oversight

More Reliable Data Needed as a Basis for Providing Federal Assistance to Economically Distressed Areas

Published by the Government Accountability Office on 1971-05-10.

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

                           llllllllllllllllllllllllllllllllllllll
                                  LM095661




Department   of Commerce
Department   of Labor
U- 133182




To the           President           of the Senate     and the
Speaker            of the      liollse     of Representatives


            This  report               presents     our        findings         on more        reliable    data
needed        as a basis               for providjng            Fcdcral         assistance          to econom-
ically      distressed               areas,         The      report        dea.ls    with    the     inadequacy
of statistical               data      furnished          by the        Departments           of Commerce
and      Labcir       that     are      used       in qualifying            redevelopment             areas        for
assistance     under  the Public     Works                              and Economic               Development
Act    of 1965, c?s amended      (42 U.S.C.                              3123).


            Our       review         was made    pursuant                  to the Dudget   and Account-
ing Act,            1921 (31         U.S.C. 53), and the                  Accountillg   and Auditing    Act
of 1950           (31 U.S.C.          67).


       Copies                of this      report       are     being     sent        to the Director,           Office
of Managernc,nt                 and      Budget;       the     Secretary             of Commerce;             and the
Secretary            of Labor.




                                                                      Co2nptroller          General




          ----               50TH        ANNiVERSARY               1921- 1971 ----                             -
-DIGEST
   _ - - -- --




       The Economic Devc:iop:iient Administration,              Departwnt     of Commerce, pro-
         ides financial     o. tccl-inical 9 and planning       assistance    to aid long-range
       economic development           cf distressed     areas.    As authorized      by 12~:~ the
       S~~~~di~j~-o~~'Comii!c-rcc designates         areas eligible       for a;sistanc,e   on the
       bi!sir, of statistical         dat a developed    by the Departments        of Labor and the
       Interior   and t?ik-%?'$u"'i,f         the Census.      (See p. 8.)

       For the most part,           i.he Economic Development           Administration's       akJility   to
       identify       properly    areas eligible        for assistance       hinges on the soundness
       of unw~plo,ym~nt 2nd -incow data.                  The desigtl,;tions     of cconcm-ic distress
       may in<'iu:=nre       thtl distrikt!on         of monies and benefits          from other federal
       agencies.          For cxanlple, -i'ikms located        in arcas of high unei;:ployment          are
       eljgitjle      for   Federal pror;urement        preference.        (See p. 10.)      There-rorc
       the Cenwn 1 Accounting              Office    (GAO) made a revsw! of the c!irrentncss              and
       accuracy of the st&istical                 data used for determin-iny          sn ar*ca's eligib-il-
       ity for assistance.
           ~__. _.-..-..Y'_




       Unanploywnt       2nd income data used by the Econc\r;;ic Pevelop;ncnt: Adminis'ira-
       tion in de1 nw1in-i ng the eligibility           of local areas ax not current         and
       arc of qucstionat~lc         accuracy.     The data should he -iiT!F:!movedto C;ISU~C re-
       alistic   eco;iwic     artpraisals     of those areas.      GAO qusstivns   :thether   the
       areas of th2 United States experqerxing               economic distress,    as indicattld
       by high uxmployn~ent          and/or 70~ irxonle levels,      are, in all instances:
       being proycr'is/    ider;tified.       (See p. 14.)     GAO did not a-ttmpt      to cvalu-
       ate the ailpt-opriateness         of' uncniployment   and incc,l;le levels as criteria
       for eligibili,Ly.



       The questionable     reliability      of the unrm:,lo~wnt     data is attk-ibut;iblc
       to conceptu;-1   ~wai:n~~sses -in the methodology        fcr estimating   tiP~ill~l~l>'tl,cnt
   -   as ~11 as to problem            in deve!oping   unmploylr;ent   rates for small r:rrc‘:!s,
       (See p. 14.)     This repel-t discusses       these weaknesses which raise
    considerable               doubt       as to the Gccuracy                   and reliability      of the unemploy-
    ment      estfinates            imde    for   sm1   1   drld       rural      arcas.       (See pp. 18 to 40.)

    The many problems associated          with the development         of current     and reli-
    able statistical        data are not subject      to ready solution.          This is espe-
    cially     true of small areas --characteristically          redevelopment        areas--
    where statistical        data normally    are not gathc;-cd on a continuing            basis
    and p/here the costs of developing           meaningful   statisticE?       data are sig-
    nificnntly      higher thsn those connected with developing              data en a national
    or regional      basis.     (See p. 10.)

    Within   reccn!: years the Department      of Labor has initiated      and sponsored
    studies   designed to produce information          which could be used to improve
    the methodology     established    for estimating      unemployment  in State and lo-
    cal areas.      Except far a modification       with respect    to one major element
    of th.is methodology,      however, the studies      have not resulted   in an im-
    proved methodology,      and it remains basically        the same as that introduced
    in 1960.

    GAO evaluated    the unemployment-estimating        practices     in two States.        In
    both States the prescribed        methodoloqy    was subjected      to varied degrees of
    modiPicat,ion   and was not applied      uniformly.     State agencies are severely
    handicapped   in their    attempts    to develop reliab'le     ur;c!:i:,loJq::ent rates us-
    ing this methcdology,       because of the lack of current          1i;bor market data
    for local areas.       (See p. 27.)

    Further  study is required    to determine    the extent     to which the practices
    and experiences    of ihc States   included   in @,0's review are indicative        of
    those in other States.      GAO believes,   however, th;lt,     afthozgh they may
    vary in degree,    the problems experienced      by tile t\s:o States are charac-
    teristic   of those in many other States,       because of the general     lack of
    labor market data for small and rural       atveas.     (See p. 27.)



    Family income data for States and local areas are available      only from
    census information    gathered once every 10 years.  There are two pro-
    nounced drawb;?,cks to the use of these data in determining   current   eligi-
    bility.

           --They          do not      provide    a reasonably                 current   measure   of   income.

           --The preciseness    and reliability    of the data                              developed     for     small
.             areas is questiondblc.       (See p. 48.)

    Because current        fai:lily income data are not available,   the Economic De-
    velopment     Admir;-istr;1tion    is not able to make the annual review of area
    eligibility     based on income that is required       by the Public Works and Eco-
    nomic Development Act of 1955, as atnended (42 U.S.C. 3127) or to base its
    determinations       of maximum grant rates on recent data.




                                                                   2
DIGEST
-----.-
-c’
c
CJ

ti
G




          l-4




      .
           GAO believes     that the Ecozotnic Devcloplrcnt               Administration         should COII--
           sidcr   the feasibility      of using per capita             incollie data (dev~'l~~p~d by
           the Office    of Rusiness Ccononiics,        C~~I~~~till~~lt     Of CGll~liil2I~‘LC)~15 Oil? lii?GrTls
           by whi ch income levels        colrld be measured more friqueMl-\~                  l;hi:n ~'very 10
           years.     Any departure     from using median family                incollIe crite,>ii;,      how-
           ever, will    require    a chanr?e in legislation.               The proc-iscnes:z         ;!nd rcli-
           ability    of the per capita        incozie data has not been fully                .ic:i.cd.      GAO
           is not necessarily       advocating      the use of these data in their                    present
           form.     (See p* 56.)


-   RECOMW~l24TIO:iS OR ,SlII;GEST3Xi&’
                                ___--

           GAO is making       several     recommendations       designed      to improve     the system.
           For example:

              --The Secretary   of Labor should ascertain    changes needed to improve
                 unemployment estimates    and to monitor State unenl~loy~~!~ni:-estimating
                 practices.   (See p0 46.)

              --The Secretary     of Commerce should study the probl~s        associated   with
                 developing   current   unemplo~~llent and income data, consider      the use of
                 the more current     per capita   income data, and re~o~~,ci;d changes in
                 legislation   as warranted.      (See p. 57.)

           Other    recommendations        are contained       on pages 46 and 57.




           The Department       of Labor said that--in       line with GAO's rccommnndations
           and within     the constraints       of budnet resources       and staffing    ceilings--
           it would take steps to ensure uniformity               -in th c application    of the prc-
           scribed    estimating     techniques    and to improve the acc\lrzicy and compara-
           bility    of data.     The Department     said also that it would consider             the GAO
           report    as part of the Department's         evaluation      of the unemployment-
           estimating     procedures     that currently    was being made.          (See p. 46.)

           The Department     of Labor agreed with GAO's recomzendation            that the firId-
           ings of research      studies    on estimating      unemployment  should be converted
           into timely     and meaningful     action,     where practicable   c?nd feasible,     and
           that the Department       should improve the review and monitoring            procedures
       r   of the unE:;;ploymcnt,-estimat'ing       practices    of the State cimp?oyl;jent secu-
           rity  agencies.      (Se2 p. 47.)

       - The Department  of Labor said further   that the -iq~rovCments    necessary     in
         the methodology  ~:ould be made by the end of fiscal    yF:dr 19?7 and would
         take account of GAO's findings   as well as the Pindin:;s    of reyt3;rch   sti!dies




                                                        3
sponsored by tht:      1.r?-?r'ttl,::r\t ;;nd by the affi
                      LP                                    1 iated   StGtc   emr:loywnt   se-
curity  agcncics.       (See pn CL)            .




Further,  a work group is studyirrg per czpita incwz  data, b!~t much re-
mains to be dox before they can be usud to il>t?as~~~~arca xonomic   d.is-
tress.   (SW p. 50.)




Also the Secretary     of Comxcrcc may seek chan:_;~s in legislation            ,c'n the
basis of the rev-iw      ~'E!col;‘~e~:d~d 13~ GAO of tile probl:<>is zssociatcd      with
developing  unemplt,yr;wtt    aad incow       data,

GAO believes    that this     rkpor-t   will    be uwTi        to -itie Congwss     in consid-
ering these    matters .




                                          -3
                                     C 0 1-1 1. e 11 t - s
                                     _--_.--




    1                                                                             5
                                                                                  5
                      Econoii:ic development     assistance                       5
                      Pleasuring  economic    distress      on basis     of
                         unemployment    arid income                             10
    2            NEE13 TO IIPKCVE UNEK'LOYKE~T ESTIMATES                         14
                     Methodology      pre::i:ribed   for estimating
                        unemployment                                             14
                     Weaknesses     in the methodology          for esti-
                        mating unemplloyment                                     18
                     Observations      on ullemployi-nent-cstinintirlg
                        pracbtices    of State agencies                          27
                     Conclusions                                                 41
                     Agency comlnents and o::r evnl,.ustion                      43
                     Recommelldations         to the Secrctxy         of Labor   46
    3            NEED TO IMPROVE INCOFlE DATA                                    48
                     Limitntioxs     of data                                     48
                     Need for periodic       reexamination of maxi-
                        mum grant rate     income criteria                       55
                     Conclusions                                                 55
             P       Agency comments and our evaluation                          56
                     Kecom!nend,ztions   to the Secretary    of
                        Commerce                                                 57

APPENDIX

         I       Letter    dated Novrmber 30, 1970, from             the De-
                    partment    of Cownerce to the General             Ac-
                    counting    Office                                           61

.   II           Lettc-r   dated Ikcember  2, 1370,   from tk I DC.-
                    partment    of Labor to the Gcrieral  Accounting
                    Off ice                                                      65
APPENDIX

  III      Principal   officials  of the Departments      of
              Commerce> and Labor having responsibility
              for the activities   discussed  in this     re-
              port                                              73

                               ABRREVIATIOISS

ARA        Area     Redevelopment     Administration

EilA       Economic     Development     Administration

GAO        General     Accounting     Office

OBE        Office     of Business     Economics
       GAO is .making     several     recommendations     designed   to im,iYove the   system.
       For cxmpi c :

          --The Secretary   of labor should ascertai'n    ch;;rqL:s needed to jmprovc
             unemployment estimate s ;,nd to mwtitor   State un~ii;ployt-i!er;t-es%-imating
             praciices.   (See p. 46.)

          --The Secretary      of COXUIC~C~ should study the probleilis associaled         with
             developitlg  current   un3iploj4xnt      and income data, cons-icier the use of
             the mot~p current    per capita     income data, and recor,iir!end ch3 iICJ?S in
             legi:,lation  as warranted.       (See p. 57.)

       Other   recom:i!endations      are conta-ined     on pages 46 and 57.


AGimCY -.----.--l_-_-__-
         4 CT rims A iY!!J L’Ih?ESOL ;%I7 .r~~xlES

       The Department       of I-ahor said that--in      line with GAO's rccoxxendations
       and with-in the constraints          of budget resources      atjd staffing    ceilings--
       it would take steps to ensure uniformity               in the appiicztion      of the pre-
       scribed    estimating     techniques    and to improve the accuracy          and ccwparci-
       bility    ol" data.    The Deparxmcnt said also that it would consider                 the GAO
       report    2s part of the Department's         evaluation     of the unempIo.~,iznt-
       estimating     procedures     that currently    was being made.          (See p. 46.)

      The Department     of Labor agreed with GAO's recommendation             that the find-
      ings of research      studies    on estimating      une:nplcyn~cnt should be convex--l:ed
      into timely     and mean-Ingful    action,     where practicable    and feasib!e,     and
      that the Departmeni       should improve the review and monitoring            piwceciures
      of the un~mploymer~t--estilil~ting       practices    of the State employment secu-
    I rity  agencies.      (See pv 47.)

      The Department  of Labor slid further      that tl-te improvwnents    ncccssary    in
    - ti;? wthodolw!y  LJ~U'I2 be triade by the end of fiscal     year 197'1 Cznd would
      take account of GAO's findings      as well as the findings      of \resp;irch s'iudics




                                                     3
sponsored by the Department and by the affiliated              State   employment    se-
curity  agencies. (See p. 47.)

The Department   of Labor noted that the reliability           of estima'ling    unem-
ploy,nent by using the prescribed    estiw?ing      techniques      tended to de-
crease for small awas.       For many seal 'I , prcdoxi'nantly      rural   areas,  the
major problem is one of urldwcmoloy writ       of available       manpowr rather
than unemployment which the me-ii~odology is intc&cd            to measure.      Because
of this 9 the Department   suggcstcd  that alternative         approaches     to measur-
ing economic distress    might be needxi.      (See p, 43.)

The Department    of Comerce agreed,      in principle,    that it would be desir-
able to have more recent incams information           on a regular    !>asis but stated
that the costs of securing      such information      by dupiicatiny     Bureau of the
Census procedures    and techniques    appeared prohibjtive.         By us.ing other
data sources3 such as Office       of Business Economics per capita          income,
however,  the Economic Devel~p~et-~t Administration        hopes to develop rcason-
ably accurate   income estitnates.     (See p. 56.)

Further,  a work group is studying  per capita income data, but much re-
mains to be done before   they can be used to mcitsure area economic dis-
tress.   (See p. 58.)




Titles  I through       IV of the Public b!orks     and Ec~nn:~iic Dcvelolxxnt      Act of
19C5 expire    at the close of fiscal      year     1977.     The House Cc~i,??littce on
Public Works plans to conduct extensive             he6:rinrJ's on thr: Economic Devel -
optnent Ad,ninistration      and its programs     in mid-1971.

Also the Secretary   of Commerce may seek changes in legislation  on the
basis of the review recommended by GAO of the prublcms associated    k/ith
developing  unemployment and income data.

GAO believes    that this   report   will   be useful    to the Congress     in consid-
ering these    matters.




                                      4
    dcvelopmcnt   of areas and regions    through    the creation                     of new
    cmplo);ment  opportunities  by developing     new facilities                      and
    resources   and expanding  existing   ones.

            EDA provides      financial,         technical,       and planniiig      assis-
    tance    through

            --grants     and loans         to help   build      or expand   public      facil-
!               ities,
/
            --business      development   loans to private    industrial       and
               comipnercial    firms  and to local gover~ncnt     a~,,;r~cic~9

            --technical    assistance          contracts        and grants   to help
                areas assess their           needs for        economic growth   and plan
                specific  projects,          and

            --planning     grants    to assist         public   bodies    in drawing          up
               and carrying      out economic          development     progr;:ms.

          As of October    31,        1970, EDA had approve?     i3hz sranting
    of financial  assistance           for 3,396 projects   total.i;:z   about
    $1.3 billion,   comprising           grants of about $949 miii ion and
    loans of $359 million.

          To be eligible      for assistance,       a project    m,:;s't: be lo-
    cated in (1) an area designated            by the Secretary       of Com-
    merce as a redevelopment         areas (2) an area which tlie Sccre-
    tary of Labor found to have been an area of substailtial                     jln-
    employment   during    the preceding       calendar    year (referred         to
    as a Title    I area),    (3) an economic       development     district,
    or (4) an economic development          region.       A redevelopment
    area may be a county,        a labor   area, an Indian       reservation,
    or a municipality      having    a population       of 250,000    or more.

           The majority        of the redevelopment        areas qualified         by
    EDA for Federal         assistance      in fiscal   year 1969 were collnties
    made up of rural         communities,        Economic development        districts
    must contain     at least        two redevelopment      areas and either          a
    redevelopment      center      or an economic development          c(Jnter.
    The development         center    must be an area or city         of sufficient
    size and potential          to foster     the economic     growth activities
    necessary     to alleviate         the distress    of redevelopment         areas
    within    the district.          Centers    within  redevelopmilt       arcns
         'I'he act provides   that the Secretary   of Labor deterwine
the rate of unemployment          and provide  the data to be used by
the Secretary       of Commerce in making determinations     of sub-
stantial       and persistent    unemployment.

       The act     defines      substantial       and persistent           unwploy-
ment   as:

       1 e Unempl.oymcilt   of 6 percent           or more during           the     latest
           calendar   year,

       2. An annual average rate Of unznployme~-lt                    of at lQa:jt
          6 percent  for one of the time periods                     specified   in
          item 3 below,

       3. An annual          average   unemployment       rate     of at      least:

           a.    50 percent    above the national     average               for     3 of
                 the preceding     4 calendar  years.

           b,    75 percent    above the national     average               for     2 of
                 the preceding     3 calendar  years,

           c,    100 percent   above the national                average      for      1 of
                 the preceding   2 calendar   years.

       The al'kiual.   av~?rage unemployment   rates    used               in 1969 for
                              substantial  and persistent                  unewp:lo~V-




                                              7
                         National
      Calendar         average unem-          50%         75%          100%
        year           p1oymznt
                             ---  ra.te      --above     above        above
                                                                      --
         1965                 4.5             6.8         7.9           9.0
         1966                 3.8             5.7         6.7           7.6
         1967                 3.8             5.7         6.7           7.6
         19G8                 3.6             5.4         6.3           7.2

      Also section       102 of the act provides           for the designa-
tion of areas suffering         from substantial         unempl-oyincnt,   Which
is statutorily      defined    as 6 percent       or more during       the pre-
ceding year.       These areas (Title       I areas> are eligible          only
for public     works and development        facilities        grant assistance
and not for public        works or business         development     loans.

        Eligible    areas must indicate   their   desire    to partici-
pate in the EDA program by making a formal             request    for des-
ignation       and by submitting  an overall    economic    development
program.

        EDA makes determinations    of area eligibility           on the ba-
sis of data supplied      by the 'Manpower ,~dmr;j~isl-.i-r,;-f.or~, Depart-
ment of Labor;    the Dureau of the Censu:;, Dep;frGX~t              of Coin-
merce; and the Bureau of Indian       Affaii:s,  Dzpart~:t~nt        of In-
terior.

       As of     February    2,     1970, 943 areas were         qualified           under
the   various     qualifying        criteria, as follows:

       Title   I areas                                                         38
       Redevelopment       areas:
             Unemployment                                                     412
             Population       loss                                            103
             Unemployment         and population       loss                    48
             Income                                                           150
             Unemployment         and income                                    58
             Indian     reservations                                           94
             Sudden rise        in unemployment                                27
             Other                                                             13

                     Total                                                    943
       EDA uses    unemp1.oyment and income data also to establish
maxim-c!m gr,ant   ra tcs for eligible  areas.   As of October  1970
the fol?-owing     criteria   were used for making grant   rate de-
terminations.

 I%xFrm!n
   grant                -- Needed
                           ---    ..--.to qrralify     for  rate _-__-
                                                        ___._-
    rate       Median fani ly                        Annual average                -
(percent)           income           or            unemplo-+ment     rate--

       80    $1,600    or less               12 percent      or higher

       70    $1,601    to $1,800             10 to    11.9   percent

       60    $1,801    to $2,000             8 to 9.9      percent     or double
                                             the U.S.      average     in 3 of
                                             the past      4 years

       50     over    $2,000                 Under 8 percent    and not
                                             double TJ, S, avc~ngc during
                                             3 of the past 4 years

         The act requiycs    EDA to conduct        an annual reviz17         of all
designated      areas to determine   whether        previously     cfuzl.ified
areas continue       to meet the statutory         criterta.      The review
is made to determine        also the maximum        grant    rate for the
follo;Jing    year.




                                        9
            Also EDA designations             may inflwnce           the geographical
    distribution         of funds of other Federal                agencies.     For cx-
    ample, under the provisions                 of part C of the Manpower De-
    veloyment         and Training      Act of 1962, as ame~~dclcl (42 U.S.C.
    26105)      9 tllc Sccrstaries       of Labor and of Health,             Education,
    and Xclfarc         are authorized        to pro~.~i Cl<: ;1 SUpplPZ2; I tarjT 13jYO-
    gram of training            and training       alloi*iances      for 11ncmployed and
    underemployed          p~ersons residing         in areas dcsig:l;-ted       as rc-
    development         areas by the Sccrctary              of Comm,?rce.     Further-
    more, Federal          procurement       prefcrencc       eligibility     is a,ut!lo-
    rized      for firms      located     in areas of high clnemp,'l:)yment under
    Deferlse Manpo\\Ter Pol.icy          4 9 revised,        and Executi;~e     Order
    10582 implementing             the Buy American          Act,

             Judging       from observations         made during     our review,        the
    problems         associated        wi.th the development     of current        and re-
    liable       statistical         data by which to measure economic              dis-
    tress      are many.         liJe recognize    that these problems          are not
     subject       to simple        solutions.     It is especially      true of
     small areas.--cll;~l_actE‘.r:i ,stically         EDA redevelopment         areas--
    where statistical               data normally     are not gathered       on a con-
    tinuing       basis and where the costs of devcLoping                    meani:lgful
.
    statistical           dat:-, are significantly       higiler   than those con-
    nected with dcv:~lopi.ng                data on a national    or re&iOilal.      basis.

            The conccl~i-s and definitions       used by the Department       of
    Labor and aff i lj ated State employment          security   agel;cies  in
    m:~asliring  uner,pl oymefit on a local-area       bar;is are identical
    with those used in estimating          national    unemployment.
                National       ~~ncm~:ioym~nt estimates,                unlike    the State     and
    local         estimates;         are b;,is12d on survey:;           of about 50,000 house-
    holds tl-1roughout                the COUiItrjr          .  The hcu sehold s arc scientif-
     ical!y          sel-ec ted e‘cch mol-Lt!l for visits               by interviewers       to
    gathzr           informaI_iu:l         on the job-holding           and job-seeZ:<.ng     acti-
    vities           during     I.!IZ peek prccding               the interview.        The national
    UII~~l~‘lOyJ!?elIt        EsSt i JJI:! I: C S , r;;h?-Ch   are those quoted in the news
    imedia, are more ccxmonly                          kliown than are local        area estimates   ~
I

               The Department       of Labor has devel.aped                a methodology
    for measuring          c:rrrcnt     unemployment          of local      and State       areas
    that makes use of labor                 market data developed               i.rl administer-
     ing tile Federal-Stntc             er:;ployment        security    progra..s.         This
    methodoln;;y        was initial.?>          designed      for large metropolitan
    labor       areas;   howe\~cr,      it wasadopted           later   for use in csti-
    mating uno~pIoyme.nt            in small areas when it bec,amc evident
     that such information              hnd to be developed              to enable       the De-
    partK~Pi?t      to meet its       re,~;ponsibif.it~~es          set i:y Ic~i.sla:-i.011      for
    econo,nic       devclopmci-it     proi;rams.           The met!~~!o?ogy , contzi ned
     in the "Handbook            on Est.imat< ng Unempl-oymeilt"              arid the "IIand-
     book on Development            of Basic Labor Narkct                Inforrr!ztiorl     for
    Small Areas,"          is basically           the same as that developed                in
     1960.

                  A modification          with respect         to one major element            of the
    methodol-ogy--             estimating      unemployment          among entrants       alld
    rcelItrarlts--s~a~              implemented        in mid-1945        to take into       account
    the structural                 changes in youth unemployment                 which had oc-
    curred             since the development             of the methodology.           Witllin      re-
    cent years the Department                       has initiated         and sponsored        several
    additional              studies     designed       to prodlxe       information     which
    could be used to improve                      the methodology;           howe\7er,  the stud-
     .
     1"::       completed        to date have not yet resulted                  in an improved
    methodology.                 The unemployment           est!'.mates     are developed        by
     the State             empInymcnt       securit-y      agencies.        Data   for about      150
    major labor               arcas and for those small areas which have an
    uncn~plojmei~t             r?%te of 6 percent           or more are furnished            to the
     Department             of Labor.

              The Eanpo;,Jer Admii-pi s tration     is responsible     for cstab-
     lislling    and i s:;:ling policies,       procedures,     and regulatlions                     to

                                                      11
State empI11.0ym2nt security                      ag,eIlcies         for ensuring      effective,
efficient,  and economical                        operatj.on           of the program       and for
as:;isting         StnLc     age-nc LCJS j.n meeting                   problems    peculiar       to
their      localities.           I'he    hp2~~:ti!lci-lt           makes     funds av8ilable          to
the States          for the administration                         of the employment          security
Pi-O.gram on the basis of their                              budget requests         and of the
Department ' s dete:-mination                         of the amount necess,ary              for the
proper       and efficient             administration                  of tile employment         se-
curity       pt‘ogram.        Department                 officials        informed    us that       staff
resources          available         to the States                 for the unempl.oyment            rate
estimatj.ng         program amounted,                     on the average,          to less than
one person for each State.

         The only median family         income data available      for local
areas are those developed             as part of the census of popula-
tion,     conducted        once every I.0 years.    Altho~~gh some consid-
eration      has been given to the development            of a system by
which ar-:l~ual changes in median family            income co~~ld be de-
termined,       to meet the annual rcv.j.ew rcquircments          of the
legislation,        little     if any progrc:ss  has been made toward
implementing        such a systern.

        The concern        TJhich we express          in the foLXow?ng          sections
of this      report     regardi.ng      the soundness        of the statistical
data used in determining                area eligib?Lity          is not without
precedent.          During    the LO-year period            of the administration
of the Area Redevelopment                 Act and the Pubf.ic Vorks and Eco-
nomic Development            Act, members of the Congress,                 Federal      and
State officials,           and other       interested       persons     have expressed
similar      concern.        Discussions        with officials        of the Depart-
ment of Labor and a former                 ARA official        and our review         of
testimony       by Department         off2ciaI.s      before    congressional         com-
mittees      indicate      to us that the available              statistical        data
on income and unemployment                 were never considered             to be fully
satisfactory         for the purpose           of area designation.

       The Tresident's      Committee  to Appraise  Employment   and
lJncmp!.oyment   Statistics    (Gordon Committee)    commented,  in its
final    report  to the President     dated September   1362, that:

        "State and local   labor-force   statistics     s-re nei-
        ther as accurate   nor as complete      as those on a na-
        tional level.    To judge by comments mad2 to the


                                                    12
          Committee,  there  i:; probably     no clement in our
          system of labor-f2    TCC reports    which is more in
          need of improvement ..'I

       Although     the need to improve the accuracy    and reliabil-
ity of tllc stat~sticai      data has been recognized    for many
YfZWS, we found that the situation        had not improved.       De-
tailed     corxnents on the adequacy of the statistical      data
f0ilow.




                                        13
         We believe     that the currency             and accuracy        of unemploy-
ment data used by           CDA    in   determining       the eligibility         of
areas for economic           development         assistance     is qclestionable
and sho[lId be improved              to ensure a realistic            appraisal      of
the economic        conditions         existing     in such a1‘ea.s.         It is
questionable        whether     tllose areas of the country                experiencing
economic distress,           as indicated         by high unemployment            and/or
low Income levels,           are, in all instances,            being properly
identified.

       The questionable      reliability       of the unemployment       data
is attributable       to conceptual       weaknesses    in tile prescribed
methodology     for estimating         unemployment   as well    as to prob-
le1n:4 in developing     the information        necessary     to compute the
unemployment     rates   for small areas 0




       The methodolo:;y     for estimating       unemployment       entails
what is commonly referred           to as a building-block          approach.
This approach     reqllires    estimates      of (1) unemployment          related
to covered    employment--     that is employment         comillg under the
Federal-State     unemployment        insurance    program,      (2) unemploy-
ment related    to noncovered         unemployment,      and (3) unemploy-
ment related    to entrants       and reentrants       to the work force.

      Unemployed    entrants    are tilose persons who have entered
the labor market For the first          time and who have not found
jobs.    Unemploy~_d reentrants     are those individuals     who have
had prior    work experience     and who are now looking     for work
but who were out of the labor         force  for some time.     The un-
empl0yx:~n-t totals    for the three blocks      are added to arrive




                                            14
at an cstimntlt    of al.1 uncnployment    within  an area.      The un-
empI.oynwnt rCte estimate      i.s computed by dividing      the total
uncmploymcnt    estimate   by the total    work force   estimate.1

        Since  t1-z work force    comprises      both employed    and unem-
ployed,     it is necessary    to develop      information    on area em-
plojrn!elYcI as wV:?.l as unczploymcnt.       Employment    data also is
an csscnt hi factor       in estimating      unemployment     for industries
not covered      by the Federal-State       unemployment    insurance
programs.

        Unemployment      figures     for covered workers            are ba:;ed on
un ,nploymrnt    compcnsati~on        claims      data filed    with the State
em?loymcnt     security      offices     and on estimates          of the number
of unemployed     :rorl:ers      who previously       held jobs in covered
establishments       but who are not receiving               benefits,        Included
are those who have been disqualified                   from receiving         bene-
fits,    have exhausted         their   benefits,     have delayed       filing,
or have never filed           for heneyits.

       Unemployment        estimates     for noncovered        workers     are de-
veloped    by (1) computing          a covered     unemploy;i,ent      rate and
 (2) applying       this   rate to employment         estimates      for non-
cof.7pre-j ';Tor:<"rs on the basis of certain              ass\*mptions     which
have been made regarding             tile relationships        between the un-
employment       rates   for covered workers          and those for non-
covered workers.

        The assump;.ions    are based           on 1957-59 studies       by the
Department     of Labor of national               unemployment   conditions
which showed that,       nationally,            there were fairly       constant
relati~onship    s in the incidence             of unemployment    between
covered     and noncovered     workers.

      The methcdology      prcscribcd     by the Department     includes
the follo:?rin g prcdetcrmined        unemployment rates    for non-
covered    groups.
                 -
1
    The term "Work force"     differs  to some extent               from the la-
    bor force   corlcept  used in measuring  national               unemploymerlt
    rates.    (ScAe pn 22.1



                                           15
         Noncovcrfd      zrout3D.S
                             ---

Sn~all   firms    and railroads
Nonprofit      institutions

Domestics

Nona~ricu?~tural.    scl'-
   employed and unpn:i.d
    family  workers
&ricu%tural       wage and
    salary  workers


Agricultural   self-employed
   and unpaid  faxily    workers
State and local     gowrnments


          The covered       c~~rploymer~t
covered      rate are ba.r,cd on informat2on              S”dLXiti(;!d     c!Iz3rterly
by firms       coming under the Fecleral-- Stutz :-incscpl. f~;;~.',::         lit insur-
ance programs.           EIPpIOp-lC~X     figures     for nonc:c!~?:\i.i.A [;roups--
st?ch as farms,         small firms,        domestics,     and nor~piol it insti,-
tut.ions-    - are derived        from  va;r-ious   source:;.        'Jy>ey;c sources
include      the censuses        of population        and a~gricultzrrc;          social
security       information;         and, in some instzllce:;,           c~~~~l.oyment
surveys      made by the State employment                security       agencies.

        Unemployment     estimates     For new entrants       and reentrants
to the work force        are developed      using an estimating        tech-
nique prescribed        by the Department       of Labor in 1965.         The
technique     is bui.l.t on relationships         noted in a study of
available     national    data on new-worker         unemployment    for the
1350-G4 period.         The factors     used to comptlte unemployment
for a State or area are dctermincd              from the yout-h-population
ratio,    which is the ratio        between the population        aged 14 to
19 years and the population            aged 20 yer:rs and over,

       The accuracy    and reliabil   ity of the UCl?l3p?.Oj~j~~Xlt rate
estim:~te  2 s contingent    on the developm,::nt  of com;~Y!.ete and
accuratr-? data for both empILoymcnt and ,l:nemp7 ojrment.         Any in-
accuracies    Fn estimates     of cmplq~~ent   and wlqzmpj.oyment for

                                                    16
covered    workers are compomded       further when the     rate de-
rived   from these cstimtes        is used in determining      u.nen?ploy-
mat for norlcovercd     wr,r?:~rs.




                                     17
       Colm?ptually     f the accuracy   of the e:;timates    yielded by
the methodology       is based on the reasonin;.     that most of the
data necessary      to pp!:oTGde a me.3::ure of uncmploj,n~r,nt    in
Staixs   ctnct areas arc available     througl~ the Federal-State
ur?~Tili3Ilo)-~l~LI~           iIlSIlT;!llC?                    C~pQr^atf9llS             nl-~d       th-lt     UlWiXp~   0)TIlPJXt   as-
sociated               txith      induc        r,ci*---; .,eS           cr:rvice
                                                                     CnVe%cd          Op-
                                                                                     by           C5lYZlOyi3C?I~t
erations           accounts      for most of the total      unemploy:-znt.            In
support           of this     reasoning,      the "Handbook   on Estimating           Un-
el-rlp1.oyinrlTL”      devclopr:d     by the Department      states     that the
unenploymcI:nt           insurance      program covers about 75 percent               of
all wage and salary                 workers.     The handbook     states     further:

               "Relative1.y    little data,    in addi.tion   to tho:;e se-
               cuj:ed as a byproduct      of em;>loymei-lis secl-rity   op-
               erations,    are ncccssasy    in order to proviclc       a rea-
               sonable measure of u.t~cmploymc~~t in States           and
               areas at a minimum cost,"

        On the basis of several   observations     we m~idc\. during
our revi2?7,T we have concluded     that the met7~od~?Lo~y does not
provide    re~sonsbiy accurate   e stinates    of conditions     cxist-
ing in small and rural     areas for purposes      of area dcsigna-
tion,

                --A substantial       number of workers    are cmpl.oyed in in-
                   dustries   which are not covered      by the une:mployment
                   insurance    program.     The employme,~t security    report-
                   ing system used as a bnse for estimating           employment
                   does not capture      the labor  changes for this      sector
                   of employment,

               --The industries          not covered     by the unemployment           in-
                  surance      programs     are characteristic          of those found
                  in small,       predominantly     rural      communities.         Such in-
                  dustries      are made up of farms,           agricultural       products
                  processing,        maI     firms,   nonprofit       organizations,
                  d0TlEStiC.s)     and State and local          governments.

               --Covered      employment                               data are             not          developed        on a
                  resident-labor-force                                    basis.


                                                                                18
            - -‘Ihe             bet:rc zcn covt~r~.d and noncovered
                      Y.-p ti0.c;                                             uncmploy-
                merit and the factor:;          for entrant:;      and reentrant-s
                used is1 the lnrthodoi.ogy           T;ere devclopcd      on the basis
                of nst:lonc;l     sam~lcs and are not sensitive.              to the
                spccis l ccononi.c        co7lditions       which xay c;:i.::t at the
                'iocal.-arEz~~ lmeY-.21c Also the .studics            on 1:11ic31 these
                ratios     and fat to-i%s were based are now old,                Ijecause
                EDA rcdevl-~lo;~~!x~nt areas generally             invo~:~*c s.i.n;slc
                counties      and because designations             are based,       in part,
                on ur,cl-,;~~1.o:~rni~ntin those count?es 3 the recognition
                of l.oc:,l r--ma economic           conditions     is essenti;~l      to
                making the p-roper designation                 of areas.




            The Manpower I&port                 of the President,       transmitted      to
    the Congress            in Pkrch 1970, reported           that in calendar          year
    1968 nearly            17 m?.I.‘ij_on ITage and salary        jobs--almost        25 per-
    cent of all jobs of this kind--were                      not covered        by unemploy-
    ment in:x!ranae .o 'iTrlc se jobs involved                mainly      ~orkcrs     on farms
    and in SixIxz and local                  governments,    domc.stic      scrv.~c~~, non-
    profit       orgai!.i.-.nt:ions       : agrTcultu.ral  products       process kg,      and
    small f~~xns n '.T%e Fk~ploym~nt Security                   A~~Y-K?E~~~~   s of 1970
     (pub0 Lo 91*-g3)              7;"l'l
                                    ,d~e..- extend coverage     to as maxy as 4,4 mil-
    lion more of t!-Je.W               zig? and salary     jobs.       Ever, with this
    legis~.r:ti.on,         ho1:eve.L: 9 a substantial       number of jobs kll            re-
        .
    marn uncovered.

            The smz~~'l, predominately   rural   areas,    such as those
    which ,c,enernlly     make up the counties      designated     by EDA as
    redevehpen      i: ~'~2.7s 3 account for much of the noncovered
    emploj;ix!nt  and 'hF17e high concentr<ations       of industries     not
    cove-red h; the FccIeral-State       un~mplojment      insurance   pro-
    grams.

          As an exnmpl.e, a research               study report    prepared     for
-   the Nor:,;1 C---, 4~ dl~ri:-t E5fiployment Sccur?.ty    Commission      :in Nay
    1968 03 i-.Je iq:~~~i-~lodology
                          :?-I-            for Estimating    Uricmployrnent     in
.   Rural. AYCGSI' pointled          out that in 1965 E 61 of the State's
    100 counties         had fewer than 50 percent          of their    workers
    covered     and "ihat 9 of i:hese 61 counties,            26 fell   below 30 per-
    Ceilt  alld  tilTE?e    b6~l~~~~ 10 pf?TCeTl’t ,
                                                                                                         i
          The ability             of the methodology               to adequately       measure           ;I
    unemployment      is        significantly             impnired      where covered         empl-oy-   !
    IllCIlt   dot? S Rot    TepIYP   SClT’r.   a   sub::t;untial       pzrt of the total         em-      i
    ployne11t.         'This is err.~ l'or a number of rcnsonc;:                        (1) the           ;
    methcdol ogy relics             upon the empl.oyment and unemployment
     data otiain~d          from co\fercd           establishments           to lend substance           ,
    and rcli.ab:i.~.i ty to the x~rk force                       and area ullcmpl~oyl*lel:t      es-
     timate s 9 (2) when the covered                      rate computed is based on a
     Low percenttrz;e         of workers,           it might not be stntir~tically
     reliable       for estimating             uncmplo-yment         re!.ated   to noncovered
    WOi~ke?:S I and (3) the actual                    ratios      between covcrc:d and non-
     Covered unemployxcnt                :‘!il the    area COuld        Vary   suhstCailtially
     from the ratios            incorporated            in the methodology.

             Department       of Labor officials          informed      us that they
    had recognized          that the methodology            might not yield       rcason-
    ably accurate         results    for rural        areas because of the low
    percentage       of cox7ered employment             generally     follnd in such
    areas and tllat         they believed        tha-t the results         for metropoli-
    tan urban areas were co~lsidcrahly                   more ~lccurate,       1"ne offi-
    cials     stated    that the Do?   ._,2rarlxnent had cx~~rcssc:! coiiccrn          to
    official     s of EDA and AX.4 .tiCth respect              I:O usi;?g l;~~i=:n;~loyrnent
    rate statistics           as mcasur: ,lTents of econcmtc: dlir:trc!ss           on
    T\rlnich to base eiigihili,ty           detc-ruination>;,

             In a letter      dated I%rch 13 p l?G3 9 the Assl.s::~~nt Secrc-
    tary for Hanpower            Dr.partmcnt   of P.?bor, COIiT?C?j?    c:\>d J-,o i-he
    Assistant      Secretary      for Economic 'kvelop?ient          ) lk~;~,r:tlr:i~nt of
    COiXJlerCJe :, that   for    small,   predominantly   rural        l.abnr areas:

            "***    thn.- major problem has always been one of un-
            dercmployment         and under-utilization        of available
            manpower resources,           rather     than unemployment,
            Intermittent         empl.oywnt     at low level     jobs and in
            subsistence        farming    may hold down the local         unem-
            ployment       rate and tend to conceal         widespread
.           poverty      in many such areas."

            The Assistant          Scretary         commented        further:

            "We believe     therefore,      that you may wish to gLvc
            Seine further     tho;lght  to the problem   of the basis
            used for de sign:.;ting     area:; which arc predomi-
            nnntly    rural   in character."

                                                    20
      As noted on p:ge 43,                    the legislative          history     of the
Public TJorlcs and E,-onpmic                  I?e~~I.opment Act indicates             that the
median family         incoxz        cri t:s ion incl.ucJcd in the act was in-
tended primari1.y           as a mx.:;u~cment             of ttilderemploymcnt           for
those cs:;cntsiai_l.~,y rural. areas xhere there are "'really                              no
llle~sui?c:-; of uncr;?-il,I.oynent        0”    Therefore      a criterion        for mea-
 suring     undermpl.oyi~:cnt           exists e Because median famLly                     in-
come d2ta for :!oml.             areas        are   developed       only    once   every       10
years as part of the census of population,                               howwer,      such in-
COIX d2-h nrc not measuring                      the cursx!nt
                                                           we..- economic         conclitions
of thox        r~mroi areas j n which un~cmploymo                          t is a crucial
elemcIlt 0 In the absence of current                         income data,         EDA has
continued        to place reliance               on the unemploymenC            rates      as
mea sure s of economic              dj.:;tress      in such areas.            The riced for
current      income statistics                is discussed        in greater       detail
i.n chapter        3.

          As of December 1970 the Department                  of 'labor,       in cooper-
ation      with the affi.liatcd          Nevada Employment            Security     Agency,
was sgo;~soring         a research       study to attempt           to dcvclop       ne'bar
method:; of measuring             underemployment      and unLneT~~Li ?.T.~;ati.on
of manpower that could be used, togetller                       wi.tJl or in lieu          of
area ~rnemployment           data 9   as  a  basis for      &2terj:linii?+;      XVI   el.i-
gibility.         The study I*511 reesamkx           rele-i7;i1    !t cJ:?ix, on rur:il
          e
countxe.';     including       those data available           from the %?~a i--km:.nts
of Comxrce,          Agriculture,        and Iabor and affiliated                State
agencies,




                                                21
Resident      l.aLoj:
                 -----   force    not    adcquatelv---m--Ameasu?:cd

         If unempIo;;-r!ent          rates    are to provide         an accurate           me:;-,
 surex-~ent of the economic                conditions      of a spcczr,.c    . C" ai-ea, it
scxms necessary            that the rates           apply to ~11 em~~l~oyccl and un-
employed residents               of the area.         The labor          force    c~nccpt       used
2~1 measuring        ;xz-iiLcnaZ       ~~-rempioy-ment rates       dots cn~lnt both
cmployed        and llneii;lp:loyod on a residence              b>::i:j.        Unlike     this
concrspt,       however,       the l~ork force        concept      1lsc.d in the met);-
odol.ogy for measuring                 unemployment.     in smz;i1 k?rei+s COU”,~S              the
employed        at thzir      places       Of  Work.     The \;;,rk force            figures,
therefore,        include        s:orkers     commuting      into the arca from
other      areas and may include               a PC;-son more than once to the
extent       that multiple           job holdings       exist    during         a payroll-      pc-
Cod,         On the other hand residents                 comz::lting        out of the area
arc excluded          from the work force             count but included                 in the
labor      force    data.




        Household       surveys    conducted      in 1969 by several        State
employment        security     agencies,     in cooperation       with the De-
par-hen-t     of Labor,      showed that the number of workers               com-
muting    between courities         was substantial.          For cx3c!ple,     in-
commuters       for one of the counties             accounted    for 1,%29 of the
total    covered      employment      of 3,456 workers.          In another
county,     in-commuters        accounted      for 431 of the 899 5:5orkers in
covcrcd     emplo;~ment e Out-coriiiluters           for these lx0 cour~ties
were 4.34 and S45, respectively,

      We noted that a report        prepared   by a university      study
team for the Mississippi         Emploj?:lznt Commission    showed that2
for 10 Mic-n:
          .3.J2.ssi.ppi counties    surveyed,   net commuting     .in 1967
ranged from -17.7 percent         to +57,5 percent.



                                               22
           The cons:istcr:cy of the ratios    and factors zlsed with
    CLETXl-li:locel conditions    is ~les~~onc25le--f~cm  t!;e stand-
    point of both variability       between the national  ss~??I.e data
    ,and locnl conditf-ons and the period of time r:hich has
    elapsed since the development of the ratios and factors.

           Early concern regarding     the uses of national   ratios for
    estimating     area u~emp?qment was expressed by the Gordon
    Coixni.tt:c?e. In its report to the President,      issued in 1962,
    entitled     "P9cas:xin.g Employwnt ,and 'Unemployment,'" the Com-
    mittee stated,      in part:
.
                 ssSlr.chratios   are obvrously      suspect l&en applied
                 to States or     localities      in which the composition
                 of the labor     force and local conditions      vary
                 markedly from      the r:ational    pattern."
                     *             J;          3(          *           Ji-
            .




                24


--..   -_
       1‘k+c-k    since the 1.957-1959 formative           years,       in-
       sured        [covered ] unemployment       in the 1J.S. has
       fallen        from more than one-half        Of   ~3.1  mcmploy-

       ment       to a l.evcl  nearer   one-third      *-?c~~~"

       Our             of estimates
                 snalysis                 of covcrcd     and total         uncmploy-
ment report4      by the State employment          sccuriLy        agenci,?s
showed that between 1364 and 1968 co,vered unempiq~nont                        had
dropped from 46 .5 to 35.7 percent           of total     Ul~li~>lO~TlWlt,        as
indicated    in the following      table.

Calendar                         Uneaplp,vment
                             _-_-__                                             Percent       of covered
  ye2.r                     Total
                            --              --Covered                      to    total ..---A--
                                                                                            uncm~loyment

                                 (000   omitted)

   1960                     4,097                1,906                                          46.5
   1963-                    5,001                2,290                                          45.8
   1962                     4,204                1,783                                          42 o4
   1963                     4,096                1,806                                          44.0
   1964                     3,802                1,605                                          4.2‘2
   1965                     3,418                1,328                                          38,8
   1966                     2,995                1,061                                          35.4
   1967                     3,173                1,205                                          38,0
   1968                     3,108                1,111                                          35.7

        With respect     to decreases   in specific    categories    of em-
ployment 9 the July 1969 report         noted that from 1958 the in-
sured (covered)       un:mpIoyment,    nationwide,    had declined     sub-
stenti.aIly    Easter    than had unemployment      among domestic
workers o The report         conta%ned the following      statistics    com-
par?ng the annual rates          of unemployment    among covered    and
domestic     wor'kers  for the calendar      years 1958-68,
                     Cal tndar              ----.----               ---. lp~lopent:
                                                           Rate --_ of           ..--.   -----
                       J’fnr                -Covered      workc~s
                                              --__ -..--___                      Domecti __c wor;:ers
                                                                                -_--_          ---_

                        1958                         6.4                                  5.6
                        1959                         4.4                                  r
                        1460                         4.8                                  ;:3
                        1961                                                              6.4
                        1962                         2::                                  5.5
                        1963                         4.3                                  5.8
                        19%                          3.7                                  5.4
                        I”65                         3.0                                  4.7
                        l&6                          2.3                                  4.1
                        19G7                         2.5                                  4.1
                        19GB                         2.3                                  4.0
             Department      of Labor offiti.als    info7.yi1ed 11s tf:.:t    in many
    il~lst.~i~ces the differences         between the !IOGSC~L~~       s;';Tv(z~ re-
    sults     and the estimates         developed  using the mc izhn,k)'logy fell
    within      tllc predicted     range of sampling      ,var2 2bi 1%ty of the
    houscho1.d surveys         and may not represent        trlle d?flerc?~ces.
    Tlley said that this         factor     would be considered      411 the anal-
    ysis of the survey results.




.
            Our cv(liuatioJ-1      of the unc~mpl.oyn;ent-(~st-ma~i1lI;            prn::tices
    of elnj~~~o~,T!ent sec\irity      ngc::nci.es in two States          SliO\SC-d    that
    the Sent&? agencies          were: severely      handicapped       in tl-ie'ir       ef-
    forts    to develop      currcn-.z unemplo>mcnt          rate:;:  because or the
    lcack of reliable         and currant       labor market data for local
    arens 8 In their          efforts    to deri17e unemployment              e~t:'im:~:2es p
    state    agency employee:;        hc?vc found it necessary             to use old
    and incompl.cte       statistical       data and to follo?~~ r:stCmzting
    practices     which are far from adequate                for tile develop:-r:cnt
    of reasonably        accurate     and reliabl      e unemr;lo)7:ient       cstimiltcs     a
    The methodology         for estimating         unemployment      prescribed            by
    the Department        of Labor is subjected            to varied        dcgrecs        of
    modification       and is not uniformly           applied,




            Further      study is required         to ascertain       the extent     to
    which the practices             and experiences       of States A and B zre
    indicative        of those in other          States.      We believe,    ho:jever,
    that,     althovgh      they may vary in degree the problems                experi-
    enced by States           A and R are characteristic            of those in many
    other     States,     because of the general            lack of current       labor
    market data for           small    and rural     areas.

.   Current
    ---       I- em-pent         data    not   readil..available

.            WC! found that much of the employment                data necessary
    for acc2~~~~1'~ly c stimating              area unemployment      rates    in States
    A and 13 .'c7:e not available                on a current  basis.       Al :;o tile
    c 0 v e 1';.'ri ~;~>i>lo~~~cI~\tdata arc not developed        on the bnsis or'
    worker :, t residences,            and this     fact had a tendency        to distort
    tile actual         employment      statistics      for an area.

                                                37
.




    28
                                                                     -                             1 .




          The changes in employment               indicated       by the moiltilly
  Sa?llpleS al-t'? CGnS!'dC?red as Sting            the total        changes for the
  industry    D It apllcars       unlikely,         however,      that such is the
  case.     For cxzlple,       if the manufacturing                industries           sampled
  reported      a change in emp'ioyment             of 20 workers          from one month
  to the next,        the 20 workers        -,iroul.d be considc-cd            as being the
  total    ch<lilges nffccting       -i-he nanufacturi;)g            industrj.es         within
  the area.        Th!L s assumption      would be made rcg,-.?.-dl css of the
  number of ~~mploycrs reporting                 or the IRlT~il:i~r     Of   wOI:!f:cTs      in-
  cluded    in a speci.f?'.c industry.

          The State labor         analyst     informed       us that the "linking
  method"    sug~cstcd       in the I?cpzr;ssncnt of L:.:bor I s 'i-innL17x~ok for
  estimating     unelrl!~,1~~j~ent for the current                m::ntl~ g!nclral.ly      was
  not usr>d by locnl         office     employees        because of tire rj_sk of
  arriving    at a pcreentagc-change               ratio      influenced        by atypi-
  cal situations        rather      than at meaningful             trends     in labor
  turnover.      The linkl:~lg        method involves           computing       the rcla-
  tionship    which the sampI.e for the current                      period     bears to
  the sample for the           prior    period     and    applying        the   percentage
  obtained    to the estimated            employment        for the prior          period.

          The exrample       shown in the follow!.ng          table    illustrates
  the differences          in the monthly         employment     estimates       result-
  ing from      the act~xl-change            method and the preferred            Iinking
  method,       The resrrlts      obtained       could affect      the covered        unem-
- ployment      r;ite,   the U~~mplOjmlerlt        es:,imates    for noncovered          in-
  dustries,        and the overall          unempl_oyment rate for the area.
   700      725           25            1.0355,       2,000        2,025            2,070




        ??ze policy        or' the employment       security        agc:nc:i_cs     of both
States     h and H is to adjust             the monthly        estcmates,         if neces-
sc?ry> when actual            quarterly     covered    emp!-oymcnt         information
becomes      avb I lablc.        State    A agency   officials         informed        us,
however,       tlict   their     experience     had shown that           such adjust-
ments    generally        were made by local         office       employees         only   at
yearly     intervals.




                                              30
Self-mployed   r:nd unpaid
   fsmfly
32
        State I3 ag;:r?cy c~-!ployees did not ~:alce nz1y effort                   to
develop      an est im<ate :?or rlnempI\?ye:! r: co>eyed f17CL:r    :: t1y TTC!L^!.x3?     s

who    were disqualific.4        from receiv in;~ r~nempl. ,?~rl::cxt i,WEf its
for nonmonetary         reasons,      such as refu:;al      of suitab~.e      work,
This estimate      F which is required          by the prescrihcd         mcthod-
01 ogy 9 could have a significant             effect      on the covered            in-
dustry     rate whichJ      in turn,      is used to develop        estizntcs             of
nOnC0~sered unemp10yrwnt,              State A agwcq' employees            developed
an estimate      of workers        ~$20 are disq:~aIified        and who delay
filing     or never fiie       claims     for benefits,       on the basis             of
5 percent      of the nuxkr         of continued      claimants,

        Department       of tibor     employees     also made an analysis                 of
the methodology          used by State G and concl.uclecl that the mcth-
odol.o;:y did not result            in unemployment        rates    significantly
diffei    .::tl: Fi-om those 17hi.cl1 resulted        frcm us<ng the T--.-x-  l)..v.
ser..-bed xl'_'! ho.d@logy,       It should be noted,           however,      that the
analysris       CC bc&h mcthodcllogies         VC-~-e mE:.de on the has.?:: of
dalx wh Lc?i :.:ei-(3 available        at the Department's           natiorial        of-
fice    and :51?<cl-1 r:7ere for State and major labo-r areas,                       We be-
?.Lm-2 ti~at i.?::: effects         of deviations       from standard


                                             33
       The current practices advers ely af feet the         measurement
of unemployment 9 and consequently   EDA eZigi.biPity         detcrmina-
tions,   by

     --failing   to give each county within      the States  an
         equal opportunity   to qualify  for designation    by mea-
         suring une.mpLoyment conditions    over a standard   period
         of time and

     --cnusEn2    unwarranted      designations,    terminations,      or
        rcvkions     in maximum grant        rates as a resul.i:  of bas-
        ing annual uncm;?loymcnt        rate averages     on incomplete
        and preliminary      data,
      The wlmpl oyment rate estkates        developed    by the State
agency eiqloy-ee.s   inccrcorate actual.   covered    emplo:xient   data
for 9 mc?n&s of the year of estimate        and the last       3 months
Of the pricer yea-r.




                                  35
    state
    ---     P,




.




                 36
      ECPL uses the unern;:loyment      rate estimates        as a bc!:;is for
making Jeterminatio:ls       as to designntions     and terminations
and for mking      revisions    to the public     Q?orlCS Ill3XilillX.I~ j;l:‘tXlt
rates  during  the Eollo~~ing     year.




             All~l~ou~l~ i-n man> instances               the revised       rates    do not
  differ          significantly          from those first          reported,      the eicects
  on area eligibiliky                   determinations        could be substantial            be-
  cause of the firm-percentage                        standards      set by statute,          As
  notctl e,~'.Iicr,             the act requires          that an area esperieilce            at
  least         a b-percent          rate of unemployment            to qualify      for dcsig-
  nation. under the ~unemp1oyfrient criteria.                           A difference       of
  onsy ol-~e tenth              of 1 percent        in the ,unemployment          rate could
'irl.Fluence            clesip.aticns       ~ A county having           an unemploym=n~
  1 ate of 6 percent                 wculd be eligible,          whereas a county !:,q;qini;
 -ii11 lltlr:T~;jPO;~TJ~~t        rate of 5 a9 percent          xould be inel-i;;i      ble , fin
  EDh oEfl.ci,a,'% inforzied               us that      a determination         as to dedc-:sik;-
 L~I-ioil   of   an   ilrea   which    csperiences        a   drop    in   the   u~~:mpl_cJy   -
merit rate for 1 year of only one tenth       of 1 percent   is de-
ftil,rcd until  unclnpLo~~~2nt rate estimates   for the folIowing
year ilre received.

                                                     37
       Our review      of ~cvi,r,ions        in Lh? 11n,~~r~~lcj~~3~~ii:
                                                                     rates    for
calendar    years 195.5-67,        for      Sta?c ii s11owed, on i.Tle llasis     of
the revised     rates,    that:

        --   one county     ~couf_d not     love   qualified        for   designation,




        Example
        --L---      1

        In April     1966 a c:ity applied               for gran';: ~nct loan funds
to const-ruct       3 water supply            1i11e aild to ilnpi'ov? ~"~ater treat-
ment and storage          fac?'.I.ftics.         The city was Ioc;2r.ed in a
co~lrlty WhTc'h had been desT.gnated                  a rede~~~elol?iL::~I1t area in
January      1956 and which q-ualLfLc d for a 70-j~.2r~:!~nt maximum
grant     rate on the basis of 211 :3!liaLic? ‘1 average              ; :.‘:cm~!-oymeilt
rate for 1965 of 4-O percent.                    Ei.lA appro=zd    t?lC      projcl_ct in
May 1967 and agreed LO finnnc::                   the tot21     p-j rr:i: cr).;t of
$608,000      by a SO--percent           direct      p,rau:t of $3c;k,!~NI,         c~ 20-
percent      supplcr~ental.      grant cf $:i.%1,000, and n 3O--pe~-ccn~
loan of $183,000.
       The qua.liFii~l      arc.a.c; and maximum grant        rate listings
dated Ail.>7 30) I.',67 9 show a revised             1965 ratr: for tllc area
of !j o3 perccllt    r~;,l:her tl~arI  the 1965 rate of 10 percent            as
sholqn in the qur.'i :i.ficd a.rcas listing            dated July L ) I-?66 e
On the basi s of the revised             1965 rate9 tile arca WOlild         not
have qun?.iiied      f;:,r designation        as a redevclopmcnt        <irea and
would not have been eligible              for   pnrtfcipation      in   the  pro-
gram.



        In May 1967 EDA approved                a project      for the construc-
tion    of a sewage collection               system,     a seX,rage lift         station,
force mai.ns, and a primary                treatment       pond.      The total         esti-
mated cost of the project               was $275,200.            On the basis of a
1965 annual average unemployment                    rate of X7$4 percent,                 the
area in ~hioh tile project              was to be locntcd             ql:;rI t*cicd for
a maxim::m grant           rate of 80 percent.             EDA ;zgre-.d i-o finance
the sewj;e        collection      system with a 50-percei?t                 d< rc~ct grant
of $74,000,        a 30-percent        supplemental         g::ant ~>f $I74 :!:40, and
a ZO-percent         loan of $29,560;           a total     OF $].48,O?i).           EDA's
participation          in the sewage lift           statl.on,      force ma l!js, and
primary      treatment        pond included       a 50-peJ:ccnt         sup;“i:-:;;:zntal
grant     of $63,600 and a %O-percent                 loan of $25,440;             a total
of $89,040.          The Federal. Water Pollution                Control       Acfm-inis-
tration      agreed to finance           the remaining          costs with a
30-percent        grant      of $38,160,

         The qualified   areas and maximum grant  rate listings
 dated July 30, 1967, show a revised       1965 annual average un-
employment      rate for the area of 7.4 percent    rather   than the
'1965 rate of 17.4 percent      as shown in the qualified      areas
 listing    dated July 1, 1966.

         If the revised       rate had been known in May 1967 and
had bee11 used by EDA in determining             the percent-ace            of grant
participation,        the area would have qualified               for a SO-
percent B rather        than an 80-percent,      grant      rate.       On the ba-
sis of the revised          rate,  EDA grant    participation            fn the
sei;:al;z collection       system would h;lve been limited               to a di-
rect grant       of 50 percent     and in the sewage lift              station,
Eo-ccc mains,       and primary    treatment    pond would have been
limited      to a supplemental       grant   of 20 percent          rather      than


                                            39
50 percent.         ln total, EDA's grant participation                      in the
project      woul.d have been $02,600 less than that                      actually
agreed     to.

        In December 1968 the same applicant                       app?.ied to EDA for
addition:il     f lmds      t 0 Cover      pro j cct    cost OVCEXi11S,          EDA agreed
to mke nvai.lab1.e            additional         loan and grant       Euncis tot,~Iing
 $126,000 p mae,cs up of a TO-percent                   direct    grant    of $63,000,
a 20-percent        supplemental           grant      of $25,200,     and a 30-percent
loan    of $37,803,           The 70-percent           maximum grant       rate was
based on the annual average                    unemployment       rate for 1967 of
10 percent      sho:;rI1 in the qualified               areas and maximum grant
rate listings          dated September             I., 1968.    The qualified          areas
listing     dated October            1, l.963, shows a revised             anm.1~~1    aver-
age unemployment            rate for 1367 of 8.3 percent                 rather      than
LO percent.         If the revised             rate had been kno!~-n and had
been used by EDA, the area would have qualified                              for n GU-
percent,     rather       than a 70-percent,             maximtim grant        rate,      On
the basis      of the revised            rate,       EDA grant   particiFaticn
would have been about $24 9500 less than that actually                                 agreed
to,
.                --the               methodology        for   developing           estimates      9

                 --the   kibor  in-i?:ket data col.Pectior:                          and report        i.np, sys-
                     tems of the .TtZte EtgellCieS   p

                 -- t11e estilmating                   practices        of   the    State      agencies,      and

                 --the   review  and monitoring                          procc3urcs         of   the    Depart-
                     ment of Labor.




                  As indicated          by our review         of the agencies        of two
      States $ substanti.aL                differEPCC?S       cxi sted between the pre-
      scribed          es.;:imating      techniques         of the Department        and those
      used by both States.                      Differences      in practices       bctveen     the
      States were also evident o State officials                              view their      prac-
      ti.ccs as being necessary                      for administ-rative        expediency      or f
      in som instances,                  because of the lack of experienced                   em-
      ployW,s.            A greater        awareness        by the Department        of Labor of
    . the problems              experienced          by State agenci.es       in applying        the
      I):-:':,. rribed       estimat Ing techilicpcs            sholild  contribu-te      greatly
    . to;:;. cd improving             the situation         O

               T'hc Department                       of Labor also needs to improve        its re-
      [Tic\, r and monitorin?,                       procedures   of the estimating    practices
      (Ii-   i 112   v;l.:>i(jTL~S        St;J[-fs      to ensure uniformity     and consistency

                                                                   41
in the np1)'I.i cation     of the prescribed                c:;timnti.ng       techniques
and to r)rovide      technical   assistance                when ileeded.

       A procrzm       of household      surveys     in selected          areas on a
cycls" cal lxsis       may be Oi-le n;ctllod    by Tqhich the Department
coul-d cvaluatc        the empIoyr :ent and uncmpIo;-;nellt;            levels
yielcle.3    by the prc scribed      methodo?_og>~ , to asc:<rtain             whether
the leveJ_s 27%~ consistent          wil:ll local      conditi     on5 O We be-
J icvct ;:llet such a program        ?~011!.Cibe    eCOlfC~i".i ca. J.Jy feasible.
In;pJ.c~cllt:ation     of such a progriim        should he made, however,
only after         the survey proccd1lros       llave bern tcstcd            and found
to be. satisfactory         and after      the comparabi'lity           of the data
has been established.

         The States        showId be able to do much to improve the
quality     of tl~ estimates                by exercising           greater       care in
gathcri.nZ      and interpreting               the labor markc:i: d:t.La and by in-
cl-easing     their      efforts        to improve their               J.~l~:~r market data
collection       and reporting              systems.      We E‘CCC~,ITL;(;C       3 howevw B
that additional            funcli~~g of the Iocal             arc.3 u-,7irxl3loyment       csti-
rn:<ting Frogram by the Dep:~r:.;ll .cnt Of Labor ji1i.;;hf be ncce:;sary
to enable the State agencies                      to signific;!ii'i.'!.y        improve their
data collection            and reporting            systems.           As :iointed     out to
us by Dep~rt.ment            officials,          a factor     limit:        Ii5 the amcunt
nnd quality        of work which can be done un~.irl~- ~lw program                            is
the available          staff       and budg~lcary resource:,                  of the States e
Department       officials          noted thziz, histori.caJ.?y,                  State staff
rcsourccs       have averaged             less than one pcr.r;~~!~ for each State.




                                               42
        The Department           -i.ndic;ited       that     thy 3_oca1 i!re,l unemp?.oy-
 ment rat2 was the most wicM.y undzrstoc!ci                           and accepted            basis
 for measuring        economic distress                and t!xt       the present           system
 of preparing       data,    ill     spite      of some defjcicncics,                 minimized
 the cost and permitted                interarea          cornpa:-;Tbility       of data and
 the utilizat-Lon         of local         k~~owledge of arca conditions.                          The
 Department      acknowl.cdged J hot:cver)                 that    the     results        achieved
 through     this methodology              te~~dccl    to    decrease        in reliability
 as the :;izo of the area decreased                        and sll;:gested         thatg       for
 very srs;;l.l area s--those             with     a pop~rlntion          of l.ess than
 5,000---it     might be desirable                to use other           approaches          to
 achieve     the deg:ce of precision                   needed.

       The DctL>artment suggested            that these altc_rnai:?-vc           ap-
 proaches    could relate        to those currenkly              being c::;)! o;-cid in
 the Sttinford    &search        Institute      proj ect to devc:!.ok, 11~37 IIIC~-
 sures of underemployment              and ~lderu,ti1.i.?,;lti.on      of m;:~r;)ow~r
 or to i:1voLve    util  ization        of the T.im~.tcd househo1.d sur~.'~l;~y
 method.

        We have indicated        that the methodology           does not ade-
 quatel.y    measure the economic         djstress    of the snnl.1      meas.
 We agree wl.th the Department            that alternative        app-tyoaches may
 be needed to properly          measure economic        distress     in such areas.
 We question,      however,     whether     the unrel_iability      of the re-
 sults    obtained   using the methodology           should be identified
 only with areas having          populations       of 5,000 or less or also
 with areas having        larger    populations.

        With regard     to questions  raised   in this report  concern-
  ing the timeliness       of some of the data used to develop     arca
' unemployment    estimates   P the Department   stated  that:

          "Some of these problems --particularly                           those re-
          lated    to the figures     on base covered                     employment
                                                                                 ---
         data--have      long been recognized,     but                    are not sub-
         ject    to easy correction."

                                                    43
45
      We recommend       that   the       Secretary     of Labor   prcl>Tide   for:




       --Improvements       in the Department        's rdtric..~ and monitor-
          ing procedures      of .-i.hc ~lner;lp1.oymen't~-estinlating      prac-
          tic?s   of the ~.~ario-ds St.a?:.cs 9 to ensure uniformity              and
          cons?stency     in the nppl.icati.on         of the prc::s~-ibed      tech-
          niques    and to provide      technical        assistxlce    r;lhen needed,




       --High  priority   to be given to the efforts       bzing made
          to improve unempI.oyment    estimates    for 7koc;;l areas,   so
          that research   findings  ~1il.2. be converted   intro timely
          and meaningful_   action 7311ere warranted    e

       --Consideration       of the filldings    of this report          in the
          Department's      evaluation     of the ~-l~~lployr.lent-estilnating
          procedures.

                                      -    -    -   -


       The Assistant        Secretary    for Administration,        Department
of Labor,    advised       us that ---j-n l,inc with our recon~nlel!dation
for a review      of the procedures          and practices    of the State
employment     seclq::i-ty   agencies    and within      the conslxaints     of
budget   resources        and staffing     ceilings--the     Department    would

                                               46
        take    appro?rinte     steps to c'n s:we uniformi.ty             in the applica-
        ti.011  of prcsCr.~bed     t;~~cIn~1.o~~r~ic~n~-ees1l:i.n;niiri~techniques    and
        to improve       the accuracy      and comparability          of the result:;     ob-
        tained*




.

               With regard   to our recommendation    that high priority     be
        given to th e efforts      being made ts improve u~-~~mpI.oyme~~t es-
        timates    for local   arcas,   the Assistant Secretary stated
        that:

                          "The Department's         present    schedule         calls      for this
                          type of conversion          of research       findiligs        into the
                          development       of improved     area unempl.oy?l:::?t est:i--
                          mating procedures         by the end of this             fiscal_ ywr
                           [1971-j.    These improvements         :qil.1 takc7 acnowlt            of
                          the findings       of the CA0 report 9 as well as oiklr
                          research     studi.es    sponsored    by the i?eparimn~;t              ai>d
                          the affiliated        State employment          st~curity        ag~;:l~s?c:;
                          to improve      the soundness       of estimates            for 'r.ocal
                          areas."

               The Department's        response     also indicated        t11;it a comprc-
        hensiyze report       by consultants       at the University         of IIcIiston,
        in cooperation        with the Department's          stazf,    was schedl:l.ed
        for completion        by January     1971,     The report      is expected        to
        include    recommendations        for improvement         in the methodology
        for estimating        unemployment      in small     areas.      The Department's
        response     indicated     also that an additional           report     relating
        to the overall        system for States        and large     areas scheduled
        for completion        by July 1971 should         include    additional        im-
        provements      in small area procedures.
    I
              The Assistant        Secretary    advi.scd us that                         the Department
      ~~::?lld conl~;idcr our findings        in its evaluation                            of the unern-
    * 131i;:?.~:~nt c:-it- Lmating procedures     which currently                           was being
        c’!ji   !C!lJC:     [:c>cj   c




                          The actions       taken and proposed             by the     Department          are
        co;~::i.~t:c.nt          with      our recommendations.
                                                                47
48
        Under the current     praci_ices)  detelmii.,:tions        of olj g5.-
 bility   base d on ic>w incc>ine are restricted         t:o thn:;e ;~~‘cas of
 the cour~rtry IThich qhaiify     at the time of 1.1~ cei;:<crs of pop-
 Ul   a’iiOI1   o   These      ci~‘~~er:ai;?3ti’~ns        are   frozen   ult;    I. t13z i‘2:<:‘11
 census         incox         rcs';alts   are made a-v;liial)le-+-a period   of I.0
 years.           llle!.'efore      9 rcgardlcss  of improveme~~ts     in inccm2
 1evcJ.s or the             general      cconqic           conditions     of     i-he   designakd
 areas   5 the areas continue        to reta,in      their    eli,glbiEity       staixs
 and possibh)       share in the limited          progrt=in funds inte;;dlzd
 expressly     for economicalLy        distressEd       areas.       Areas guaran-
 teed long-term       elzgibility      are in sharp contrast               to those
 areas designated        on the basis of high unemployment                   that are
 subjected     to annual       rate determinations         and reviews         to as-
 certain     whether    continued    eligibility        is xl'arranted.

            Althou&      the system serves to benefit                  those arcas
' whi ch q-u2 ii fy at the time the C~;ISUS                is   taken,       it  creates
  an inequity         for those areas x&i&,            although          not quz 3 '.L'yi.!Jg
 .at the tGxe of the census,               later   experience            economic       down-
  turns With resultant            effects     on the areas'            income l~~\~cls.
  Eligibility        determinations        do not rest        solely        on m~r!r;ures
  of    i.ncomc ~ If an area's          economic      dislsress        is in the form
  of undercmpl-oflent,           howey,fer, it is doubtful               that the other
  criteria--       such as uncmpIoyinent--       will      prove to be sui.Zable

                                                      49
    substitutes      for measuring, the distressed     condition     of the
    area.      As noted on page 20 S widespread     poverty      may exist  in
    some areas even thou;_;h the local       unemployment      rates mzy be
    relatiwly      low.

              Concern with using census data for EDA purposes              also
    arises      i&en onz con;;idcrs    the degree of p~eci.scncss         and re-
    liability       of ~hc income ::aca deveI.opcd s 'The dectlnnia'l~        cen-
.   su .ses are desTgi;~sd prim:~~i.l,.y to obtain     infomati.on      on the
    del:logr':;qKc    and soci.al. chn~actcris   -tics of   the  po~e~lation
                                                                    *           and
    not to provide       median Ec~ni.1.y income statistics.          The Income
    data therefore       are by-prcduci-.    data that are gathered        from
    only a sampling       of the popnla?i.on.




                                                                                                       kior~:~t by hict:         true median
                                                              Interval      rCt!lin                  fomilv , inco1.c could be iover
                                      Estimate              Khich trur: nxdian                           or hirJ,c,r     t!>an c:;tii.:atcd
                Number of            1960 median                family    incoxle                            ineclj an ic v,iivA___-
                                                                                                     -----_____.                   i1ir.c~~ __
    County       fanilirs
                 ---               f_amily        _
                                              j nccme      could     fall    (note a.1                     Ifiwer                    Ilip,llcr

         A             1,833           $2,254               $2,099       to      $2,40sb                 $155                       $155
                                                             1,961       to       2,56?c                  293                        313
         B         27,092               4,272                4,227       to       4,318b                    45                        46
                                                             4,181       to       4,363'                    91                         91

    aTruc median  family   income               is   the   median       family       irxome     that      would       be obtained      in a
     census of all iamilies.

    b
        Chances are about GR out of lG0 that                    median           family    incomes        obtained        from   a com-
        plete  census Vould fall within this                   range.


    CChances     are     about     95 out of 100 tl~:~t median                   family    incoxes         obtained       frcm   a cc..!-
     plcte     census     :ir~ld    fall within this   ranc,e.
.




              The     nzcd   to     deveLop          r:urrent         estimati~s           of   incoxe        1?as
      been    recognized          for         some   tixz        by   both         ARA   a!ld   EDA.        Tile     Pub-
      lic Works and Econom:ic Dcvelqxent      Act specifically         reguircs
      that an annzlrll revi.e~ be made of designations-       -including
      those designarions     based on income.

              The Rurcau of the Census collects             annual data on mc-
      dian family     income in the interim          bct~srccil dcc::?rxii;: 1 ccn-
      suscs througll     saii:ple surveys       of about !jC?,OC;O flous:i12r: Yc;s in
      the Uni-;:eJ States,         The survey iS krl0~~Tl aS the C1L-c7'<      il.K pop-
      ulation    survey.      SiIlCe these data are coSS.eci.~:~l for oi'ly a
      small samp1.e of the Kation 9 however,             the sta Listic:;      arc
      meaningful     only as they relate          to the Nation      as a ~~;lo,l.e or
      to reg-ional    scgme11ts of lzhc Nation,          7.ke sai.;:p1c: is no t
      large enough to develop          statistics      on a St.~~tc or local        basis.

           In the %y 1365 hearings                                referred    to earlier,                  the Dep-
      uty Administrator,     J&4, with                          regard     to adjusting                  census in-
      come data,     stated:

              "We hope within    a few years to have a system by
              which 1272 can determine   ;-he changes in annual     in-
              ccme 9 county by county,     -L!lrough the use of In-
              ternal   Revenue figures.     WC do not now have it,
              but we are experimenting      with it,"

             An EDA o-[Eicial       informed        us that as of Hay 1970                                    EDR
    . did not have a system ~cxzh as that described                         by the                          Deputy
      Administrator.      ELM officials             later        informed   us that                            it
      iK1u1.d be very  costly       to lmilfd       nn    income       data system                          on    the
      basi:: of Internal      Revc~xe       Servj ce data and that                 the                      results
      would be of c;:.-cstionable         reliability              because not all                             income
       As noted earlier,      entries    for census data frequently
are based on a person's        Irtemoz7.    OX data,     however, are
based mainly    on   ~cco~~ds  of   bus?ne.ss  2nd governmcn t which
sho:q disbursements      made to individ1LaI-s    rather    than on data
furnished    by individuals.




         In considering   the effect    on the geogra'Fhica1      disper-
sion of EDA areas of using per capita            income,   :QC noted tht
mob,t of the areas \&ich         wouXd qualify   under tile loi:-income
criteria     wxuld fall   within    the 10 State.s which the census
identified      as having   the I_owest median family      income,




                                    53
        AS indicated      above, use of median famLly           income or per
capita    income wou!d yield          substantially     the same results
(or   the same period       of time.       Assuming that this       relation-
ship re;.iains     constant    fro;‘rl year to year,      it can bc concluded
that chznnges in 1cvel.s of per capita              income will     approxi-
mate changes in ilieriit-02      family    income D

         Using OBE income data for the yecars 1953 and 1967, we
dete-zzined       wh?t changes in income had taken place bEtween
tl1e ts:o years and t;rhat effects           such changes would have had
on area designations             based on income had the changes been
recognized.           The results     showed that 71 of the 160 counties,
nationwide,         whLch would have qualified        on the basis of 1959
per capita        income criteria       would have lost their    income des-
ignation      eligibility        and 37 additional    counties would have
qualified       for eligibility.          On a basis other than income,
27 of the 71 counties             and 25 of the 37 counties    would have,
qualified       for eligibility.          The table  below summarizes   the
results.

                                                     Number of counties
                                                     ~-_-I___---
Qualified    on 1959 per     capita   income                          160
Terminated    (1959-67)                                    71
Designated    (1959-67)                                   -37          34

     Qualified    on 1957 per     capita
       income                                                        x25
                                                                     --




                                      54
             EDA procedures          do not rcrluirc:         rcvi sions in the ma2):imum
    grant     rate incoiw        c-t-1.terj.3    to rcco;;.nizc      changes in income
    lCLVC?LS   0    The income lt?17els          cui-rently      used as a bosh-s for giv-
    ing up to 30 percent               additional        grwt     funds arc the sang as
    those used when thy program Y/AS first                        cstablishcd.            JkrLr;;=
    periods       of i11FLatii;p.      and rilpid      econorilj c3 grOW.h, SIlch ;1.C;Were
    experienced         during     1965-70,        it is doubtful         that income lc.vels
    can rem,zin unchanged              and still       stand as valid          critc>ria       for
    disbursing         pro;;ram funds * For example,                 data resulting            from
    the current         population         surveY conducted          by the Bureau of the
    Census showed that the rhedian income of households                                 in the
    United       States     was $8,359 in 1969 compnrcd with $5,660 shoT*rn
    by the 1960 census o The rntj~o of poor persons                              to the total
    population         was 12 percent          in 1969 compared with 22 pcrcc?nt
    in 1959.

            Ide noted daring        our review        that th2 OffIce       of Econor::ic
    Opportunity      issued uniform          income c:uidelinps         for ~L~~>liii::+tion
    in all of its programs             where family         income is used to i:fztcr-
    mine prcgram       cbigihility,          These guidelines        ,;re bCl.SCd  Oil   pOV-
    erty thresholds         derived     from a defiaition          of poverty     dc>vc:l-
    oped Lor stattstical            purposes        by the Social    SecIlrj.ty   Adr!:i.llis--
    tration     in 1954.        The Office       of   Economic Cpportunity        sdjusts
    the income guidelines            periodically         for changes in cGnr,umzr
    prices.

    CONCLUSIONS

            The decennial      income data used by EDA as a basis             for
  eligibility         and grant rate determinations         do not    provide     a
  current        measure of economic     distress,      It is also question-
  able whether         the census income data accurately           represent      the
  actual       levels    of income for some small and rural           areas.      EC-
  forts       should be made, therefore,         to dcvclop   a means for im-
' proving        income data.
.           If the     practicalities   of the situa'cion     warrant  a Con-
    clusion    that     income data cannot be developed        with greater
    p1 ecision    in    measurenicnt  or currentness,      we believe  thnt  the
    Sccrctary     of    Commerce should   so advise     the Congress   and
      Furthermore              WC   belALer7e         that    EDA should institute        tcch-
iliquas  for    periodically                      3djusirin,:   the   max. imuii g-ant rate in-
come criteria          to      bri:-lg,         i:lx,m lno3-e j n l.I.ne with current     colldi-
tions.




                                                        56
         The Assistant         Secretary      directed          our atteni-ion     to tl:e
fact tl1at the %t,l ic NorJrs and Economic Developiic~lt                           Acst         -
qu!'.rc.d EDA to use only median family                       income data.        We be! LCW
that cog~lIzanee           sh~l d be given to the fact that tile per
cap i t3 illCOiCC       series   developed       by OBE did not become avaLl-
abILe until       1.968    md %-Ls not     an    alJcernative         sollrce  of    ‘;.nccJlne

data at the time t:hat the A?'& and EDR legislation                            was drafted
and consider-cd          by the Congress,



       We recommend         that    the    Secretary       of Commerce provide             for:

        --A review       and study,   in cooperation     with the Department
           of Labor,      of the problems     associated    with developing
           curr~~~t    urwmploymcnt     and income data.

        --Considxation     of the feasibility       of using,  as a basis
           for area d esignation,     the more current     per czpita   in-
           come data devcl.opcd     by OJ3E instead   of, or as a suppl.e-
           merit to, the median family       income developed   as part
           of d:+ccennial censuses.

                                               57
            The Assistant     Sec::t?tary    did not conmxlt        specifically      on
    the need to institute          techniques     for periodically          adjusting
    the rnaxim:.~m grant rate income criteri,i,             Since tl3ese criteria
    arc used by EDA for givi-11;: up to 30 percent                in add?.tionaX
    grant    funds,   we contirlue     to bel.icyre that there        is a need to
    period-i c~l-1~ adjust      the criter5a      to recognize       changes in in-
.   come levels *
.




    .




        s9
Mr.    Ijcnry       Eschwege
Associate         Director
Civil    Division
General       Accounting        Office
Washington,           D, C.     20548

Dear    Mr,         Eschwege:

This  is in reply   2.0 your     letter     of Septembc:r         22, 1970,      rc:q\~~.~ting
comments      on a draft     report     entitled      “Need     to Irnp~*o\~e    Data
Used   as  a Basis   ior Providing           Fed~r:il     Assistance       to   Ecol-!wlj?i-
tally  Depressed      Areas,      Dcpartmcnts           of Commcrcc           and L;:bor,      ”

We have rcviewcd       the comments               of the Economic      Dcv~;lopr~-,el!t
Administration    and believe    that           they are appropriately         responsive
to the matlers   discussed    in the           report.

Sincerely       yours3




A&chm         ent




                                                     61
NOV 16 1970

Mr. wnry       Eschn~egc
Associate      Director,      Citli.1  Di.vision
United    States      General     Accouni.ilrg             Office
Washington,       D. C.       2054s




Representatives           of !73*4 and GAL! have met on two occasions                       to
disc\lss     the drnT t report.            1~s a l~csult      of these     meetings       ,
ce r ta i n c ha i3 6 cs have been made j-1) L;hc draft             report       by the
GAO repr~~~tn1,ati.v~~          to refl.ect,        amoi-lg other   things,        EDA'~
concern      r!itIr   income    ds-ta cur~reni;ly       llr;ed in designating           rc-
development         areas    and the fact         that     EDA is presently         cxplor-
ing the use of othc~-             income     data    as a basis     for     designation.




        1)     reviews      and study          i;he    grob,lcm     of   developing   current
               income       data,

        2)     consicic?r       the   fc:;isibilitg  01 using    more current               per
               capita       income      dab developed     by the Office      of
               Business    Ikoncriics        (013X) instead    of decennial               census
               Yigu ros for     dcsip;natiou       of areas    on the basis               of
               low income,      and

        3)


GAO a.I. s o
in-depth




                                                      62
                                                                                                                                  _, ..




  With    respccl;    to the use by ;?A of median          family  income      data
   (1959    data    as shown in the 1960 Census        of the Populat;on)          ,
  we call      your   attention      to the fact  that   section   ,103 (n)(2)       of
  the EDA Act presently            rcquircs   EDA to use only     metlian    L'amily
  income     data    and does not allow       DA to use pc'l* r,:aI)j tn incollie
  data   as sug;;;ested        by GAO.

  As the draft              report      acl;nowlt:dges          , 1960 Census                  cjata arc the:
  most recent            data      available          for     all      areas        (cou?i$,ie:;)           in the
  United       Sta-tcs.           The dx+aPt report                also       notes       -ti?a L to da tc
  nothing        has been accompi ishcd                     to update             the I!)GO Census                    mc>dinn
  family       income         data.       While       this      is     coi*rcct         ant1 ous          opcl*altl i>g
  procedures            continue        to be based on the 1960 Cetlsus                                   dat,a,         K:C
  have explored               the possibility               of developing                 and u I ilj zing;
  other      sources          of income         data.         The draft             report         recognizes
  that     in 1967 hmA initiated                      a program            with       the Office               of Busi-
  ness Economics                of the Department                  of Commerce               for      the develop-
  ment of per capita                  income        estimates            on a county               basis.              PC!r
  capita       income         data    for many,            but not all              counties,             became
                      .
  avallablc          in 1968.           EDA has been the primary                             single         contributor
  to this        effort         and through           Fiscal         Year 1971 has obligated
  $645,000         to OBE to secure                 these       data.           A joint          EDA/ORE work
  group      is wor!cing            on the devcl.opment                  of per capita                  income           data.
  However,         there        remains       the difficult                 task      of eval.uai:ing                  the
. quality        of tllc estimates                and their            ccmparal,iIity                 with       tho:;ct
  c~c~velopcd        II?JQII Census         p3.-ocedurcs           which        yield        eutima t2r-r; oi'
  mc>dian fa1~1-1~ incow                 s Much work               remains          before         this        ~~cfil;;l;,-
* clc>veloped        :.:c~r-tcs can be considered                       as an acceptable                    s\I~:< I ; I 11; (2
  -i‘or the fn:'~ 3-y income                statistics             currently            used       to     re1'3    ('i' i
  i! l*ca   ecol10.:1-i   c   distrcns         .




                                                             03
h'ls.   Ilcnry    Fschwege


We would      also    poiilt    to an earlier          attempt       in 1963 - 1064 by
the Area Redcvelopmc~~t            Administration              (AM) , JCDA':~ pr~!d~c:~s:-:o~~
agency,     to secure        and utilize        Internal        Revcnuc       data          to clc-
vclop   income      estimates.          This    effort       was lal';:cl     y ~~ns;\:cce:-zs      1‘~?1

due to the difficulties               in securing          and p~!'occ~:isi.ng          ihc illfo-I:-
mati.on   as well       as the many problems               invulvctl      in resoI.ving
the techni.cal        questions       associated         with     th5 s data       source .
During    this     period     RR,4 also      favored       the p~~opo:;ccl c!uincjuenni::l
census    in order        to secure      more recent          income      estimates.

Our recent          investigations          suggest         that     the relative              income
rankings        of counties          remain      fairly       stable       and certailily            ch:~nf;es
arc no-t as dynamic               as are rates          0-f ui7erny,loyiii":i-i;.           Ncvert~hel     c11-;:-:
                                                                                                                 )
we agree        in principle           that    it would         lx desirzl,l        e to have more
recent       income      information        on a rcgu'l.ar           l3asi.s.       UnfOstun;?-te1y)
the costs         of securing          such   information            by    duplica        t-ins Ccnsl!:;
procedures          and techniques           (e.g.      surveys)        ;ippc~r~r;       to be pro-
hibit.ive.          I1owevc 1' ) by utilizing             other      data     sources,           such as
from ODE, we hope to devel.op                      reasonably          ncc:iratc         income      esti-
mates      to serve        as mcasurcs        of area economic                clistrcss.

EDA is currently         studying           its    Iegis1ntivc         authori     ;:y wi.th            a
view    toward    changing       the criteria             for  tile tlcsi~;n:~i:iol1            of      l:e-
development       areas.      II 0 v: e v e r    we arc not in a pesi i;i.on at                         t1,i.s
time    to make definitive              rcc&nmendations            for    S~lCll  ClIall~~:i-:s
pending      completion     of this           study.

Sincerely,



Robert    A. Podesta
Assistant     Secretary
 for    Economic    Development




                                                     64
We have 31~0 taken cognizance,     in OLIT I*espwse,     of the   cmmnts     of S,hu
two State empboy9nes-t secrti2;y ngeacfes where the GAO conducted CL COW-
prehensive   audit Of RIYX unemp~o~ent      Cs~-ix~~t~rlg procedures     in COXl?CtfGiI
with the prepzratiow    of this report.     Copies or thcsc S-hte mpiies            hsve
been furnished    to members of your staff,

You my be certain    that we shzll--in    Hnc ~5%. your recommendations
and gtithin the constraints  of budget resources      and staffin  ceilP1Gs--
take n~~Yopr.hte  steps to t'asSGUe? -!x-~iforx;fty in 3Fplicaiion of prescribed



.
APPENDIX II

estimatirq   technicpx      znd to .$-prove the accurxy      md conrp1-~I'i::il.ity
of the resuits    0Staincd."       183 shall alrn,  as you rccoimeridc~i,   cmsidcr
the comrmts     in the GAO report       "as part. of the Depnrt~ynt's     eva.l:mtTon
of the umm$!qmnt         estimtir~      procedux!s"  which is cuzeni;Q      bcilp
                                                                                    >
conducted.

Sincerely,
                   .\                I-
         _-                         i. ,
              ,I!’ , (    ti’ L .

llm’x:        !JIlmTs
Assistant               Secretary          for   Adrxin:'Lstration

Ehxlosure




                                                               66
         APPENDIX   11




.




    67
-   _   .   _   . .   I   .,   .,   .!.-
APPENDIX II
APPEPSIIIXII