13 cases--suggest that one
should be cautious in generalizing about debtors across all 13
locations in the Center's report. 


--------------------
\30 AOUSC's data do not correspond directly to the locations included
in the Center report.  AOUSC's analysis included all cases filed or
terminated in each district.  Due to resource constraints, the Center
selected petitions from 1 location in each of the 13 districts
included in its analysis.  For example, the Central District of
California includes bankruptcy court locations in Los Angeles, San
Bernardino, Santa Barbara, Santa Ana, and Woodland Hills.  The Center
selected petitions from Los Angeles but not the other locations. 


   THE METHOD OF SELECTING
   LOCATIONS AND PETITIONS DOES
   NOT SUPPORT GENERALIZATION OF
   THE REPORT'S RESULTS
------------------------------------------------------------ Letter :7

The Center's researchers selected the 13 bankruptcy locations and
3,798 personal bankruptcy petitions without using scientific random
sampling techniques.  As a result, the national estimates presented
in the report's conclusions were not based on representative
probability sampling methods.  In addition, standard statistical
methods cannot be used to evaluate the likely accuracy of the Center
report's results.  Consequently, the methods used in the Center's
analysis do not provide a sound basis for generalizing the Center
report's findings to the annual 1996 filings in each of the 13
locations nor to the national population of personal bankruptcy
filings. 


      LOCATIONS JUDGMENTALLY
      SELECTED
---------------------------------------------------------- Letter :7.1

The 13 court locations used in the report were judgmentally selected
from large urban areas with a Credit Counseling Center\31 and large
bankruptcy caseloads.  The locations were also chosen to include
variations in other characteristics, such as the growth in bankruptcy
filings, the split between chapter 7 and chapter 13 filings, and
state-specific asset exemption levels for chapter 7.  Indeed, the
Center report showed that the courts that were included differed
considerably in the total number of filings, the proportion that were
chapter 7 and chapter 13 personal bankruptcy filings, and the change
in the total number of filings from 1995 to 1996.  Neither the court
locations nor petitions were chosen with the objective of identifying
the range of debts--lowest to highest--that bankruptcy debtors could
repay. 

The total number of personal bankruptcy petitions filed in 1996
varied greatly among the 13 court locations.  To account for this
fact, the Center report stated that the sample was weighted so that
the report's weighted estimates that combined information from all
locations represented the total filings from these 13 court
locations.  This means that the Center report's estimates were
strongly affected by those court locations that had the highest
number of personal bankruptcy petitions filed in 1996.  For example,
about 41 percent of all 1996 chapter 7 filings in the 13 court
locations were from Chicago and Los Angeles.  The 17 percent of the
sampled chapter 7 filings from Chicago and Los Angeles were therefore
inflated to correctly represent the relative size of the Chicago and
Los Angeles locations among the 13 locations.  All of the Center
report's weighted estimates, including those labeled as national
estimates, were weighted to represent only these 13 locations. 

The Center report's authors provided us with data, not included in
the report, that indicated that the predicted abilities of those who
filed for chapter 7 personal bankruptcy to repay debts varied
considerably among the 13 court locations (see table 3).  For
example, the percent of chapter 7 debtors whom the report determined
had some income available to repay debt ranged from a low of about 32
percent in San Diego to a high of about 67 percent in Dallas.  The
considerable variation among locations indicates that the repayment
rate at other locations, and for the nation as a whole, could differ
from the combined, weighted estimate for these 13 locations. 



                          Table 3
          
            Percent of Chapter 7 Debtors in Each
           Court Location Whom the Credit Center
           Determined to Have Positive Net Income
          That Could Be Applied to the Payment of
                      Nonhousing Debt

                         Percent of chapter 7 debtors with
                      positive net income available to pay
Court location               nonpriority, nonhousing debts
------------------  --------------------------------------
Atlanta, GA                                           52.7
Chicago, IL.                                          49.8
Dallas, TX                                            67.1
Houston, TX                                           60.2
Hartford, CT                                          52.2
Indianapolis, IN                                      43.1
Kansas City, MO                                       56.9
Los Angeles, CA                                       32.9
San Diego, CA                                         31.7
Memphis, TN                                           50.0
Phoenix, AZ                                           50.4
Pittsburgh, PA                                        34.0
Tampa, FL                                             36.8
----------------------------------------------------------
Source:  Credit Research Center data. 

The Center report's authors stated that its results cannot be
generalized to all personal bankruptcy petitions filed nationally
because the sample was not designed for this purpose.  Consequently,
the national estimates presented in the conclusion of the Center
report are not supported by the report's study methods. 


--------------------
\31 According to the report's authors, locations with Credit
Counseling Centers were chosen because the report we reviewed was
part of a larger research effort.  One purpose of this larger effort
was to compare debtors who did and did not use credit counseling in
these 13 locations. 


      PETITIONS NOT RANDOMLY
      SELECTED AMONG MONTHS OF THE
      YEAR OR DAYS OF THE MONTH
---------------------------------------------------------- Letter :7.2

The Center report states that the sampling procedures used to select
petitions from the 13 court locations resulted in a sample that was
representative of all petitions filed in those locations.  From our
review of available information on the report's sample design, we
have determined that statistical probability sampling methods were
not used to select the petitions filed within each court location. 
The Center's petitions were gathered from several months and
generally included the petitions filed in the first few days of the
months of May and June (eight locations); June only (three
locations); or July only (one location) of 1996.  In the remaining
location, the petitions were selected by the clerk of the bankruptcy
court during April, May, and June 1996.  Because the sample procedure
for selecting filings within bankruptcy court locations was not
random, the characteristics of the petitions drawn may be
systematically influenced by variation in the types of filings that
can occur (1) in different months throughout the year and (2) for
days within the month.  Consequently, standard statistical sampling
methods cannot be used to determine whether the results in the Center
report were likely to be representative of all bankruptcy filings in
each of the 13 court locations. 

The Center report evaluated the possibility that the petitions from
May to July that were included in the analysis might differ from
those filed during other months of the year by examining
supplementary data for other seasons from Indianapolis.  On the basis
of the Indianapolis analysis, the authors conclude that "a concern
that seasonal differences in petitions could lead to an overstatement
of the ability to repay debt across all petitions filed during 1996
is unwarranted." The Center report provided no basis for judging
whether the lack of monthly variation in Indianapolis could be
expected in all 13 court locations. 

The petitions within each court location were not selected from
filings over complete monthly periods and, therefore, could be
affected by variations in the characteristics of petitions filed at
different times of the month.  In a few court locations, because of
especially high filing volumes, the sample quotas were reached in the
first day or two of the month.  For example, our analysis of the
Center's data showed that about 95 percent of the petitions selected
in Dallas and Houston, Texas, were filed by the third day of the
month.  At both of these locations, the petitions drawn had been
filed prior to the first Tuesday of the month, the date on which
mortgages are foreclosed in Texas.  Thus, the petitions used from
these two locations may have included a disproportionate number of
debtors who sought to avoid mortgage foreclosures under chapter 13. 
The income and expenses for such filers may vary from those of
debtors who filed in these locations later in the month.  The
report's authors told us that they planned to sample additional
petitions in Dallas and Houston to examine this possibility. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :8

The comments and observations in our report are based on a review of
the final version of the Center report, dated October 6, 1997;\32
some additional information we requested from the report's authors;
data and analyses provided by the Federal Judicial Center (FJC) on
bankruptcy filings in the 13 locations used in the Center report;
telephone interviews with bankruptcy judges and trustees; and our
experience in research design and evaluation.  On November 13, 1997,
we met with Professor Michael Staten, coauthor of the report, to
discuss our questions and observations about the report.  Following
this meeting, Professor Staten and his coauthor, Professor John
Barron, provided additional information about the report's
methodology and some additional data that we requested.  We received
the last of these data on December 23, 1997.  The authors declined to
provide a copy of the automated database used for their analysis,
citing their interest in retaining its proprietary value. 

The team that reviewed the report included economists from our Office
of Chief Economist and specialists in program evaluation, statistical
sampling, and statistical analysis from our General Government
Division's Design, Methodology, and Technical Assistance group.  We
did our work principally between October 1997 and January 1998 in
Washington, D.C. 


--------------------
\32 John M.  Barron, Ph.D., and Michael E.  Staten, Ph.D., Personal
Bankruptcy:  A Report on Petitioners' Ability-to-Pay, October 6,
1997.  Earlier reports of the research findings were included in
testimony before the National Bankruptcy Review Commission on
December 17, 1996, and January 23, 1997, and before the Senate
Judiciary Subcommittee on Administrative Oversight and the Courts,
April 11, 1997. 


   COMMENTS FROM STUDY AUTHORS
------------------------------------------------------------ Letter :9

Professors Michael E.  Staten and John M.  Barron, authors of the
Center's report, provided written comments on a draft of this report. 
(See app.  I.) The authors discussed each of the report's five areas
of concern that, together, led to our conclusion that additional
research and clarification would be needed to confirm the accuracy of
the Center's report's conclusions regarding the proportion of debtors
who may have the ability to repay at least a portion of their
nonpriority, nonhousing debts and the amount of debt such debtors
could repay.  In discussing the five areas of concern, the authors
agreed with some concerns but believed that other concerns were
either overstated or unwarranted. 

Their specific comments on the concerns raised in this report are
discussed and evaluated at the end of appendix I.  We focus here on
the authors' major comments and our evaluation of those comments. 
Basically, the authors disagreed with us over the implications of the
concerns we raised.  They believe that the sample of bankruptcy cases
they examined was large enough and was taken from a sufficiently
varied cross-section of cities and courts to (1) reveal a significant
number of chapter 7 petitioners with some capacity to repay their
debts and (2) suggest a need for policymakers to reexamine whether
the current bankruptcy statutes should be changed.  They also believe
that determining debtors' ability to pay their eligible dischargeable
nonhousing debts, which the Center report did not do, was an
interesting but unimportant side issue.  Although they agreed that it
would be difficult to use their results to estimate with any
precision the repayment ability of chapter 7 debtors outside of their
sample, they believed that their sample results, regardless of the
concerns we found, strongly suggest a widespread substantial
repayment capacity.  They provided additional data and analysis, not
included in the Center's report, on reaffirmations of secured
nonhousing debt to further support their conclusions. 

We continue to believe that the concerns we found strongly suggest
that additional research and clarification are needed to determine
the accuracy of the Center report's conclusions regarding the
proportion of debtors who may have the ability to repay at least a
portion of their nonhousing debts and the amount of debt they could
potentially repay.  We note in this regard that the Credit Research
Center is currently conducting additional research with its
bankruptcy database, and the accounting firm of Ernst & Young is
conducting a study to address the concerns we raise in this report. 
The Center commented that the study clearly indicates a widespread
and "substantial" repayment capacity across all 13 locations in the
study.  We agree that the data and indicators used by the Center
showed that the percentage of debtors in each location with at least
some positive net income available for debt repayment was not so
small as to be negligible.  However, the assumptions, data, and
sampling procedures used in the Center report raise questions
concerning the accuracy and usefulness of the report's estimates and
require the reader to use caution in interpreting the types of firm
conclusions found in the Center report.  For example, the Center's
estimate of the percentage of debtors who have at least some capacity
to pay included all debtors whose monthly net income after expenses
was greater than zero, whether that amount was $1 or $1,000.  We were
not able to conclude, as the Center did, that there is a
"substantial" repayment capacity in every city because (1) we do not
have a basis for determining how much repayment capacity should be
considered substantial; and (2) we cannot conclude that the
petitioners' net income, as derived from data in their initial
schedules, can be accepted as an accurate estimate of debtors' net
income available for debt repayment for the following 5 years. 

Several factors suggest to us that those debtors with at least some
capacity to pay would not be able to repay as much debt as the Center
report assumed.  For example, historically only about one-third of
chapter 13 debtors have completed their repayment plans, suggesting
that for two-thirds of debtors something changed between the time the
plans were confirmed by the bankruptcy court and the time the actual
repayment plan was to be successfully completed.  To the extent that
debtors are unable to maintain their debt repayments for the full
5-year period assumed in the report, the amount of debt repaid would
be less than that assumed in the report.  In addition, the Center's
estimates of repayment capacity do not include any provision for the
administrative costs of administering a repayment plan.  In fiscal
year 1996, 14 percent of the payments from chapter 13 debtors was
used to pay administrative and legal costs. 

The Center report provided an estimate of the potential repayment
capacity of debtors who have filed for bankruptcy to pay their
nonpriority, nonhousing debts.  We do not agree with the Center that
identifying the universe of dischargeable debts that a debtor may
have the capacity to repay is an interesting, but unimportant, side
issue in assessing a debtor's ability to repay his or her nonhousing
debts.  It is the debtor's total eligible dischargeable debts that
represent the potential loss to creditors if the bankruptcy court
grants the debtor a discharge of all his or her eligible
dischargeable debts.  The Center report did not attempt to identify
this universe of debts in its analysis.  Creditors are not at risk in
the bankruptcy process for debts that are nondischargeable or debts
that the debtor reaffirms.  Similarly, creditors are not at risk
through the bankruptcy process for the dischargeable debts of those
debtors whose bankruptcy cases are dismissed.  With few exceptions,
these debtors remain personally responsible for all their debts.  The
relevant universe of debtors who pose a risk of nonpayment to
creditors through bankruptcy are those who complete the bankruptcy
process and receive a discharge of all or part of their eligible
dischargeable debts.  The Center report did not attempt to estimate
the capacity to pay of this universe of debtors.  Instead, the
Center's assessment of capacity to pay included those debtors who may
have received a discharge plus those debtors whose cases were
dismissed and did not receive a discharge.  Consequently, the Center
report's universe of debtors included debtors who remained
responsible for their eligible dischargeable debts because their
cases were dismissed. 

The Center agrees that the Center report's findings were not based on
data from a nationally representative scientific, random sample.  The
Center comments that the researchers did not intend to obtain a
nationally representative sample and that much useful information can
come from samples that are not nationally representative.  Although
decisions with nationwide implications could be based on evidence
from selected locations, we believe that the assumptions, data, and
methods used in the Center report require that its
conclusions--which, in some cases, are stated as broad national
estimates--be interpreted with caution. 

The additional data provided in the comment letter are helpful; but,
as discussed in our comments at the end of appendix I, we did not
have the database used for these analyses to verify the results. 
More importantly, these new data do not resolve many of the concerns
we raise in this report.  For example, the weighting methodology used
to develop the weighted estimates presented in the new tables is the
same methodology used for the Center report's other estimates and is
subject to the same limitations we discussed in our report.  As with
the Center's other estimates, the assumptions used in the new
analyses assumed that 100 percent of debtors' discretionary income
and 100 percent of the proceeds from the sale of the debtors'
nonexempt assets would be used to repay debt.  In practice,
administrative costs would reduce the amount paid to creditors. 

Thus, notwithstanding the comments and additional information
provided by the Center report's authors, we continue to believe that
more research would be needed to verify and refine the Center
report's estimates of debtors' repayment capacity to better inform
policymakers. 

We are providing copies of this report to the Chairman and Ranking
Minority Member of the Senate and House Committees on the Judiciary;
the Chairman and Ranking Minority Member of the Subcommittee on
Commercial and Administrative Law, House Committee on the Judiciary;
and to the authors of the Credit Research Center report.  We will
also make copies available to others upon request. 

If you have any questions, please call me at 512-8777. 

Richard M.  Stana
Associate Director
Administration of Justice Issues




(See figure in printed edition.)Appendix I
COMMENTS FROM THE CREDIT RESEARCH
CENTER
============================================================== Letter 



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on specific issues included in the
letter dated January 21, 1998, from Professor Michael Staten, on
behalf of himself and his coauthor, Professor John Barron.  Other
issues discussed in the letter have been included in the report text. 

GAO COMMENTS

1.  The authors agreed that there is a need to validate debtors'
income, expenses, and debts in developing assumptions of future
income and expenses but stated that a researcher currently has no
recourse but to accept what the debtor advises the court under oath
at the time the petition is filed.  We understand that researchers
must use the best available data and that, currently, verifiable data
on debtors' income and expenses during bankruptcy have not been
developed.  However, our intent was to indicate that the Center
report should have discussed how the use of data from debtors'
initial schedules could affect the Center report's results and, thus,
how those results should be used.  In this case, it seems the
researchers could have used more recent data, at least for some
debtors, because debtors may amend their initial schedules at any
time prior to the final disposition of their bankruptcy cases.  Such
amendments could alter the estimated income, estimated expenditures,
and debts that debtors reported on their initial schedules.  We
recognize that obtaining these amended schedules would have required
additional time and resources.  However, we believe that the
importance of these data to the overall conclusions in the Center's
report would justify such an effort. 

The authors also said that to the extent there is a bias in the
debtors' initial schedules, it would be expected that the debtors
would understate their capacity to repay debt.  Although this may
seem logical at first glance, it is important to note that there are
no empirical data on the accuracy of the data reported in debtors'
initial schedules.  Nor is there any empirical basis for assuming
that debtors would consistently attempt to understate their capacity
to pay their debts.  In fact, there is no empirical basis for
assessing whether debtors generally overstate or understate their
capacity to repay on their initial schedules or the general amount of
the overstatement or understatement.  There may be several reasons
why some debtors would actually overstate their capacity to pay.  For
example, some people may simply not want to admit how serious their
financial situation has become in order to protect certain assets. 
Also, mistakes could be made in the schedules used in the Center's
analysis, which are not easily interpreted by debtors who might
proceed without legal or financial assistance.  For example, in Los
Angeles, a location whose data contributed significantly to the
Center's final weighted estimates, Center data showed that about
one-third of debtors reported they had no lawyer.  Through mistakes
in filling out the schedules, debtors could report information that
would have the effect of either overstating or understating their
capacity to pay their debts. 

2.  The Center stated that its calculations provided a benchmark of
debtors' ability to pay that could easily accommodate whatever
assumptions about possible income changes the reader wished to make. 
It also agreed with us that incorporating a cushion into a chapter 13
repayment plan to guard against income interruptions or unexpected
expenses seemed to be a prudent step. 

We agree that the Center report provided a baseline estimate of
debtors' ability to pay that would change as the report's basic
assumption--that debtors' income and expenses would remain unaltered
for 5 years--changes.  However, the Center provided no estimates
based on alternative assumptions of repayment capacity, and without
the Center's database, it is not possible for anyone to estimate the
effect of such alternative assumptions on the Center report's
estimates of debtors' potential repayment capacity.  Since many
economic factors can change in a debtor's financial situation during
5 years, it would seem prudent to base any policy decisions on a
wider range of assumptions than the somewhat optimistic set of
assumptions used in the Center study.  For example, the assumption
that debtors' reported income and expenditures would remain unchanged
for 5 years had the effect of providing optimistic estimates of
debtors' repayment capacity in two ways:  (1) it did not allow for
situations where the debtors' income decreases or expenses increase,
thus discretionary income available to pay debt was assumed to remain
unchanged for 5 years; and (2) 100 percent of this discretionary
income was assumed to be used for 5 years to repay debt, when in fact
a portion of the debtors' discretionary income would be used to pay
the expenses of administering the debtors' repayment plans. 

There is some additional evidence that the Center's assumption that
debtors' income and expenses would remain unchanged for 5 years may
be optimistic.  For example, the AOUSC report discussed on page 8 of
our report showed that only about 36 percent of chapter 13 debtors
completed their repayment plans.  The reasons for this low completion
rate are unknown, but it illustrates the high level of discrepancy
between the amount that debtors could potentially repay, based on the
data and assumptions used in the Center report, and what has actually
occurred over a 10-year period.  In addition, in virtually all cases,
creditors do not receive 100 percent of debtors' payments under
chapter 13 repayment plans.  Fiscal year 1996 data from the Executive
Office of U.S.  Trustees showed that 14 percent of payments were used
to pay the debtors' lawyers, the chapter 13 trustees' statutory
operating expenses in administering the plans, and other
administrative expenses. 

3.  The Center said that it believes it has clearly identified the
universe of debts for which it estimated debtors' ability to pay as
all "debts not secured by real estate, without drawing a distinction
between secured vs.  unsecured, priority vs.  non-priority, or
dischargeable vs.  non-dischargeable." The Center commented that the
distinction between dischargeable and nondischargeable debts is
simply "an interesting side issue." The Center said that such
distinctions between categories of debt were not necessary if the
report's intent was to assess debtor's overall ability to meet their
obligations.  The Center also said that unsecured priority debt was
not included in the base of total unsecured debt for many of the
repayment calculations, because the report assumed that unsecured
priority debt would be paid before unsecured nonpriority debt. 

We do not agree that the distinction between dischargeable and
nondischargeable debt is just an interesting side issue.  The
distinction is important if the Center's data are to be used for
considering the need to alter existing bankruptcy statutes.  It is
the debtor's total eligible dischargeable debts that represent the
potential loss to creditors if a debtor is granted a discharge of his
or her eligible dischargeable debts.  The Center did not attempt to
identify this universe of debts in its analysis.  Creditors are not
at risk in the bankruptcy process for debts that are nondischargeable
in bankruptcy or for eligible dischargeable debts that the debtor
reaffirms.  Total dischargeable debts are total debts less total
nondischargeable debts.  As discussed on pages 9 and 10 of our
report, the Center report may not have fully identified all eligible
dischargeable debts, because it excluded data on unexpired leases
from Schedule G, such as automobile leases.  Thus, the Center did not
identify that universe of debts for which creditors are at risk in
the bankruptcy process. 

As we note on pages 9-10 of our report, to assume that all unsecured
priority debts would be fully paid over 5 years but that no other
class of nonhousing debts would be fully paid creates a disparity in
the treatment of nonhousing debts that does not reflect actual
bankruptcy practice.  In chapter 13 repayment plans, secured debts
would ordinarily be paid before or concurrently with unsecured
priority debts.  Consequently, the Center report's calculations did
not provide an estimate of the amount of unsecured nonpriority debt
that could be repaid.  If the Center report's purpose was simply to
identify debtors' overall ability to pay nonhousing debts from net
income after reported expenses, then the report should have included
unsecured priority debt with all other nonhousing debts-- secured and
unsecured--and calculated debtors' ability to pay the resulting total
nonhousing debt. 

4.  The Center agrees with us that its calculations of debtors'
ability to repay their nonhousing debts did not consider the payments
required to pay the nonhousing debts that debtors stated it was their
intent to reaffirm (repay).  The Center notes that the February 1997
testimony of law professors Marianne Culhane and Michaela White
before the National Bankruptcy Review Commission stated that about 50
percent to 60 percent of intended reaffirmations (the data used in
the Center report) actually result in signed reaffirmation agreements
in which debtors reaffirmed their debts.  Thus, they noted it is
possible that the number of final reaffirmations could be less than
that reported in debtors' statements of intent. 

We agree that the number and dollar value of debts that debtors
ultimately reaffirm could be more or less than those found in
debtors' statements of intent.  We believe that this further supports
our overall conclusion that the results in the Center's report should
be viewed with caution.  In September 1997, professors Culhane and
White reported updated results of their study, which were based on
debtor reaffirmations in only 7 of the 90 bankruptcy districts, and
thus must be considered illustrative, not conclusive.  Nevertheless,
the reaffirmation report's findings provide additional evidence that
one should be cautious in interpreting conclusions based solely on
debtors' initial schedules, such as schedules of income and expenses
as well as reaffirmations.  For example, the reaffirmation report
found that debtors filed fewer reaffirmations than indicated in their
statements of intent and that the debts that debtors ultimately
reaffirmed were often quite different from those that debtors stated
it was their intention to reaffirm.  The reaffirmation report and the
Center's data indicated that debtors rarely stated their intention to
reaffirm unsecured debts.  However, the reaffirmation report found
that debtors in fact ultimately reaffirmed unsecured debts as well as
debts that were not listed in their initial schedules at all.  The
reaffirmation report also noted that court records provide an
incomplete picture of reaffirmations, because debtors may also sign
reaffirmations with creditors that the creditors fail to file with
the court, as required. 

In addition, the reaffirmation report reinforces our concern that
local court bankruptcy practice and rules may affect the data that
debtors report on their initial schedules and in the data found in
debtors' court files generally.  For example, the reaffirmation study
found that the number of final reaffirmation agreements filed with
the bankruptcy court in each district appeared to be affected by
governing court decisions for the districts studied.  In two
districts, the debtor could keep property, such as a car, by simply
maintaining ongoing contractual payments on the property.  Thus, it
was not necessary for the debtor to file a reaffirmation agreement
with the court in order to keep the property.  In two other
districts, court decisions required the debtor to file a
reaffirmation agreement or surrender or redeem the property.  The
number of final reaffirmation agreements was lower in those districts
that did not require a reaffirmation agreement in order for the
debtor to keep the property.  However, the report said that the data
did not permit an empirical evaluation of the extent to which such
controlling court decisions affected the number and type of
reaffirmations that debtors in the report ultimately filed with the
bankruptcy courts. 

5.  The Center comments included data and analyses, not previously
provided, that the Center said address the impact of reaffirmations
on debtors' ability to pay their nonhousing debts.  These new
analyses are based on weighted data for the 13 locations included in
the Center's study.  We cannot assess the accuracy of the data in the
tables because we do not have the database used to develop these
tables and, therefore, cannot replicate how the new estimates were
derived.  However, we do have some overall observations on these new
data. 

First, the weighted data are based on the same weighting methodology
used for the Center report's other estimates and, therefore, are
subject to the same limitations of that weighting methodology that we
noted in our report.  The weights are heavily influenced by filings
in two locations--Chicago and Los Angeles--which accounted for about
41 percent of all bankruptcy filings in the 13 locations. 

Second, the tables presented in the comments need clarification in
their presentation.  For example, table 1 of the comments does not
indicate that all dollar amounts in the table are averages, which
they are.  The table also does not clearly indicate that the amount
of nonhousing debt shown is the total amount of nonhousing
debt--secured and unsecured--less unsecured priority debt. 

Third, the assumptions underlying the data in table 2 are not
explained.  For example, line "D" of table 2 is supposed to represent
the amount of unsecured nonpriority debt that could be paid over 5
years from future income after liquidating all of the debtor's
nonexempt property, if any.  The calculation appears to assume that
(1) when surrendered and liquidated, the collateral would bring 100
percent of the value of the collateral, as listed in the debtor's
initial schedules; and (2) 100 percent of the proceeds realized from
the liquidation would be used for repaying the debt secured by the
collateral.  We found no basis for either of these assumptions.  For
example, when a debtor's nonexempt assets are liquidated to pay
creditors, the asset may bring more or less than the value of the
collateral as listed in the debtor's schedules.  Moreover, there are
usually expenses associated with liquidating a debtor's nonexempt
assets, such as statutory bankruptcy trustees' commissions and
appraiser or auctioneer fees.  Such expenses would reduce the amount
paid to creditors because these costs would be paid before any
remaining proceeds were distributed to creditors. 

The data in the new tables are subject to the same limitations as
other estimates of debtors' ability to pay included in the report. 
The tables are based on the assumptions, used throughout the report,
that debtors' income and expenses would remain unchanged over a
5-year period and that 100 percent of a debtor's discretionary net
income will be used for debt repayment.  As previously discussed,
both logic and available evidence would suggest that these are not
realistic assumptions.  For example, the Center provided us data, not
included in its report, which showed that the majority of nonhousing
secured debt was vehicle debt.  The data in the new tables 2 and 4
provided with the Center's comments assumed that the debtor's
automobile would be sold, and no replacement obtained.  The absence
of an automobile could very well affect a debtor's employment and,
thus, a debtor's future stream of income. 

6.  The Center's comments noted that it would be difficult to
estimate with precision debtors' ability to pay their nonhousing
debts in any location other than the 13 locations included in the
Center report.  On the other hand, the Center concluded that debtors'
data in all 13 locations showed a substantial repayment capacity,
despite the great diversity in the characteristics of the 13
locations, such as unemployment rates and the percent of total
personal bankruptcy cases that were chapter 7 cases.  The Center
stated that this showed that substantial repayment capacity is a
widespread phenomenon, whether or not the report's findings are
applicable to other locations. 

We agree that the Center's data show that some debtors who file for
bankruptcy under chapter 7 may have some capacity to repay their
debts.  But, from a policymaking standpoint, the more relevant
questions are whether the Center report's findings provide a
reasonable estimate of that repayment capacity and whether the
Center's defined universe of debtors and debts used to estimate
repayment capacity was appropriate for assessing the need for a
change in current bankruptcy laws.  As previously discussed, we
believe the Center's universe of both debts and debtors may not be
the appropriate ones for assessing whether current bankruptcy
statutes should be changed.  In answering these questions, it is also
important to note that the data used for the Center report were based
on information debtors provided at a single point in time--the time
they filed for bankruptcy--regardless of whether or not they
completed the bankruptcy process and received all or part of the
relief they sought in filing for bankruptcy.  Thus, the report
included data from debtors who may have withdrawn their petitions
voluntarily, had their petitions dismissed by the court, or who
received bankruptcy court discharges of all or part of their eligible
dischargeable debts.  For example, in Los Angeles, of those chapter 7
petitions filed on the same days of May and June 1996 as those
petitions used in the Center sample, about 5 percent had been
dismissed by September 30, 1996.  For chapter 13 petitions, more than
30 percent had been dismissed during the same period.  In contrast,
not more than about 4 percent of chapter 7 and 13 petitions in San
Diego had been dismissed within 90 days.  Because the report's
findings include debtors who did and did not receive a discharge of
their eligible debts, the report's findings cannot be used to reach
conclusions about the most relevant public policy question--the
potential ability to pay of debtors who received a discharge of all
or part of their eligible dischargeable debts. 

7.  The Center agreed with our general conclusion that scientific,
random sample methods were not used to select the bankruptcy
petitions used in the Center's analysis.  However, the Center said
that the lack of a scientific, random sample did not necessarily
diminish the usefulness of the Center report's findings.  The Center
commented that it did not intend to obtain a nationally
representative probability sample and agrees that it did not use a
scientific random sampling methodology to select the 13 bankruptcy
locations or the bankruptcy petitions used in the analysis.  The
Center also states that most social science research is conducted
with samples that are not nationally representative probability
samples and concludes that much useful information comes from samples
that are technically less ambitious than the standard that we
applied. 

Our evaluation assumed that the Center report may be used for
important policymaking on a national scale.  As a result, we believe
that it is appropriate to inform the Committee that the Center
report's data do not meet scientific standards for estimating the
characteristics of bankruptcy debtors for the United States as a
whole or for all bankruptcy debtors in each of the 13 locations. 

8.  The Center discussed our observations on its methods of selection
for each of the three steps at which petition selections were made
without probability selection methods. 

The Center agrees with us that the 13 locations were not selected
using probability selection techniques and, thus, may or may not be
representative of the remaining courts in the United States.  The
Center commented that nonprobability samples have been used in some
previous studies of bankruptcies, including a study by GAO, and that
the purpose of the Center study was to form and test hypotheses about
potential causal factors.  The Center also stated that the study has
potential value for policymakers because the large sample from a
varied cross-section of courts identifies significant numbers of
petitioners with some capacity to repay debts and because there was a
finding of substantial repayment capacity in every city in the study
despite the great diversity in city/court characteristics. 

We agree that this diverse set of 13 locations demonstrates that
based on the data and assumptions the Center used, the Center's
indicators of debtors' ability to repay debts are found at greater
than negligible rates at all locations.  However, we also concluded
that users of the Center data should consider the variation among
locations and the lack of a national estimate as limitations.  The
important variations between the studied cities might be of
importance for some policy purposes.  In addition, we cannot confirm
the Center's conclusion that there is a "substantial" repayment
capacity in every city, because we do not have a basis for
determining how much repayment capacity should be considered to be
substantial and because, as explained above, we can not conclude that
petitioners' reports of income on bankruptcy petitions can be
accepted as an accurate estimate of income for the following 5 years. 

9.  The second sampling issue on which the Center commented was what
the Center referred to as seasonality--the fact that the Center's
petitions were filed in the spring and summer months and might not be
representative of petitions filed at other times of the year.  The
Center stated that ample evidence from previous researchers and
supplemental testing in the Center study suggest that the potential
bias from focusing only on cases filed in the spring and summer
months is negligible.  The Center cited as evidence the findings from
two previous studies and the Center's analysis of Indianapolis
petitions from its current study. 

The Center and we agree that the study petitions were drawn from a
limited part of the year.  We did not have a sufficiently strong
basis to conclude that seasonality factors could or could not have
possibly affected the Center report's estimates of the debtors'
repayment capacity.  The two previous studies cited in the Center's
comments do not address the effect of season on the debtors' ability
to repay debts but only examined the number of chapter 7 and chapter
13 bankruptcy filings by season.  The results of the Center's
analysis do show that season of the year did not affect estimates of
the ability to pay in one city, Indianapolis.  We agree that we do
not have a strong theoretical reason for expecting a seasonal effect. 
However, in the absence of evidence from more than one location, and
in view of the fact that the present study is strongly concentrated
by season, we continue to believe that the season in which the
petitions were selected should be considered a limitation in
interpreting the results from the study. 

10.  The third sampling issue addressed in the Center comments was
the time of the month from which the Center's petitions were drawn. 
The Center agrees that the bankruptcy petitions used in the study
were generally drawn from days early in the month.  The Center
explains that the petitions were drawn from the beginning of the
month to maintain tight control over the petition selection procedure
and to minimize uncertainty about the characteristics of cases that
were not studied.  The Center maintains that Texas is the only one of
the 13 locations where there is evidence or reason to believe that
cases early in the month might differ from those late in the month. 
In Texas there were a disproportionate number of past-due home
mortgage chapter 13 petitions early in the month, and the Center said
it was now drawing additional cases in Texas.  In addition the Center
notes that although chapter 13 petitions might differ by time of
month in Texas, it is not clear why the characteristics of the
chapter 7 cases would differ over the month.  The Center also notes
that the study did not find differences in the values of variables
that measure ability to pay at the one location, Indianapolis, that
could be tested with the data from the Center study. 

We believe that the differences among petitions at different times of
month in Texas should be considered.  Those debtors who file for
chapter 13 early in the month to prevent a mortgage foreclosure may
have different financial characteristics from chapter 13 debtors
filing later in the month.  The only clear evidence of the absence of
a time-of-month effect comes from a single court in Indianapolis.  We
are concerned that there may be other court-specific factors of which
we are unaware.  For example, working with the Federal Judicial
Center, we learned that mortgage foreclosures early in the month also
could affect the type of filings early in the month in Atlanta. 
Although the Atlanta filings for this study did not happen to be
concentrated at the beginning of the month, the Atlanta example
indicates that it is difficult to exclude time-of-month effects. 
Thus, we believe that the lack of representativeness by time of month
should be considered in evaluating the study. 


*** End of document. ***