Quality-control program questions and answers

+----------------------------------------+-----------------------------------+
| U.S. DEPARTMENT OF LABOR | CLASSIFICATION |
| Employment and Training Administration | |
| Washington, D. C. 20213 | UI |
|
+-----------------------------------+
| | CORRESPONDENCE SYMBOL |
| | |
| | TEUQI |
|
+-----------------------------------+
| | ISSUE DATE |
| | |
| | August 14, 1986 |
+----------------------------------------+-----------------------------------+
| RESCISSIONS | EXPIRATION DATE |
| | |
| None | August 31, 1987 |
+----------------------------------------+-----------------------------------+

+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
| DIRECTIVE | : | UNEMPLOYMENT INSURANCE PROGRAM LETTER NO. 52-86 |
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
|
|
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
| TO | : | ALL STATE EMPLOYMENT SECURITY AGENCIES |
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
|
|
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
| FROM | : | DONALD J. KULICK |
| | | Administrator |
| | | for Regional Management |
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
|
|
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+
| SUBJECT | : | Answers to Questions and Comments Raised Concerning the Unemployment Insurance Quality Control (UI/QC) Program. |
+-------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------+

1. Purpose. To transmit answers to questions raised concerning the UI/QC system.

2. Reference. UIPL 19-84, UIPL 3-85, UIPL 21-85, ET Handbook No. 395.

3. Background. The VI/QC system was implemented voluntarily on March 30,1986. Many questions and concerns have been raised regarding ET Handbook No. 395, analysis and corrective action, and the ADP software. The majority of the questions and comments came from the States and Regions when they reviewed the draft handbook.

   In keeping with our commitment to keep SESAs informed and encourage uniform QC methodology, the answers to the questions and concerns raised by SESAs and others are being transmitted to the SESAs for their information.

   The questions, concerns, and answers are grouped by subject matter for easy reference.

4. Action Required. SESAs are requested to share this information with appropriate staff and other interested parties.

5. Inquiries. Direct questions to your appropriate Regional Office.

6. Attachment. Copy of questions and answers.

Attachment: questions and answers (https://oui.doleta.gov/dmstree/uipl/uipl86/uipl_5286a.pdf)

Attachment to UIPL No. 52-86

A. Policy and General Questions

1. Q. Proposed QC funding is inadequate--insufficient to fund extraction of additional data
   elements, or for supervision given larger samples, or to recognize State differences in
   interstate requests and travel or fund differences in staff hours worked.

   A. The Department believes that funding is adequate, overall, for QC. Regarding the specific
   issues raised, it should be noted that:

              Writing data extraction routines is a one-time effort.      It should not involve
               continuing use of ADP staff time.
              The high degree of automation should make the QC supervisor much more productive
               than his RA counterpart. Case management and most administrative functions are
               computerized, whereas they were manual in RA. In addition, in an average State,
               approximately 30% of the data elements can be downloaded, leaving only about 70
               elements to be entered manually versus 88 in the RA program.
              As far as taking full account of travel need s and State differences in work hours per
               week, interstate requests, etc., we are committed to monitoring the situation closely
               after QC becomes operational to ensure that the full costs of the system are met.
              Another effort being conducted by the Department is to work with selected States to
               develop less costly da ta collection alternatives. This is an effort to find other ways to
               collect the necessary data without sacrificing data integrity or utility.
   

2. Q. Concern has been raised that the National Office has not taken the lead in analyzing all the
   Random Audit data collected over the past few Years. Are actions contemplated to address
   these concerns?

   A. Under the RA program the emphasis was placed on the col lection of data. The RA data also
   had a number of limitations in terms of analysis. Nevertheless, the National Office has
   supplied SESAs with examples of how RA da ta can be analyzed and have encouraged them to
   use RA data for this purpose to the extent practical and feasible in each State.

3. Q. What has been done by States to f ix the problems found by the Random Audit program?

   A. As a result of the RA findings many States have taken a number of actions to tighten
   administrative procedures and reduce improper payments. Examples include:

              Procedures modification and forms revision to clarify requirements:
              revision of benefit rights interview to clearly communicate the individual UI eligibility
               responsibilities;
              increased inter-agency coordination to more closely monitor claimant "work search"
               activity and
              revised agency policies regarding acceptable "work search" efforts by individuals.
   
                                                  1
   
     The Department strongly encourages States to initiate appropriate corrective action in
     response to problems identified through RA; final responsibility for such action rests with the
     States. The same policy will apply under QC.
   

4. Q. Excessively rigid Federal requirements make QC unworkable.

   A. The Department has attempted to establish the m1n1mum of procedural restrictions
   necessary to ensure data Quality, while allowing States flexibility to modify the program for
   their own internal needs. The data system allows States to add State-use digits to further
   identify problems and to add up to 60 State data items to those Federally specified.

5. Q. Concern has been raised with the direction the program appears to be taking. That is an over
   concern with the numbers themselves rather than basic program integrity. An example of
   this is that in States with stringent work search requirements. QC will thoroughly review and
   verify the work search and most likely find some degree of inadequate work search and
   hence. improper payments. In States with no work search requirement, there will be no QC
   review and therefore no “errors" are possible in this area. This philosophy places States in
   the uncomfortable position of potentially sacrificing basic program integrity in an effort to
   reduce so called errors, i.e., if States do not have a policy to cover a situation they cannot be
   faulted for not carrying out that policy to the letter. What is the National Office’s response
   to this situation?

   A. Because UI operates in each State based on the State’s own law and policy, we have no choice
   but have each State assess the accuracy of its administration of its State laws. Employers and
   other interested par ties of each State will have to assess trade-offs of stringent rules/complex
   laws with the cost and accuracy associated with administering those rules and laws.

6. Q. Don’t establish standards and don’t make any State-to-State comparisons [ in data] due to
   differences in State laws.

   A. The following, received as a consensus recommendation from Washington based
   representatives of SESAs (ICESA), organized labor, and the business community, is now ETA
   policy:

     “States would be required to release the results of the QC program at the same time each
     year, providing calendar year results using a standardized format to present the data. States
     would have the opportunity to release this information prior to any release a t the national
     level. “
   
     The procedures to implement this policy are under development now by the Department.
     As stated early in the QC design. the Department has no intention of using QC data to
     establish across-the- board performance standards or sanctions.
   

7. Q. Congress has enacted legislation requiring fiscal sanctions against States not meeting specified
   error-rate levels in Welfare and Food Stamp programs. Does the Department of Labor favor
   similar legislation to reduce erroneous payments in the UI program?

   A. The Department does not favor Federal legislation to reduce erroneous UI payments because

                                               2
   
       it does not believe such legislation is needed. States set benefit standards and impose taxes
       to finance them and thus have ad equate incentives to implement corrective act ions to
       reduce errors. The Department’s objectives in establishing a UI QC system are to provide a
       diagnostic tool for State use in identifying and correcting errors in benefit payment and tax
       collections to improve program quality and integrity, and to provide an accurate and reliable
       information source to guide State and Federal policymaking and legislation.
   

8. Q. Does the Department of Labor plan to provide financial incentives to States which improve
   their productivity and reduce errors?

   A. No. As part of the consensus agreement referenced in question No. 6 above, it was
   recommended that no funding incentives be used.

       A strong incentive to the States is built into the system: by identifying problems and
       correcting them, leakages from the system will be reduced, benefits will be paid in a timely
       and proper fashion to those who meet the criteria, and taxes will be paid by employers when
       due.
   

9. Q. State privacy act provisions prohibit the disclosure of selected data elements proposed for
   inclusion in the QC data base.

   A. The Department has decided that the best approach is to eliminate the social security
   number. With the elimination of the social security number, it is assumed that State privacy
   provisions no longer are a problem. The Federal government will only have a State encrypted
   number to identify a case.

10 Q. Use of the Social Security Number ( SSN) identifier , particularly "unscrambled", undermines security of certain data that should remain confidential.

A. See response above.

11. Q. UIS has not provided adequate opportunity to review and comment on regulations.

    A. Proposed regulations are to be published for comment prior to final promulgation. ETA has
    tried to provide as many additional avenues of review and comment as possible--including
    publishing the Advance Notice of Proposed Rulemaking and circulating drafts of ET Hand
    book No. 395. These vehicles gave States and other interested parties an opportunity to
    express their views on QC policy issues and also gave ETA a good sense of the substance of
    much State thinking regarding QC and influenced the way the regulations have been drafted.

B. Investigative Procedures

1. Q. Are States required to verify all claimant work search with employers? Often if no
   applications are taken, no verifiable documentation exists.

   A. All work search contacts for the key week are required to be verified. QC coding provides a
   field for "number of work search contacts for key week that could not be verified as either
   acceptable or unacceptable".

                                                3
   

2. Q. If a claimant tells you that he only telephoned the employers as a contact for the key week,
   can the QC investigator verify this by phone or must he still go down to the company in
   person to obtain a signed statement from the employer showing that the claimant made a
   phone inquiry for work?

   A. All work search contacts must be verified in person. Even if telephone contacts are acceptable
   work search methods per State law and policy, the QC investigator must still verify the contact
   in per son. Data collection alternatives will be conducted to explore the possibility of
   modification of “in-person interview" methodology.

3. Q. Can States use telephone information (it's acceptable for hearings) in place of in-person inter
   views? Suggest companies previously contacted in-person with no discrepancies could be
   called for separation information and wage report could suffice for wage verification.

   A. Current QC methodology calls for all information including claimant and employer statements
   to be obtained in person. The consensus pa per from Washington-based public interest
   groups (prepared by ICESA) and the QC Round table both recommended that testing of
   alternative methods of collecting the information be conducted.

4. Q. Question the need for in-person claimant interview in all cases. It may be more efficient to do
   by mail and may not lose accuracy. The QC Supervisor should have discretion to waive in-
   person interview requirement.

   A. See response to question No. 3 above.

5. Q. Do earnings for all weeks need to be investigated or only those in the Key Week?

   A. Remuneration/earnings must be investigated for those weeks which could, based on SESA
   written law and policy, directly affect the Key Week. The requirement relating to
   remuneration has been reworded to make it consistent with the policy that investigations will
   be limited to what is “doable” within funding levels. The current statement of the
   methodology is:

     “Verifications must be made of receipt of all remuneration which could directly affect the Key
     Week for which claimants could be disqualified or have benefits reduced. “
   

6. Q. Is it necessary to have all official act ions for issues outside the Key Week completed before
   the supervisor can sign off on the case as complete?

   A. No. QC methodology calls for QC investigation of only Key Week related issues. "A case is
   complete when the investigation has been concluded as required. All official actions for the
   Key Week (except appeals) have been completed, the supervisor has signed off and the
   results have been entered into the computer." (See Chapter VI of ET Hand book No.395.)
   Non-Key Week issues should be referred to other SESA staff for pursuit and resolution unless
   adjudication by QC staff would only involve incidental time and resources; these need not be
   concluded in order for a case to be considered complete.

                                              4
   

7. Q. Regarding nonmonetary re-investigation. Do not see need to re-investigate every nonmon
   already issued.

   A. The QC investigative methodology does not call for reinvestigating every nonmon. Instead,
   investigators must take "complete fact finding on al l issues that have not been detected
   previously or for which it is questionable that they were properly treated".

8. Q. Question the requirement that “all base period employers must be contacted regard less of
   whether or not such employment was used in the monetary. “ -State doesn’t feel need to
   audit/investigate employers not involved in calculation of monetary-not required in RA.

   A. This method ology has been modified to now read "all prior employers with whom
   employment could affect the claim".

9. Q. Is it necessary to pursue “moot” issues, i.e. claimant is ineligible for a few weeks due to
   monetary ineligibility and for one or more of those weeks claimant is not A&A? State policy
   may say that the A&A issue is “moot”?

   A. In the example given, where a claimant is monetarily ineligible and also not able and available
   (A&A), the A&A issue need not be pursued since it could not affect the Key Week. (A Key
   Week is a paid week.) An issue is pursued by the QC unit only when there is a potential that
   it could affect the Key Week. Any issue that could affect the Key Week is to be pursued.

10. Q. Doesn't the QC requirement regarding verifying "referrals by union halls" refer only to
    referral unions?

    A. Yes. In order for a union to be considered as a work search contac.t, most States require that
    the union have a hiring hall.

11. Q. Regarding the part of the QC methodology that states “upon completion of interviews with all
    base period employers, the investigator must deter mine what is the proper program code for
    the claim.” If after investigation program code changed, should correct code be entered?

    A. Yes. This data element does not have a "before investigation" and "after investigation" entry.
    Therefore, entry must be checked for accuracy when the relevant part of the investigation is
    finished.

12. Q. Must the QC investigator attend all- appeal hearings as stated in the QC methodology? Why
    allow explanation of absence then? Also, QC Analyst may have no direct evidence to provide
    to the referee.

    A. QC investigators are required to attend appeal hearings in order to explain how the QC
    investigation was conducted and present evidence to explain why the claimant’s eligibility or
    benefit amount changed. The explanation sentence is to allow for a rare exception situation
    where the investigator cannot be present.

13. Q. The completion time lapse objective appears unrealistic due to increased workload, data
    elements, new codes and computer equipment. Also, what will happen if SESA doesn't have

                                                5
    
      good reason for untimely investigations and/or data entry?
    

    A. The time lapse standard was based upon RA experiences in sampled States with consideration
    for the differences between RA and QC. Based on preliminary State experience, it appears
    that the QC standards are realistic; however, a review will be made after one year to
    determine if modifications to the standards are necessary.

14. Q. The proposed QC manual states that 98\ of cases pulled must be completed within 90 days.
    Concern exists that the criteria of 98\ of cases being completed within 90 days of sample
    selection can never be met since the number of cases pulled each week is small. For example,
    if a state pulls 14 cases per week and completes 13 of them. they will have a 92.8\ completion
    rate. The criteria might as well be 100\ completion within 90 days of sample selection.

    A. The criteria will not be applied to a single weekly sample. The case completion criteria is an
    annual measure covering all cases drawn in a year. If the annual sample size is 600 cases. 12
    cases could go over the 90 day time standard and the SESA would still meet the 98\
    completion criteria for the year. It will be possible for SESAs to track performance against the
    criteria during the year using the spreadsheet function.

15. Q. Can a State eliminate Items P and G from Summary of Investigation (Item P is "form sent to LO
    Mgr. on case disposition": Item G is Dependents).

    A. Yes. QC methodology calls for a Summary of Investigation to be included in the case file.
    There are no specifications concerning what the summary must contain. A format was offered
    as a sample. States may develop their own format as long as it meets the needs of the
    program.

16. Q. Is it possible to have the investigator’s explanation of why an in-person interview was not
    conducted only once; not in both questionnaire and in summary?

    A. No. It should occur in both places. Everything in the summary is located elsewhere. Its
    purpose is to organize and summarize.

17. Q. The QC design does not appear to recognize the impact of interstate requests and state
    differences in policy and travel requirements.

    A. QC interstate requests will be monitored closely to obtain information in this area. If changes
    to design are merited based on operational experience, they will be made.

18. Q. There is the concern that interstate requests via terminals will need hard copy documentation
    when adverse action likely.

    A. This is possible, in which case hard copy documentation will have to be provided by mail.

19 Q. Can the Systems Manager be outside the QC unit or must the QC Supervisor be the Systems Manager? References are made throughout the hand book that only the QC Supervisor can perform certain ADP functions.

6

A. This is a State decision. However, the QC software was designed under the assumption that the QC supervisor would be the System Manager. The System Manager is the only account that is allowed access to all functions. The System Manager, for example, is the only one allowed to establish user accounts. The individual users then are responsible for establishing their own password for the account.

C. Error Classification

1. Q. Regarding Error Classification, if two errors occur in a case, both of which result in a total OP,
   when the two OP amounts are totaled can the results be grossly overstated?

   A. Because this is possible, this concern will be addressed by a workgroup during the development
   of the forma t for State release of QC data. It may be appropriate to compile data one way for
   State use in identifying problem areas and another way for public release.

2. Q. How do you code a case if another SESA unit refuses to issue an official agency action based on
   QC findings?

   A. The QC methodology states that official action for Key Week findings must occur as a result of
   the QC investigation in order for the case to be complete. If official agency action which would
   be appropriate when State law and policy are applied to QC findings is not taken, the case is
   incomplete. There is no code for such circumstances.

3. Q. How do you code fraud cases if the fraud unit has a policy not to pursue cases of under a
   certain dollar amount or under a certain number of weeks?
   A. If there is an official written agency policy that declines the pursuit of fraud in certain
   circumstances then that policy would be followed. If there is no such written policy, a fraud
   determination must be written to complete the case.

4. Q. Regarding Prior Employer Action codes and definitions, should the wording be changed to
   cover rebuttals?

   A. No, this is not necessary. The key word that is used in the definition is "determination". Within
   UI, the term determination includes the total process: fact-finding, rebuttal, adjudication and
   documentation. Not all determinations have or need rebuttal or written adjudication for QC
   purposes, but "determination" covers these processes when they must be pursued. A separate
   definition is not necessary.

D. Claimant Questionnaire

1. Q. Reword question i37 to stand alone with respect to shift work claimant is willing to accept and
   delete any reference to question i23 that addresses the claimant' s prior shift work.

   A. Question i32 now limits the claimant’s response to shifts that they are willing to accept.
   Question il7 provides for identification of hours or shift for the most recent job. Both questions
   are used in determining claimant's work search in relation to work history to determine if they
   meet the continuing eligibility requirements.

                                                7
   

E. Data Elements-Record Type One

1. Q. (B-1)-Some interviews involve both in-person and telephone interviews. How should multiple
   methods be coded ?

   A. When more than one method is used, code the method by which the majority of the
   information is obtained.

2. Q. (B-3)-Education level is not necessary for QC and others outside the State.

   A. This element provides information necessary for the investigator's assessment of the
   claimant’s "reasonable and prudent" work search efforts. It is an element which was
   provided under RA.

3. Q. ( B-4)-Voc/tech training is not necessary for QC and others outside the State.

   A. This element also provides information necessary for the investigator's assessment of the
   claimant’s "reasonable and prudent" work search efforts. Although it is a new element, it is
   similar to education level which was reported for RA.

4. Q. ( B-4)-Is verification still needed if training has nothing to do with past or seeking work?
   Verification is not required of college graduates.

   A. The investigator should compare the claimant's response with other available SESA records.
   If a significant difference between the claimant's response and the records occurs, the
   investigator should resolve this with the claimant. Third par ty verification with a technical
   school or college may be necessary in those rare instances when the issue cannot be resolved
   with the claimant.

5. Q. (B-6,B-7 & B-9)-Why are all these occupation codes needed?

   A. These codes assist the investigator in assessing the claimant’s employment history and its
   relation to the work search.

6. Q. (B-7)-There needs to be a standardized definition of usual occupation.

   A. The investigator should compare the claimant’s response on the questionnaire with other
   available SESA records and information obtained from employers. If a significant difference
   between the claimant’s response and the records occurs, the investigator should resolve this
   with the claimant prior to data entry.

7. Q. C B-8)-Normal Hourly Wage- What kind of verification expected? From claimant
   Questionnaire or ES records?

   A. The investigator should compare the claimant’s response on the questionnaire with other
   available records i.e., wage information, ES records. If a significant difference between the
   claimant’s response and the records occurs, the investigator should resolve this with the
   claimant prior to data entry.

                                                8
   

8. Q. C B-11, B-12 and B-13)-Sex, date of birth, ethnic are not relevant to QC.

   A. On the surface these items do not appear to be relevant to QC but, these items are needed to
   validate the representativeness of the sample selected and were collected under RA for the
   same purpose.

9. Q. (C-7)-What code applies when no ERP is given?

   A. If the claimant is a potential ERP candidate and no ERP was given, this item should be coded
   0. If an ERP was waived for the claimant, the appropriate code is "x" (not applicable) .

10. Q. (C C-8 and C-9)-Nonseparations prior to the investigation are not relevant to QC.
    A. Information regarding prior nonseparations indicates the effectiveness of the agency' &
    normal detection procedures: therefore, it is an important secondary indicator for agency
    purposes.

11. Q. ( D-1 and D-6)-The definition calls for coding only last employer; where should another
    separation affecting the key week be coded?

    A. The investigation will ascertain the reason for al l relevant separations. However only the
    most recent separation (State definition) will be coded as an indication of the type of
    separation which has caused the spell of unemployment. Other separation information will
    be part of the investigation file. If the State desires additional separation information entered
    into the system for analytic purposes, it may elect to use the Option section of the menu
    software.

12. Q. (D-4)-The tax rate of the last employer has no bearing on entitlement. It is based on State
    law and should not be part of national data base.

    A. In the majority of cases the last employer is also the separating and chargeable employer. It
    will be helpful when formulating corrective action to deter mine if a correlation between
    separation errors and maximum tax rates exist.

13. Q. (D-5)-0ur State system provides a 2-digit SIC and relates to the major base period employer
    not last employer.

    A. Under these circumstances. a State should "download'' the first two digits followed by two
    zeroes. For example, if the record shows employer code of 17, then it should be
    "downloaded" as code 1700.

14. Q. (E-2)-Does "all wages" just mean the wages defined by State law?

    A. Yes.

15. Q. (E-4 and E-12)-High quarter wages are not needed in this state and therefore should not be
    required.

                                                9
    

    A. This item provides useful information about the claimant’s work pattern. Little additional
    burden is anticipated in those States that maintain quarterly wage data even if high quarter
    per se is not used to determine monetary eligibility.
    This item is not applicable for States with other than quarterly wage data.

16. Q. (E-5-Weeks Worked)-Does this come strictly from the monetary? Sometimes will be over a
    hundred due to several simultaneous part-time employers.

    A. This should represent the number of weeks worked under the State's definition. For example,
    if 20 hours constitutes a week of work under State .law and the monetary shows 100 weeks
    for a 52 week period, enter the number of weeks in that period which met the State' s 20-
    hour definition.

17. Q. ( E-8)-Is it before investigation (as in title) or before Key Week (as in definition)?

    A. All items entitled "before investigation" mean up to and including the Key Week.

18. Q. (E-9)-Is BYE considered in calculating the Remaining Balance?

    A. No. The purpose of this item is to deter mine how much money remains after the Key Week
    ending date regardless of the BYE. If the key week is the last week of the benefit year, code
    the amount that would have been left after the key week payment had the benefit year-end
    not intervened. Use code 00(z) only if key week payment was last payment due to
    exhaustion of benefits.

19 Q. ( F-5)-Should reduction under Section 404 of Pennsylvania Unemployment Compensation Law be coded?

A. The reduction mandated legislatively should be included in the amount paid plus offset coding element. The reduction is analogous to an over payment offset in that the combination of the amount paid to the claimant plus the offset is used to reduce the claimants balance by such total. Do not code this information in element 1''-9, "Other Income Deductions for Key week Before Investigation", as the issue does not relate to income deductions at all.

20. Q. (F-6 thru F-13)-We need more clarification of income and earnings.

    A. F-6. F-7, F-10 and F-11 all relate to wages received for services during the claim series. These
    wages are usually from a new employer and are for intermittent or casual employment. F-8,
    F-9, F-12 and F-13 pertain to all other types of income even if the State law defines separation
    pay, etc. as wages.

21. Q. (G-6 and G-7)-Registration/referrals with private employment agency should not be part of
    national data base.

    A. All efforts to regain employment should be investigated and coded for State analytic
    purposes. Registration with a private employment agency was collected and reported under
    RA. The number of referrals was an item requested at a Data Collection meeting of State

                                                 10
    
       administrators because it provides valuable information to States on what is occurring outside
       the agency with claimants and may lead to a policy change in how claimants’ work search is
       pursued.
    

22. Q. (G-9)-Definition says number of union referrals indicated by claimant, but matrix says from all
    sources.

    A. The wording of this element in Chapter IV of the Hand book has been changed to be
    consistent with the matrix. Therefore follow the matrix. The investigator should compare the
    claimant's response on the questionnaire with other available records. If a significant
    difference between the claimant’s response and the records occurs, the investigator should
    resolve this with the claimant prior to data entry.

23. Q. The State should be able to differentiate among claimant, employee or agency error.

    A. A. That is the purpose of H-3 in the error classification section.

F. Record Types Two and Three

A number of questions were raised about record types two and three prior to and during the Secretary’s review of the QC program. However, with the decision to make the longitudinal files optional for QC, the questions become irrelevant. Record type two (wage activity) is optional while only parts of record type three (claim activities) are optional since it is also needed to pull the sample of Key Weeks for investigation. See page III-2 of the State Operations Hand book for any additional information on record type three. If any issues are still unanswered in this area, they should be addressed through the Regional Office.

G. Analysis and Corrective Action

1. Q. ETA needs to fund more analysis of RA data and provide States with technical assistance on da
   ta analysis so that existing C and new) data can be used properly.

   A. The Department has recently allocated funds to each State for an analyst position.
   Additionally. UIPL 29-86 presents guidelines and procedures to be followed in developing and
   designing proposals for special studies related to the UI QC program. After approval of the
   proposed special study. States would be allowed to redirect staff for these purposes.

2. Q. Once the State QC staffs have learned the basics of the program and begin collecting data, will
   the National and Regional Off ices devote themselves to assisting States in information analysis
   and take the lead to show States how the data can become a useful management tool?

   A. During CY 1986 the National Office will be providing a technical assistance guide on statistical
   analysis for data analysis and corrective action. The publication of this material will be
   accompanied by training. Beyond that, active technical assistance will be provided in the
   development of management reports.

                                                11
   

3. Q. Corrective action will be developed to address the problems identified by QC. Have funding
   sources to pilot these plans been identified?

   A. Many corrective actions can be undertaken at little or no cost to States. Several States
   accomplished reductions to their error rates based on RA data without special resources.
   Beyond the limited-cost corrective actions, the Department will consider seeking funds in
   subsequent years based on information produced by the QC program. In addition, the
   automation fund is available for funding needed requests and QC findings would be a sound
   rationale for submitting a proposal.

H . Organization and Authority

1. Q. QC in general, and in particular its organizational placement, should not be subject to Federal
   compliance monitoring.
   A. To ensure that data are useful for nationwide analysis, the conduct of QC needs to be subject
   to continuous scrutiny; otherwise, data quality could suffer and entire quarters of SESA data
   would have to be thrown out (as happened with RA). Similarly, an organizational setting for QC
   too close to off ices responsible for UI program management could lead to conflicts of interest
   and compromising QC 1 s role as an unbiased source of program assessment.

2. Q. If a QC staffer is also a union official engaging in union activity, will this conflict with QC 1 s
   regulation that funds be used exclusively for QC purposes and put the State out of compliance?

   A. The language in the proposed draft regulations which have not been published addressed the
   use of QC funds exclusively for QC activities. The Quality Control State Operations Hand book
   (ET Hand book No. 395) states “All staff positions and resources allocated to QC must be
   utilized exclusively for QC purposes." Its intent was to ensure that QC work is completed in a
   timely and effective manner within the full funding provided for this activity. It was not
   intended to constrain States’ personnel choices. Thus, for a QC staffer to be a member (or
   official) of a union does not by itself cause a problem. The issue, if any, would be whether or
   not the State were performing mandated QC activities and staying within budget. It would be
   inappropriate, however, for QC resources to be used for the performance of that persons union
   activities (i.e. travel).

3. Q. States are required to develop written QC procedures. If a State feels ET Hand book 395 is
   sufficient, can it be used as the State procedures?

   A. States may use the recommended procedures issued in the companion Technical Assistance
   Guide as part of their written QC procedures, but the procedures must be tailored to
   accommodate particular SESA laws and processes. See UIPL 26-86 attachment 2 for a checklist
   of areas where particular tailoring is expected.

4. Q. Hand book No. 395 requires each State to develop written procedures for QC. Who should
   write these procedures and how extensive should they be?

   A. States may assign the task of writing the procedures to whomever they wish. Most States will
   probably assign this to either the QC supervisor or staff who normally write procedures. The

                                              12
   
     procedures may be patterned after the recommended procedures sect ion of the Technical
     Assistance Guide, or they may be mad e more or less extensive, depending on the needs of the
     State. The procedures are necessary to document the way the QC program will operate in
     the State so that the QC methodology is tailored to particular State law and policy.
   

I. Data Processing

1. Q. Is a random number table available for use in the COBOL Program?

   A. The National Office will supply each year a printed table of random numbers which provides
   unique number s for each week in a given calendar year. SESAS must use the number in the
   control record input to the COBOL.

2. Q. The State cannot provide the month of birth.

   A. If month of birth is not captured by the normal SESA system, the State should code the month
   of birth as 0 6 (June) as the default.

3. Q. Assumption is incorrect that data elements are available and accurate for downloading
   process.
   A. The Record #3 format is almost identical to the RA format (17 of 27 elements), which allows
   for sampling to occur, these should be available. The remaining should be completed if
   available (information supplied by States in response to a survey revealed most are available).

      For the Record #1 downloading. SESAS should follow the instructions for formatting in the
      Hand book Chapters I, II and IV, completing as many fields as possible. Some fields cannot be
      downloaded and the instructions indicate these must be left blank. If downloaded data in
      Record #1 is wrong. the QC unit will make corrections manually as data is entered from case
      investigations.
   

4. Q. Should there be a delete routine if. once sampled, we discover that an individual should not
   have been selected?

   A. Yes. The National Office will establish a routine for the QC supervisor to notify both the State
   DP unit and the National Office if this occurs. The State DP unit would have to delete that
   individual from the "Cumulative File" (sampled SSN's) that is maintained on the mainframe.
   The national office would correct any files passed to the VAX system.

5. Q. How much mainframe data must be aggregated? How many files and what are volumes? This
   affects costs.

   A. Mainframe files and storage is minimal. Each week a file of initial claims and weeks claimed
   must be extracted for all transactions in the week. This file, similar to the RA Population file is
   used to create a single record for statistical counts, array and select the weekly sample, and
   aggregate for statistical proportions tests. These functions are all handled by the COBOL
   program provided by the National Off ice. The SESA DP unit must save the weekly transaction

                                                13
   
     files (cumulate on tape) for a period of four months to allow backup should sampling prove to
     be not representative of this universe.
   
     Other than extracting data temporarily held prior to passing to the PR0-380, the only other
     file maintained on the mainframe is a cumulative file in Record 13 format (120 bytes each) of
     those weeks selected in the sample.     This small volume files (8-16 new records per week).
     There are no sophisticated update or maintenance routines required. Volume will depend on
     SESA UI workloads.
   

6. Q. What is the ADP manual referred to, we have not received it? Everything is marked “draft’.

   A. The ADP manual referred to in the QC Hand book is the manual for using the PR0-380.
   Documentation for SESA DP staff includes Chapters III and IV in the QC Hand book plus
   additional documentation sent with the COBOL program from the National Off ice. All
   documentation is marked draft until all official approvals for QC have been obtained.

7. Q. Adjustment indicator (Record 13) needs a better explanation. Why does Record 12 have no
   adjustment indicator?

   A. The Record 12 format has been revised and now contains an adjustment indicator; however,
   creation and use of this record is completely optional for each SESA.

     These indicators will be used the same way in both record formats. It indicates, by coding as
     a “2”, that the record is the first transaction for a particular week, claim, or wage item for a
     particular SSN. If a second transaction (e.g. a monetary redetermination or wage adjustment)
     occurs, the indicator is coded. "l" to indicate that claim or employer wage item had been
     reported earlier and now is being "adjusted”. This eliminates duplicate statistical counts and
     allows the adjusted data to be kept in the longitudinal file to show changes.
   

8. Q. QC calls for the sample to be pulled no later than each Monday. The timing of transactions in
   the SESA does not allow this to occur until Wednesday at the earliest.

   A. The QC investigation is hindered in that more time has elapsed since the week was claimed
   possibly making some verification harder to obtain. SESAs should make all possible attempts
   to run the process on Monday. In addition, the QC software is identical for all States and
   uniformity is required wherever possible.

9. Q. Is there more of an explanation of security regarding the QC micro computer system? And
   how is the security of the SESA mainframe effected?

   A. The short explanation in the Hand book deals with physical security of the PR0-380 computer
   system. When talking of a small, easily movable piece of equipment such as this. physical
   security is very important. The computer should be in an office that can be locked when
   someone is not at the console terminal.

     It is suggested that the States establish a password requirement for each logon ID, and the QC
     supervisor should routinely check that passwords are required for all IDs. Additionally, the QC
     supervisor should have their staff frequently change their password s. These steps should
   
                                              14
   

prevent any unauthorized logon to the microcomputer. Even after an allowed logon, within the QC software, there are permissions required for access to certain data or functions. Users are asked, for instance, to supply a system password for access to certain menu functions within the system utilities.

Any phone line into a microcomputer also presents a possible security problem. QC supervisors should be alert to all users signed onto the system from telephone ports and verify missions.

15