oversight

Fort Worth Housing Authority, Fort Worth, TX, Did Not Properly Apply Its Section 8 Subsidy Size Standards for Existing Tenants

Published by the Department of Housing and Urban Development, Office of Inspector General on 2005-10-21.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                                                           Issue Date
                                                                    October 21, 2005
                                                           Audit Report Number
                                                                        2006-FW-1001




TO:         Justin Ormsby
            Director, Office of Public Housing, 6APH


FROM:       Frank E. Baca
            Regional Inspector General for Audit, 6AGA

SUBJECT: The Fort Worth Housing Authority, Fort Worth, Texas, Did Not Properly Apply
         Its Section 8 Subsidy Size Standards for Existing Tenants


                                   HIGHLIGHTS

 What We Audited and Why

             As part of the Inspector General’s annual audit program, we audited the Fort
             Worth Housing Authority’s (Authority) Section 8 program. We wanted to
             determine whether the Authority properly applied the subsidy size standards in its
             administrative plan.


 What We Found


             The Authority had sound policies regarding assignment of tenant subsidy size.
             However, it did not follow its policy when it neglected to change voucher sizes
             for as many as 382 tenants between January 2003 and March 2005. This resulted
             in $5,951 in ineligible payments, $174,667 in unsupported payments, and
             potential overpayments of $521,744 over the next 3.7 years. The Authority could
             better use its Section 8 funding and avoid future overpayments by strictly
             applying the subsidy size standards in its administrative plan, as required by U.S.
             Department of Housing and Urban Development (HUD) regulations. Further, the
           Authority’s Section 8 department enacted policy changes without approval from
           its board of commissioners.

What We Recommend


           We recommend that you require the Authority to: (1) repay ineligible housing
           assistance overpayments of $5,951; (2) repay or support questioned costs of
           $174,667; (3) develop and implement procedures to ensure it assigns the correct
           subsidy size for all tenants to better use $521,744 in Section 8 funding; and (4)
           institute controls to ensure that the board of commissioners approves any program
           changes before being implemented.

           For each recommendation without a management decision, please respond and
           provide status reports in accordance with HUD Handbook 2000.06, REV-3.
           Please furnish us copies of any correspondence or directives issued because of the
           audit.

Auditee’s Response


           We provided the Authority a draft report on September 20, 2005, and held an exit
           conference on September 23, 2005. At the exit conference, Authority officials
           said that they had provided us with the wrong subsidy standards during the audit.
           In its October 4, 2005 written response to the draft audit report, the Authority
           apologized for providing incorrect subsidy standards and disagreed with the audit
           results, which were based upon the incorrect subsidy standards. As a result of the
           Authority’s response, we modified Finding 1 of the report to reflect analysis using
           the correct subsidy size standards. Further, we added a finding on the Authority
           implementing policy not enacted by its board of commissioners. The Authority’s
           response and our evaluation of the response are in Appendix B of this report.




                                            2
                            TABLE OF CONTENTS

Background and Objectives                                                            4

Results of Audit
        Finding 1: The Authority Did Not Properly Apply Its Section 8 Subsidy Size   5
        Standards for Existing Tenants
        Finding 2: The Authority Enacted Policy Changes without Approval from Its    9
        Board of Commissioners

Scope and Methodology                                                                12

Internal Controls                                                                    14

Appendixes
   A. Schedule of Questioned Costs                                                   16
   B. Auditee Comments and OIG’s Evaluation                                          17




                                            3
                     BACKGROUND AND OBJECTIVES

Fort Worth Housing Authority

The Fort Worth Housing Authority (Authority) was created in 1938 by the City of Fort Worth
and is governed by a five-member board of commissioners. It administers more than 4,800
Section 8 vouchers with funding of $29.9 million in fiscal year 2005. The majority of this
funding is for the Housing Choice Voucher program. The U.S. Department of Housing and
Urban Development (HUD) allocated an additional $2.5 million in funding for the Authority to
administer its Section 8 programs.

Section 8 Tenant-Based Assistance: Housing Choice Voucher Program

The Authority uses its Section 8 funding to provide rental subsidies so eligible families can
afford decent, safe, and sanitary housing in the private market. Participants may choose any
housing that meets program requirements. The Authority pays a housing subsidy directly to the
landlord on behalf of the participating family, who then pays the difference between the actual
rent and the subsidy amount. The Authority determines eligibility based on income and family
size in accordance with its administrative plan. The Authority verifies family income and
composition annually and ensures the unit meets minimum housing quality standards.

Section 8 Project-Based Certificate Program

Project-based assistance has the same purpose and eligibility requirements as the Housing
Choice Voucher program, but the assistance is tied to the housing unit. Families who move from
the assisted unit do not have a right to continued assistance. The Authority managed funding for
four project-based developments during the audit period that were included in the audit scope.

Our audit objective was to determine whether the Authority properly applied the subsidy size
standards in its administrative plan.




                                               4
                                         RESULTS OF AUDIT

Finding 1: The Authority Did Not Properly Apply Its Section 8 Subsidy
Size Standards for Existing Tenants
The Authority paid for tenants to live in units larger than allowed under its administrative plan
and HUD regulations. Although the Authority consistently applied its subsidy size standards for
new Section 8 tenants, it did not reissue vouchers as required when existing tenants reported
decreases in family size. As a result, from January 2003 to March 2005, the Authority made
$5,951 in ineligible overpayments and $174,667 in unsupported housing assistance payments. In
addition, the Authority may incur up to $521,744 in overpayments over the next 3.7 1 years
unless it properly applies its subsidy size standards to existing tenants.




    Authority Establishes Subsidy
    Sizes


                   HUD regulations required the Authority’s board of commissioners to adopt a
                   written administrative plan that establishes local policies for administration of its
                   Section 8 program. The Authority must administer its program in accordance
                   with its administrative plan. 2 In its plan, the Authority must establish subsidy size
                   standards that determine the number of bedrooms needed for families of different
                   sizes. 3 Table 1 shows the subsidy size standards in effect during the audit period.

                                       Number of        Number of Persons
                                       Bedrooms        Minimum Maximum
                                           0              1          1
                                           1              1          2
                                           2              2          4
                                           3              3          6
                                           4              5          8
                              Table 1: Fort Worth Housing Authority subsidy size standards




1
      The median length of time in program for those receiving assistance in the 50 largest metropolitan statistical
      areas is 3.7 years. HUD Office of Policy Development and Research (2003).
2
      24 Code of Federal Regulations 982.153.
3
      24 Code of Federal Regulations 982.54(a) and (c).


                                                            5
    The Authority Did Not
    Change Vouchers When
    Required


                 The Authority consistently assigned subsidy size to new tenants entering the
                 program within its approved standards. However, the Authority did not always
                 reissue smaller vouchers to existing tenants when circumstances required it to do
                 so. Also, some tenants held vouchers that were too small for their families. HUD
                 regulations allowed the Authority to grant exceptions to its subsidy size standards
                 under certain conditions. 4 The Authority’s administrative plan required tenants to
                 request exceptions in writing and provide documentation as to why a larger
                 subsidy size was necessary. In addition, the Authority’s policy required that
                 approval of an increased voucher size would be dependent upon the financial
                 feasibility of the program. However, the Authority’s staff allowed exceptions for
                 existing tenants without explanation or documentation of tenant requests for
                 exceptions. The files contained no supervisory approval for the exceptions or
                 evidence the Authority weighed the financial feasibility of the voucher size.
                 While the Authority did not require supervisory approval of exceptions, such a
                 procedure would serve to enforce the Authority’s subsidy size standards and
                 facilitate a financial feasibility review.

                 A review of the tenant files indicated incorrect subsidy size assignments occurred
                 when:

                      ƒ   A tenant reported a decrease in family size;

                      ƒ   The decrease in family size reduced the subsidy size for which the family
                          qualified, requiring the Authority to issue a new voucher; and

                      ƒ   The housing counselor neglected to change the voucher and recalculate the
                          subsidy payment.

                 In most instances, the unit rent was less than the maximum subsidy for the correct
                 voucher size. For example, a family consisted of a man and his son with a two-
                 bedroom voucher. The son moved out after reaching adulthood, leaving the man
                 as the only person in the unit. Under the Authority’s subsidy size standards, a
                 single person only qualified for a one-bedroom voucher. The son’s move
                 required the Authority to change the man’s voucher from two bedrooms to one.
                 In this case, the man’s $615 rent for a two-bedroom apartment was less than the
                 $643 maximum subsidy for a one-bedroom voucher. It appears the housing
                 counselor did not change the voucher because the rent was below the maximum
                 subsidy amount and there was no monetary impact for the Authority if the man
                 remained in his $615 apartment.


4
     24 Code of Federal Regulations 982.402(b)(8).


                                                     6
                In a similar case, a housing counselor simply included a note on the recertification
                checklist that indicated the rent was within the maximum subsidy amount.
                Because rent amounts were often lower than the maximum subsidy amounts, this
                practice did not frequently result in subsidy overpayments. The Authority’s
                Section 8 managers acknowledged this was not appropriate and commented that
                staff needed additional training on the issue. The Authority should require its
                housing counselors to change vouchers under these circumstances to prevent
                future subsidy overpayments resulting from rent increases or tenant relocations.

                Analysis of the Authority’s housing assistance payments records showed 382 of
                6,914 tenants (5.5 percent) held vouchers larger than allowed. The records further
                indicated that the Authority made possible overpayments for 125 of the 382 tenants.
                In addition, the Authority made $5,951 in ineligible subsidy overpayments for four
                tenants because it did not reduce voucher sizes when tenants reported changes in
                family composition. Unless the Authority can provide support that the subsidy
                payments were appropriate for the remaining 121 tenants, it should repay $174,667
                in questionable payments related to these vouchers. The Authority did not overpay
                for the remaining tenants whose vouchers were too large because the rents were
                below the maximum subsidy amounts for the appropriate subsidy sizes.
                Nonetheless, the Authority should correct or support the voucher assignment for all
                382 tenants identified in the analysis to prevent possible future overpayments and to
                comply with its administrative plan.

                Analysis of the Authority’s tenant data also identified tenants whose vouchers
                were too small for their families. In March 2005, the Authority had 88 tenants
                whose vouchers were too small based on its subsidy size standards. In 13 cases,
                the Authority assigned a family of seven a voucher for three or fewer bedrooms.
                Under its subsidy size standards, the Authority should have issued a family of
                seven a four-bedroom voucher. The Authority should take steps to ensure it
                issues vouchers of the appropriate subsidy size for all families.

    The Authority Could Better Use
    Funds by Applying its Subsidy
    Size Standards Correctly

                In addition to unsupported payments during the audit period, the Authority will
                continue to overpay housing assistance for tenants whose vouchers are too large
                until it corrects them. Based upon an analysis of the Authority’s March 2005 tenant
                data, 185 of its 4,809 tenants (3.8 percent) held vouchers larger than allowed under
                the Authority’s subsidy size standards. For 108 of the tenants, the Authority may
                have overpaid as much as $11,751 5 in March 2005. It did not overpay for the
                remaining 77 tenants whose vouchers were too large because their rent was less than
                the maximum subsidy amount for the correct voucher size.


5
    Amount included in the $174,667 questioned above.


                                                        7
                  To avoid future overpayments, the Authority should correct or support the
                  voucher sizes for all tenants whose vouchers were too large. Because the tenant
                  files showed the Authority had not correctly applied its subsidy size standards for
                  some tenants, the same situation will likely continue unless the Authority takes
                  corrective action. If the Authority correctly applies its subsidy size standards for
                  all tenants, it could put as much as $521,744 in housing assistance to better use
                  over 3.7 years 6 by avoiding subsidy overpayments. The Authority could better
                  document and enforce its compliance with the requirements in its administrative
                  plan by requiring supervisory review and approval for all exceptions and
                  following up on any exceptions identified through quality control procedures.

    Conclusion


                  For new program participants, the Authority applied its subsidy size standards in
                  accordance with its administrative plan. However, when circumstances required
                  the Authority to change an existing tenant’s subsidy size and issue a new voucher,
                  it often did not. As a result, the Authority made $5,951 in ineligible
                  overpayments and $174,667 in unsupported housing assistance payments between
                  January 2003 and March 2005. In addition, if the Authority corrects the voucher
                  sizes for all 382 tenants whose vouchers were too large as of March 2005, it could
                  put an estimated $521,744 in housing assistance to better use over the next 3.7
                  years.

    Recommendations



                  We recommend that HUD require the Authority to:

                  1A. Repay ineligible housing assistance overpayments of $5,951.

                  1B. Support or repay unsupported housing assistance payments of $174,667.

                  1C. Develop and implement procedures to ensure it assigns housing vouchers in
                      accordance with the subsidy size standards in its administrative plan, which
                      could result in an estimated $521,744 in funding being put to better use.




6
     Housing subsidy calculations are based on a number of interrelated factors that often change over time,
     including family composition, income, and unit rent. As a result, the potential overpayments identified here are
     estimates of potential cost savings from downsizing tenants’ vouchers in accordance with the Authority’s
     subsidy size standards ($11,751 x 12 months x 3.7 years = $521,744).



                                                          8
Finding 2: The Authority Enacted Policy Changes without Approval
from Its Board of Commissioners
The Authority’s Section 8 department enacted policy changes without approval from its board of
commissioners. In an attempt to save money, the former director of assisted housing 7 (director)
established and instructed staff to implement subsidy size standards stricter than those defined in
the approved administrative plan. Although the director’s initiative began to achieve the desired
cost savings, the board of commissioners never adopted the stricter policy. By implementing
policy changes without approval from its board of commissioners, the Authority acted outside its
administrative plan and HUD regulations. However, if the board of commissioners were to
approve the stricter standards, the Authority could save as much as $2.12 million over the next
3.7 years.



     Policy Changes Enacted
     without Board Approval


                   In HUD’s fiscal year 2005 appropriations act, Congress required housing authorities
                   to control the increasing costs of vouchers. In February 2005, HUD issued a notice 8
                   providing several cost-savings suggestions to housing authorities in response to the
                   congressionally imposed budget constraints. Among them was a suggestion to
                   revise subsidy size standards to two persons per bedroom, regardless of sex or age.

                   Early in 2005, the director established and instructed staff to implement stricter
                   subsidy size standards than those defined in the approved administrative plan. 9
                   According to the Authority, managers and counselors considered the director’s
                   instructions to be the policy of the Authority and implemented them as such. The
                   stricter standards were in accordance with HUD guidance and, after only three
                   months of implementation, reduced program costs by as much as $35,875 10 per
                   month. The Authority’s policy allowed it to grant exceptions “for generations,
                   unusual family concerns or medical reasons.”11 Managers trained staff on the
                   stricter standards and counselors began to apply them as they recertified tenant
                   eligibility. Counselors stated tenants quickly became aware of the change and over
                   time reluctantly accepted the stricter standards and smaller voucher assignments.
                   Although the Authority applied the stricter standards for all new tenants entering the
                   program, it did not consistently apply them for existing tenants at recertification.



7
      The director retired on August 1, 2005.
8
      HUD Notice PIH 2005-9 (HA).
9
      The director presented to us these subsidy standards as approved by the board of commissioners.
10
      December 2004 per unit cost of $503.14 less March 2005 per unit cost of $495.68 multiplied by 4,809 tenants
      as of March 2005 ($503.14 - $495.68) x 4,809 = $35,875.
11
      Refer to the Authority’s policy in Appendix B, page 18, section 6.3.


                                                         9
                  In discussing the draft audit report, the Authority disclosed that its board of
                  commissioners had never approved the stricter standards even though the Authority
                  had begun implementing them. Although the Authority had the flexibility to define
                  its own subsidy size standards, these policies must be clearly stated in the
                  administrative plan. HUD regulations required that the board of commissioners
                  formally adopt the administrative plan and any revisions to the plan.12

                  When questioned about the implemented policy, the executive director
                  acknowledged the director was not authorized to make changes without board
                  approval. By implementing policy changes without approval from its board of
                  commissioners, the Authority acted outside its administrative plan and HUD
                  regulations. The Authority should institute controls to ensure the board of
                  commissioners approves any program changes before they are implemented. In
                  addition, if the board of commissioners were to approve them, the Authority could
                  resume its implementation of the stricter standards, which could save approximately
                  $59,000 per month, 13 when it recertifies existing tenants.

     Conclusion



                  The stricter standards the director implemented began to achieve desired results by
                  cutting program costs while allowing exceptions “for generations, unusual family
                  concerns or medical reasons.” The Authority should not have implemented the
                  stricter standards without approval of its board of commissioners. By doing so, the
                  Authority did not comply with HUD requirements or implement its Section 8
                  program in accordance with its administrative plan. To ensure that it effectively and
                  efficiently administers its $29.9 million in annual Section 8 funding, the Authority
                  must establish controls to guarantee that policy changes are approved by its board of
                  commissioners before being implemented. In addition, if the board of
                  commissioners were to approve the implementation of the stricter standards for both
                  new and existing tenants, the Authority could save more than $2.12 million over the
                  next 3.7 years. 14

     Recommendations



                  We recommend that HUD require the Authority to:

                  2A. Institute controls to ensure that the board of commissioners approves any
                      program changes before they are implemented.


12
      24 Code of Federal Regulations 982.54.
13
      Based on tenants as of March 2005.
14
      $59,603 x 12 months x 3.7 years = $2,646,373 - $521,744 from Finding 1 = $2,124,629.


                                                       10
2B. Officially adopt, by board resolution, the stricter subsidy size standards. If the
    Authority applies the stricter standards to existing tenants at recertification, it
    could save as much as $2.12 million over the next 3.7 years.




                                  11
                              SCOPE AND METHODOLOGY

The audit covered the period from January 2003 through March 2005. To accomplish our
objectives, we reviewed federal regulations, the Authority’s administrative plan, and its audited
financial statements. We analyzed data provided by the Authority, reviewed tenant files, and
interviewed Authority and HUD program staff. We performed fieldwork at the Authority’s
administrative offices in Fort Worth, Texas, from May to July 2005.

We obtained computer data files from the Authority that contained Section 8 housing assistance
payments and related information for all tenants during the audit period. We validated the data
in accordance with professional standards. 15 We analyzed the data using ACL software to
identify the audit universe of tenants whose vouchers were too large based on family
composition and the Authority’s subsidy size standards. Because of difficulty obtaining accurate
policy information from the Authority, we revised our methodology during the audit. As
discussed below, our final analysis of the Authority’s data is based on unadjusted housing
assistance payments during the audit period.

We performed comprehensive analytical testing of the Authority’s data to identify potential
overpayments attributed to tenants whose vouchers were larger than allowed under the
Authority’s subsidy size standards. This resulted in identification of unsupported payments of
$180,618 for 125 tenants from January 2003 through March 2005.

We reviewed a representative sample of 34 tenant files to support the data analysis. We used
EZ-Quant software to generate the sample. From the tenant files reviewed, we identified $5,951
in ineligible overpayments resulting from the Authority’s failure to ensure tenant voucher
assignments complied with its subsidy size standards. To illustrate the ongoing impact, we
analyzed the Authority’s March 2005 tenant data to identify possible overpayments for tenants
whose vouchers were larger than allowed. The analysis showed 382 of its 4,809 tenants held
vouchers larger than allowed and the Authority potentially overpaid $11,751 in housing subsidy
for 108 tenants. Over 3.7 years, 16 the Authority could overpay as much as $521,744 ($11,751 x
12 months x 3.7 years) unless it strictly enforces its subsidy size standards.

We used statistical concepts to identify the sample files we reviewed. Therefore, the sample was
expected to be representative of the population. A small representative sample was sufficient to
support the conclusions reached from the data analysis in an objective manner; as such, we did
not project the results of the sample testing to the population. We reduced unsupported amounts
identified through data analysis by the ineligible amounts identified in the tenant file reviews so
as not to duplicate the amounts.



15
     Government Accountability Office, “Assessing the Reliability of Computer-Processed Data,” GAO-03-273G,
     October 2002.
16
     The median length of time in program for those receiving assistance in the 50 largest metropolitan statistical
     areas is 3.7 years. HUD Office of Policy Development and Research (2003).


                                                         12
We evaluated the potential cost savings the Authority might achieve if it obtained board approval
and fully implemented the stricter subsidy size standards discussed in Finding 2. We calculated
the Authority’s unadjusted per unit cost in December 2004 and in March 2005. A comparison of
the two showed the Authority’s implementation of the stricter standards reduced its per unit cost
by $7.46, providing a realized cost savings of $35,875 per month. In addition, as of March 2005,
582 tenants held vouchers larger than allowed by the stricter standards. If it adopts this policy,
the Authority could realize an additional monthly savings of up to $59,603.

We conducted the audit in accordance with generally accepted government auditing standards.




                                               13
                             INTERNAL CONTROLS

Internal control is an integral component of an organization’s management that provides
reasonable assurance that the following objectives are being achieved:

   •   Effectiveness and efficiency of operations;
   •   Reliability of financial reporting; and
   •   Compliance with applicable laws and regulations.

Internal controls relate to management’s plan, methods, and procedures used to meet its mission,
goals, and objectives. Internal controls include the processes and procedures for planning,
organizing, directing, and controlling program operations. They include the systems for
measuring, reporting, and monitoring program performance.



 Relevant Internal Controls


              We determined the following internal controls were relevant to our audit objectives:

              •   Effectiveness and efficiency of program operations - Policies and procedures
                  that management has implemented to reasonably ensure that a program meets
                  its objectives.

              •   Validity and reliability of data - Policies and procedures that management has
                  implemented to reasonably ensure valid and reliable data are obtained.

              •   Compliance with laws and regulations - Policies and procedures that
                  management has implemented to reasonably ensure resources are used
                  consistent with laws and regulations.

              We assessed the relevant controls identified above.

              A significant weakness exists if management controls do not provide reasonable
              assurance that the process for planning, organizing, directing, and controlling
              program operations will meet the organization’s objectives.

 Significant Weaknesses


              Based on our review, we believe the following items are significant weaknesses:




                                               14
•   The Authority does not have adequate internal control processes for ensuring
    its housing counselors assign housing vouchers in accordance with the subsidy
    size standards in its administrative plan (see Finding 1).

•   The Authority does not have adequate internal controls to ensure that the
    board of commissioners approves policy changes before they are implemented
    (see Finding 2).




                                15
                                    APPENDIXES

Appendix A

                 SCHEDULE OF QUESTIONED COSTS




                                                                     Funds to be
           Recommendation                                            Put to Better
               Number            Ineligible 1/ Unsupported 2/           Use 3/

                  1A                   $5,951
                  1B                                    $174,667
                  1C                                                      $521,744
                  2B                                                    $2,124,629
                 Totals                $5,951           $174,667        $2,646,373




1/   Ineligible costs are costs charged to a HUD-financed or HUD-insured program or activity
     that the auditor believes are not allowable by law; contract; or federal, state, or local
     polices or regulations.

2/   Unsupported costs are those costs charged to a HUD-financed or HUD-insured program
     or activity when we cannot determine eligibility at the time of audit. Unsupported costs
     require a decision by HUD program officials. This decision, in addition to obtaining
     supporting documentation, might involve a legal interpretation or clarification of
     departmental policies and procedures.

3/   “Funds to be put to better use” are quantifiable savings that are anticipated to occur if an
     Office of Inspector General (OIG) recommendation is implemented, resulting in reduced
     expenditures at a later time for the activities in question. This includes costs not incurred,
     deobligation of funds, withdrawal of interest, reductions in outlays, avoidance of
     unnecessary expenditures, loans and guarantees not made, and other savings.




                                              16
Appendix B

        AUDITEE COMMENTS AND OIG’S EVALUATION


Ref to OIG Evaluation   Auditee Comments




Comment 1




Comment 2


Comment 3


Comment 4




                         17
18
19
                         OIG Evaluation of Auditee Comments

Comment 1   We thank the Authority for its positive response. Unfortunately, the Authority
            had implemented the incorrect subsidy standards in an effort to reduce costs. As
            discussed in Finding 2, we agree with the Authority's effort to administer its
            Section 8 program efficiently and economically. However, the Authority should
            only implement policy enacted by its board of commissioners. We modified the
            report as necessary.

Comment 2   We reviewed the information the Authority provided and changed the amount we
            reported as ineligible. We omitted the Authority’s attachments because they
            included private tenant information. We separately provided the Authority with
            detailed information regarding all ineligible amounts presented in the report.

Comment 3   We re-evaluated the Authority’s tenant data using the standards the Authority
            provided with its response, which it asserts are those approved by its board of
            commissioners. Our review of sample tenant files during the audit showed the
            Authority did not consistently apply its subsidy size standards for existing tenants
            (see further information in Finding 1). We modified the report to reflect those
            changes.

Comment 4   As discussed in Finding 1, the audit identified specific instances where the
            Authority did not correctly apply its subsidy size standards for existing tenants.
            We discussed the details of the individual cases with Authority staff throughout
            the audit.




                                             20