Federal Paperwork: General Purpose Statistics and Research Surveys of
Businesses (Letter Report, 09/20/1999, GAO/GGD-99-169).

Pursuant to a congressional request, GAO provided information on the
federal paperwork burden, focusing on the: (1) paperwork burden
associated with federal agencies' general purpose statistics and
research (GPS/R) surveys that are directed towards businesses; (2)
nature, use, and burden of selected GPS/R surveys; and (3) agencies'
efforts to reduce the burden associated with the selected surveys.

GAO noted that: (1) GPS/R surveys account for an extremely small
proportion of the estimated paperwork burden that federal agencies
impose on businesses; (2) nevertheless, federal agencies estimated that
businesses spent 12.6 million hours responding to 180 of these surveys
in fiscal year (FY) 1998; (3) GAO estimates that all businesses'
financial costs to complete these 180 surveys were between about $219
million and $305 million; (4) two federal agencies--the Bureau of the
Census and the Bureau of Labor Statistics (BLS)--accounted for more than
half of the estimated governmentwide GPS/R burden-hour total; (5) within
these two agencies, 14 large surveys, each with at least 100,000
estimated burden hours, accounted for nearly two-thirds of the
governmentwide GPS/R burden-hour estimate; (6) the characteristics of
the 14 large Census and BLS GPS/R surveys directed at businesses varied
widely; (7) six of these surveys were part of the Economic Census, which
is taken every 5 years, covers virtually all businesses, and serves to
update many other business statistics; (8) other large GPS/R surveys
were more frequent, more limited in the number of businesses surveyed,
covered only certain types of businesses, or addressed more specialized
topics; (9) survey topics ranged from changes in the selling prices of
goods and services to the characteristics of businesses owned by
minorities and women; (10) the statutes requiring or authorizing the
surveys generally provide the agencies with a substantial degree of
discretion to determine the scope, substance, and, in some cases,
frequency of the surveys; (11) all of the surveys provide economic
information that is widely used by federal, state, and local
governments, businesses, and the general public; (12) although the
agencies indicated that responses to 5 of the 14 surveys were voluntary,
some of these voluntary surveys are mandatory in certain states under
the laws of those states; (13) Census and BLS estimated that these 14
surveys imposed $179 million in financial costs on businesses in FY
1998; and (14) both BLS and Census have taken steps to minimize or
reduce the burden associated with these 14 GPS/R surveys, including: (a)
designing and administering the survey instruments to minimize burden on
the survey respondents; (b) using information technology to enable
businesses to respond to surveys electronically; and (c) using
administrative records in other agencies as a substitute for or a
supplement to surveying the businesses directly.

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

 REPORTNUM:  GGD-99-169
     TITLE:  Federal Paperwork: General Purpose Statistics and Research
	     Surveys of Businesses
      DATE:  09/20/1999
   SUBJECT:  Data collection
	     Reporting requirements
	     Statistical data
	     Statutory law
	     Surveys
	     Projections

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  This text was extracted from a PDF file.        **
** Delineations within the text indicating chapter titles,      **
** headings, and bullets have not been preserved, and in some   **
** cases heading text has been incorrectly merged into          **
** body text in the adjacent column.  Graphic images have       **
** not been reproduced, but figure captions are included.       **
** Tables are included, but column deliniations have not been   **
** preserved.                                                   **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************

    United States General Accounting Office GAO               Report
    to the Honorable Sue W. Kelly Chairwoman, Regulatory Reform and
    Paperwork Subcommittee, House Committee on Small Business
    September 1999    FEDERAL PAPERWORK General Purpose Statistics and
    Research Surveys of Businesses GAO/GGD-99-169 United States
    General Accounting Office GAO                 Washington, D.C.
    20548 General Government Division B-281700 September 20, 1999 The
    Honorable Sue W. Kelly Chairwoman, Regulatory Reform and Paperwork
    Reduction Subcommittee Committee on Small Business House of
    Representatives Dear Madam Chairwoman: The information collected
    by federal agencies is an integral tool for accomplishing their
    missions. Agencies collect information for a variety of purposes,
    including to verify regulatory and tax compliance, determine
    eligibility for benefits, evaluate the effectiveness of federal
    programs, develop economic statistics, and conduct research.
    However, these information collections also impose a burden on
    those individuals, businesses, and others asked or required to
    provide the information. Federal paperwork is commonly measured in
    terms of "burden hours," and federal agencies estimated that they
    imposed nearly 7 billion burden hours of paperwork in fiscal year
    1998.1 Previous studies have indicated that, among businesses,
    small businesses can be disproportionately burdened by federal
    regulatory requirements.2 Some small businesses have expressed
    concerns about the number of surveys federal agencies ask them to
    complete, the amount of time it takes them to complete the
    surveys, and the compulsory nature of some of the surveys. To help
    address these concerns, you asked us to describe (1) the paperwork
    burden associated with federal agencies' general purpose
    statistics and research (GPS/R) surveys that are directed towards
    businesses; (2) the nature, use, and burden of selected GPS/R
    surveys; and (3) the agencies' efforts to reduce the burden
    associated with the selected surveys. GPS/R surveys account for an
    extremely small proportion (0.3 percent) of Results in Brief
    the estimated paperwork burden that federal agencies impose on
    businesses. Nevertheless, federal agencies estimated that
    businesses spent 12.6 million hours responding to 180 of these
    surveys in fiscal year 1998. 1The burden-hour figures used in this
    report are, unless otherwise indicated, for information
    collections approved by the Office of Management and Budget (OMB)
    at end of fiscal year 1998 (September 30, 1998). We refer to those
    collections in this report as burden imposed in fiscal year 1998.
    2Thomas D. Hopkins, A Survey of Regulatory Burdens, June 1995
    report to the Small Business Administration. Page 1
    GAO/GGD-99-169 Federal Paperwork B-281700 We estimate that all
    businesses' financial costs to complete these 180 surveys were
    between about $219 million and $305 million.  However, the costs
    imposed on an individual business to complete even several of
    these surveys may be quite small. Businesses are frequently
    required to respond to the surveys and may face possible criminal
    or civil sanctions for failure to do so. Two federal agencies-the
    Census Bureau in the Department of Commerce and the Bureau of
    Labor Statistics (BLS) in the Department of Labor (DOL)-accounted
    for more than half of the estimated governmentwide GPS/R burden-
    hour total. Within these 2 agencies, 14 large surveys, each with
    at least 100,000 estimated burden hours, accounted for nearly two-
    thirds of the governmentwide GPS/R burden- hour estimate. The
    characteristics of the 14 large Census and BLS GPS/R surveys
    directed at businesses varied widely. Six of these surveys were
    part of the Economic Census, which is taken every 5 years, covers
    virtually all businesses, and serves to update many other business
    statistics.  Other large GPS/R surveys were more frequent, more
    limited in the number of businesses surveyed, covered only certain
    types of businesses, or addressed more specialized topics. Survey
    topics ranged from changes in the selling prices of goods and
    services to the characteristics of businesses owned by minorities
    and women. The agencies' estimates of the time needed to complete
    each survey varied from a few minutes to 6 hours. The estimated
    number of respondents ranged from about 27,000 for one survey to
    more than 2 million in others. The statutes requiring or
    authorizing the surveys generally provide the agencies with a
    substantial degree of discretion to determine the scope,
    substance, and, in some cases, frequency of the surveys. All of
    the surveys provide economic information that is widely used by
    federal, state, and local governments; businesses; and the general
    public. Although the agencies indicated that responses to 5 of the
    14 surveys were voluntary, some of these "voluntary" surveys are
    mandatory in certain states under the laws of those states. Census
    and BLS estimated that these 14 surveys imposed $179 million in
    financial costs on businesses in fiscal year 1998. Both BLS and
    Census have taken steps to minimize or reduce the burden
    associated with these 14 GPS/R surveys, including (1) designing
    and administering the survey instruments to minimize burden on the
    survey respondents, (2) using information technology to enable
    businesses to respond to surveys electronically, and (3) using
    administrative records in other agencies as a substitute for or a
    supplement to surveying the businesses directly. However, both
    agencies said that statutory confidentiality restrictions are an
    obstacle to greater burden reduction Page 2
    GAO/GGD-99-169 Federal Paperwork B-281700 through data sharing.
    The executive branch has proposed and Congress has introduced
    legislation to address this issue. The Paperwork Reduction Act of
    1995 (PRA) requires federal agencies to Background    minimize the
    paperwork burden that they impose on individuals, small
    businesses, and others through their collections of information.
    At the same time, the PRA recognizes that information is a
    critical resource enabling the federal government to perform its
    most basic functions. The original PRA in 1980 established the
    Office of Information and Regulatory Affairs (OIRA) within OMB to
    provide central agency leadership and oversight of governmentwide
    efforts to reduce unnecessary paperwork and improve the management
    of information resources. The 1995 PRA gave OIRA and executive
    branch agencies significant new responsibilities,3 but continued
    the requirement that agencies obtain OIRA approval before
    collecting information from the public. It also required OIRA to
    set annual burden-reduction goals totaling at least 25 percent for
    the 3-year period ending on September 30, 1998. As part of its
    review process, OIRA requires that agencies submit a completed OMB
    Form 83-I along with other materials to justify a proposed
    information collection. The Form 83-I provides the agency's
    estimate of the number of burden hours for the collection, which
    is based on (1) the estimated amount of time needed by each
    respondent to provide the requested information, (2) the number of
    respondents for the collection, and (3) the frequency of the
    collection. The Form 83-I also indicates the primary purpose of
    the collection (e.g., general purpose statistics, research, or
    regulatory or compliance monitoring); the primary target of the
    collection (e.g., businesses, individuals, or state and local
    governments); and whether responding to the survey is voluntary or
    mandatory. When OIRA completes its review, information from the
    Form 83-I is to be entered into a database maintained by the
    Regulatory Information Service Center (RISC).4 OIRA uses the RISC
    database and other information to develop its annual Information
    Collection Budget, which contains governmentwide data on the
    number of approved information collections each year and the
    agencies' burden-hour estimates as of September 30 of each year.
    3For a discussion of some of those responsibilities, see
    Regulatory Management: Implementation of Selected OMB
    Responsibilities Under the Paperwork Reduction Act (GAO/GGD-98-
    120, July 9, 1998). 4RISC is part of the General Services
    Administration but works closely with OMB to provide the
    President, Congress, and the public with information on federal
    regulations. RISC maintains a database that includes information
    on all regulatory actions and all information collection review
    actions by OIRA. Page 3
    GAO/GGD-99-169 Federal Paperwork B-281700 The PRA also makes OIRA
    responsible for governmentwide statistical policy and coordination
    of the federal statistical system. The PRA requires OIRA to ensure
    the efficiency and effectiveness of the system and the integrity,
    objectivity, impartiality, utility, and confidentiality of data
    collected for statistical purposes. The PRA also requires OIRA to
    develop and oversee the implementation of governmentwide guidance
    concerning statistical collection procedures and methods.
    Additionally, the PRA mandated the establishment of an Interagency
    Council on Statistical Policy, composed of the heads of major
    statistical programs and representatives of other statistical
    agencies, to advise and assist OIRA on statistical issues. We have
    issued several reports and testimonies on the implementation of
    Previous Reports and    the PRA, noting that the burden reduction
    goals envisioned in it were not Testimonies             being
    met.5 Agencies frequently pointed to new and existing statutory
    requirements for information collection as impediments to
    achieving the goals. We also noted a variety of problems with the
    "burden-hour" measures that are used as indicators of the
    paperwork burden and said that it is important to recognize the
    difference between the government's "measured" paperwork burden
    that is reflected by the number of burden hours an agency reports
    and the "real" burden that is felt by the public. Users of
    paperwork burden-hour estimates need to proceed with great caution
    because it is often unclear exactly what the burden estimates
    represent and what factors can cause changes in the totals.
    Nevertheless, these measures are the best indicators of paperwork
    burden available, and they can be useful as long as their
    limitations are considered. Although one of the key purposes of
    the PRA is to reduce paperwork and minimize burden on the public
    and businesses, the PRA also requires agencies to balance burden
    reduction goals against agencies' needs for information critical
    to performing their missions. For the federal statistical system,
    after years of retrenchment, there has been renewed national
    recognition in the 1990's of the importance of federal statistics
    and a commitment to improve the quality of the system. We have
    issued several reports and testimonies on major initiatives to
    improve the quality of federal economic statistics.6 In 1990 and
    1991, the Council of Economic 5Paperwork Reduction: Burden
    Reduction Goal Unlikely To Be Met (GAO/T-GGD/RCED-96-186, June 5,
    1996); Paperwork Reduction: Governmentwide Goals Unlikely To Be
    Met (GAO/T-GGD-97-114, June 4, 1997); and Paperwork Reduction Act:
    Implementation at IRS (GAO/GGD-99-4, Nov. 16, 1998). 6For example,
    Economic Statistics: Status Report on the Initiative to Improve
    Economic Statistics (GAO/GGD-95-98, July 7, 1995); Federal
    Statistics: Principal Statistical Agencies' Missions and Funding
    (GAO/GGD-96-107, July 1, 1996); Statistical Agencies: Statutory
    Requirements Affecting Government Policies and Programs (GAO/GGD-
    96-106, July 17, 1996); Statistical Agencies: Consolidation and
    Page 4
    GAO/GGD-99-169 Federal Paperwork B-281700 Advisors, chaired by
    Michael Boskin, reviewed national economic statistics and
    announced a group of recommendations to improve the quality of
    economic statistics. The recommendations from the two "Boskin
    reports," formally known as the Economics Statistics Initiative,
    have subsequently been a key part of the federal government's long
    range statistical plan described in the Statistical Programs of
    the United States Government: Fiscal Year 1999 that is produced
    annually by OMB. Recommended improvements in those reports would
    require more frequent collection of some data, revised measures of
    key economic indicators, and the development of new measures for
    emerging industries and services. However, collecting more
    information to improve these economic statistics may conflict with
    agencies' efforts to reduce paperwork burden. The Boskin reports
    and OMB also recommended increased data sharing among statistical
    agencies to reduce burden. OMB developed a legislative proposal to
    permit greater data sharing while preserving the confidentiality
    protections in existing law. Relatedly, data-sharing legislation
    has been introduced in each recent session of Congress. We have
    supported these proposals to revise existing law to permit greater
    data sharing in our prior reports and testimonies. We used RISC
    data to describe the paperwork burden associated with Scope and
    federal agencies' GPS/R surveys that were directed toward
    businesses as Methodology    of September 30, 1998 (the end of
    fiscal year date that OMB uses in its annual Information
    Collection Budget). RISC provided data on agencies' burden-hour
    estimates for all information collections, those directed
    primarily at businesses, and business surveys whose primary
    purpose was to develop general purpose statistics or conduct
    research. Of the GPS/R surveys on business, RISC provided data by
    agency, by whether responses were voluntary or mandatory, and by
    individual collection within certain agencies with large burden-
    hour totals. Using Census and BLS estimates of wage rates from 14
    large information collections, we developed estimates of the
    annual governmentwide costs to businesses of responding to all
    GPS/R surveys. To describe the characteristics of the 14 large
    GPS/R surveys within Census and BLS, we obtained information that
    agencies submitted on the Form 83-I and in supplementary
    information that described the need for the surveys, how the data
    would be used, the agencies' burden-hour Quality Issues (GAO/T-
    GGD-97-78, April 9, 1997); and Statistical Agencies: Proposed
    Consolidation and Data Sharing Legislation (GAO/T-GGD-98-91, March
    26, 1998). Page 5
    GAO/GGD-99-169 Federal Paperwork B-281700 estimates, and the
    agencies' efforts to reduce paperwork burden, particularly on
    small businesses. We also reviewed the relevant federal statutory
    provisions for the 14 surveys to determine the extent to which the
    agencies had discretion in administering these surveys. We
    discussed each of the 14 surveys with responsible agency officials
    at BLS and Census (e.g., to determine whether these surveys were
    mandatory or voluntary within certain states), and discussed
    related issues (e.g., how national economic statistics are used)
    with OIRA's chief statistician and staff from the statistics
    branch. We obtained information on agencies' efforts to reduce the
    burden associated with the 14 GPS/R surveys of businesses through
    discussions with BLS, Census, and OIRA officials, and by reviewing
    documentary evidence. We also asked these officials about barriers
    to agencies sharing the data collected through these information
    collections and initiatives to eliminate these barriers to data
    sharing. Our review focused on GPS/R information collections in
    effect at the end of fiscal year 1998.  Because the burden
    associated with the economic censuses is primarily imposed only
    once every 5 years (which included fiscal year 1998), the burden-
    hour estimates for the information collections would be less in
    other years. Also, it is important to keep in mind that the
    agencies' burden-hour estimates are estimates; it is unclear how
    accurately they reflect the real burden felt by the targets of the
    information collections. Some of the information in this report is
    drawn from the RISC database. We did not verify the accuracy of
    this database, but we did obtain updated and corrected information
    for certain elements from BLS, Census, and other agencies. We did
    not evaluate the effectiveness of the agencies' burden reduction
    efforts or assess whether other reductions could have been made.
    Neither did we independently determine how the information
    collected through the GPS/R surveys is used.  We conducted this
    review between November 1998 and August 1999 in the Washington,
    D.C., headquarters offices of Census, BLS, and OIRA in accordance
    with generally accepted government auditing standards. We provided
    a draft of this report to officials responsible for GPS/R surveys
    policy and administration in Census, BLS, and OMB for comment.
    Their responses are presented in the Agency Comments and Our
    Evaluation section at the end of this report. Page 6
    GAO/GGD-99-169 Federal Paperwork B-281700 The estimated paperwork
    burden on businesses from GPS/R surveys in Most GPS/R Survey
    fiscal year 1998 was a very small portion of the estimated burden
    on Burden on Businesses businesses from all types of collections.
    Nevertheless, the 180 GPS/R surveys approved by OIRA represented a
    substantial aggregate burden on Was Imposed by
    those businesses that had to complete the surveys-an estimated
    12.6 Census and BLS                        million burden hours,
    or between $219 million and $305 million in business' compliance
    costs. Fourteen surveys administered by Census and BLS accounted
    for nearly two-thirds of the governmentwide GPS/R burden-hour
    estimate on businesses. As of September 30, 1998, OIRA had
    authorized federal agencies to collect GPS/R Surveys Are Small
    information from businesses, individuals, and others through a
    total of Portion of Businesses'                6,755 information
    collections. The agencies estimated that the paperwork Paperwork
    Burden                      burden associated with all of these
    information collections was nearly 7 billion burden hours. As
    figure 1 shows, roughly two-thirds (4.6 billion) of these burden
    hours were from information collections that were primarily
    directed toward businesses. Figure 1:  Burden Hours by Type of
    Respondent Source: RISC database. As figure 2 shows, most of the
    4.6 billion burden hours on businesses were associated with
    information collections primarily intended for regulatory or
    compliance reporting. About three-tenths of 1 percent of the 4.6
    billion Page 7                                        GAO/GGD-99-
    169 Federal Paperwork B-281700 burden hours on businesses (12.6
    million burden hours) was from 180 information collections whose
    primary purpose was to develop general purpose statistics or
    research. Figure 2:  Burden Hours on Businesses by Type of
    Information Collection Source: RISC database. Most GPS/R Burden
    Hours Are              For more than three-quarters (9.8 million)
    of the 12.6 million burden hours From Surveys With Mandatory
    associated with GPS/R surveys primarily directed toward
    businesses, the Responses
    businesses were required to respond to the surveys. Although
    failure to respond to the surveys could result in possible federal
    criminal or civil sanctions, officials at OIRA, BLS, and Census
    told us that they were unaware of any such sanctions being imposed
    for failure to respond to a mandatory survey. In nearly all of the
    remaining surveys (about 3 million burden hours), the respondents
    were under no federal obligation to respond.7 However, as will be
    discussed later in this report, some of these "voluntary" surveys
    were made mandatory under the laws of certain states. 7Surveys
    accounting for less than one-half of 1 percent of the
    governmentwide GPS/R burden-hour estimate were coded in the RISC
    database as voluntary but required for benefits. Page 8
    GAO/GGD-99-169 Federal Paperwork B-281700 As shown in table 1,
    more than 80 percent of the burden associated with Census and BLS
    Imposed                    GPS/R surveys on businesses was from
    collections administered by two Most GPS/R Survey Burden
    departments-the Department of Commerce (about 66 percent) and DOL
    on Businesses                             (about 15 percent).
    Within the Department of Commerce, 7.8 million of the 8.3 million
    GPS/R burden hours were associated with surveys administered by
    Census. Similarly, BLS accounted for almost all of DOL's 1.9
    million GPS/R burden hours and surveys on businesses. Table 1:
    Number of Total Burden Hours
    Burden hours        Percent of total     Number of by Agency
    Department/Agency                    (millions)         burden
    hours         surveys Commerce
    8.3                  66%             61 Bureau of the Census
    (7.8)                (62)             (41) Labor
    1.9                  15              15 Bureau of Labor Statistics
    (1.9)                (15)             (14) Treasury
    0.6                       5          24 Health and Human Services
    0.5                       4           18 Energy
    0.5                       4          12 Other departments and
    agencies              0.8                       6          50
    Total                                     12.6
    100             180 Source: RISC database. Within Census and BLS,
    the paperwork burden is similarly concentrated in a few surveys.
    Eight Census surveys and six BLS surveys account for about 85
    percent of each bureau's GPS/R survey burden on businesses.
    Together, these 14 surveys accounted for nearly two-thirds (8.4
    million hours) of the governmentwide GPS/R burden-hour estimate.
    Although paperwork burden has traditionally been reported in terms
    of GPS/R Surveys Impose                      burden hours, federal
    agencies' information collection requirements Substantial
    Aggregate Labor impose real financial costs on individuals and
    organizations providing the Costs on Businesses
    information. For businesses, those costs include the (1) financial
    resources expended to cover capital, operating, and maintenance
    costs for equipment, supplies, and nonlabor services (e.g.,
    information technology systems and monitoring equipment) and (2)
    wages of employees responding to the collection. BLS and Census
    officials indicated that employee wage costs account for virtually
    all of GPS/R survey respondents' financial costs. In recent years,
    OMB has required agencies to convert burden hours into dollar
    costs and report these estimates in their submissions to OMB. To
    do so, agencies must estimate a wage rate that would be applicable
    to the burden hours associated with the collection, including
    overhead and fringe benefit costs related to the employee's time
    (e.g., health insurance and Page 9
    GAO/GGD-99-169 Federal Paperwork B-281700 contributions to a
    retirement plan).8 As will be discussed in detail later in this
    report (see table 6), Census and BLS estimated respondents' wage
    rates for the 14 large GPS/R surveys range from $9.59 per hour to
    $30.00 per hour. Using these agency estimates, we calculated that
    businesses' cost for completing all 14 of these surveys would be
    about $179 million. The estimated wage rates in these 14 surveys
    can serve as a rough guide for estimating the costs to businesses
    of responding to all GPS/R surveys. As table 2 shows, if the wage
    rates associated with the other 166 GPS/R surveys of businesses
    (4.2 million burden hours) were as low as $9.59 per hour, the cost
    associated with completing those surveys would be about $40
    million. However, if the wage rates were as high as $30 per burden
    hour, the businesses' cost for the surveys would be about $126
    million. By adding these figures to the $179 million costs of the
    14 largest surveys, we estimated that businesses' cost associated
    with all 180 GPS/R surveys was between $219 million and $305
    million. Table 2: Estimated Labor Costs of
    Estimated range of labor GPS/R Surveys of Businesses
    Burden                                   costs (millions) Number
    of                      hours Source
    surveys (millions) Lower bound Upper bound Large BLS and Census
    surveys                                                14
    8.4                      $179                        $179 Other
    surveys
    166                   4.2                       $40a
    $126b Total
    180                 12.6                       $219
    $305 aThis estimate assumes labor costs of $9.59 per hour. bThis
    estimate assumes labor costs of $30.00 per hour. Source: The
    number of surveys and burden hours are from the RISC database.
    Cost estimates for the l4 large BLS and Census surveys are from
    the agencies' Form 83-I packages. (See table 6.) Upper- and lower-
    bound labor cost estimates are based on wage rate ranges from BLS
    and Census surveys. The characteristics of the 14 large Census and
    BLS GPS/R surveys varied Census and BLS GPS/R widely. Some of the
    surveys applied to all businesses, but others were Surveys Vary
    but                     limited to certain sectors or types of
    businesses. The statutes requiring or authorizing the surveys
    generally provide the agencies with a substantial Collectively
    Impose                  degree of discretion to determine the
    scope, substance, and in some cases, Substantial Burden on
    frequency of the surveys. Survey topics ranged from changes in the
    selling Businesses                           prices of goods and
    services to the characteristics of businesses owned by minorities
    and women. The agencies' estimates of the time needed to complete
    each survey varied from a few minutes to 3.5 hours. Also, the
    frequency of data collection differs across the surveys.  Some of
    the surveys are conducted monthly, but others are administered
    only once 8In the fiscal year 1999 Information Collection Budget,
    OIRA said "the proper accounting of such overhead and fringe
    benefits has been an area of inconsistency and weakness in the
    present system of burden estimation." Page 10
    GAO/GGD-99-169 Federal Paperwork B-281700 every 5 years. The
    number of respondents ranges from about 27,000 for one survey to
    more than 2 million in others. All of the surveys provide economic
    information that is widely used by federal, state, and local
    governments; businesses; and the general public. Although the
    agencies indicated that responses to 5 of the 14 surveys were
    voluntary, some of these surveys are mandatory in certain states.
    Together, Census and BLS estimate that these GPS/R surveys of
    businesses account for about two-thirds of the governmentwide
    burden-hour estimate for those surveys, or $179 million in
    aggregate costs. As tables 3 and 4 show, the scope and subject
    matter of the 14 Census and Surveys' Subject Matter and BLS
    surveys of businesses that constitute the bulk of the agencies'
    GPS/R Scope Differ                      burden-hour estimates
    varied substantially. Some of the surveys collected information
    about all businesses (e.g., the Producer Price Index), while
    others focused on particular sectors (e.g., the various economic
    censuses) or ownership groups (e.g., the Surveys of Minority- and
    Women-Owned Business Enterprises). The topics covered by these
    surveys were equally diverse, ranging from changes in businesses'
    selling prices to comprehensive information on establishments'
    employment, payroll, and geographic location. Page 11
    GAO/GGD-99-169 Federal Paperwork B-281700 Table 3:  Eight Surveys
    Accounted for Most of Census' GPS/R Burden-Hour Estimate in Fiscal
    Year 1998 Survey title
    Description 1. Economic Census of the Wholesale Trade Sector
    Every 5 years, this survey is to be conducted on the universe of
    businesses in the wholesale trade industry. It is to collect facts
    about the structure and functioning of the economy and features
    unique to this industry, such as sales by commodity line and class
    of customer, the number of establishments, sales, payroll,
    employment, and geographic location. 2. Economic Census of
    Manufactures                             Every 5 years, this
    survey is to be conducted on the universe of businesses in the
    manufacturing industry. It is to collect facts about the structure
    and functioning of the economy and features unique to this
    industry, such as materials consumed, inventories held, the number
    of establishments, payroll, employment, and geographic location.
    3. Economic Census of Utilities; Transportation;
    Every 5 years, this survey is to be conducted on the universe of
    businesses Information; Finance and Insurance; and Real Estate,
    in all of the listed sectors. It is to collect facts about the
    structure and Rental, and Leasing Sectors
    functioning of the economy and features unique to these
    industries, such as amounts and sources of revenues/receipts,
    insurance benefits paid to policyholders, purchased
    transportation, the number of establishments, payroll, employment,
    and geographic location. 4. Economic Census of Retail Trade and
    Food Services           Every 5 years, this survey is to be
    conducted on the universe of businesses in the retail trade and
    food services industry. It is to collect facts about the structure
    and functioning of the economy and features unique to this
    industry, such as sales by merchandise line and customer class,
    the number of establishments, receipts/revenues, payroll,
    employment, and geographic location. 5. Economic Census of
    Professional, Scientific, and            Every 5 years, this
    survey is to be conducted on the universe of businesses Technical
    Services; Management, Support, Waste               in all of the
    listed industries. It is to collect facts about the structure and
    Management, and Remediation Services; Educational
    functioning of the economy and features unique to these
    industries, such Services; Health and Social Assistance; Arts,
    as the number of establishments, receipts/revenues, payroll,
    employment, Entertainment, and Recreation; and Other Services,
    and geographic locations. except Public Administration Sectors 6.
    Economic Census of Transportation/Commodity Flow            Every
    5 years, this survey is to be conducted on a sample of businesses
    in Survey
    the commodity flow industry.  It is to collect facts about the
    structure and functioning of the economy and features unique to
    this industry, such as the value of shipments, commodities
    shipped, modes of transportation used, number of establishments,
    receipts/revenues, payroll, employment, and geographic location.
    7. Survey of Minority-Owned Business Enterprises
    Every 5 years, this joint survey is to be conducted as part of the
    economic (SMOBE) and Survey of Women-Owned Business
    census program. Unlike the other economic censuses, it surveys
    only a Enterprises (SWOBE)
    sample of the businesses to obtain information on businesses owned
    by minorities and women. 8. Shipper's Export Declaration (SED)
    Program                  This survey is to collect information on
    individual shipments exported from the United States. Shippers are
    to complete a survey form for almost every shipment, and this
    information forms the basis for the official export statistics
    compiled by Census.a The form itself serves as the export control
    document used by Customs, the Bureau of Export Administration, and
    the State Department. a As discussed later in this report, exports
    between the United States and Canada do not require a SED form to
    be completed. The United States and Canada have an agreement that
    the United States will rely on Canadian import data for its export
    statistics. Source: Supplementary documentation to the Form 83-I
    for each survey. Page 12
    GAO/GGD-99-169 Federal Paperwork B-281700 Table 4:  Six Surveys
    Accounted for Most of BLS' GPS/R Burden-Hour Estimate in Fiscal
    Year 1998 Survey title
    Description 1. Current Employment Survey
    This survey is to collect data each month on employment, hours,
    and earnings from a sample of nonfarm establishments (including
    governments). This is a joint effort between BLS and the State
    Employment Security Agencies (SESA). 2. Occupational Employment
    Statistics                     This annual survey is to collect
    occupational employment data on workers in a sample of businesses
    by industry. It is a joint effort between BLS, DOL's Employment
    and Training Administration, and the SESAs, which collect the
    data. 3. Multiple Worksite Report
    This quarterly survey is to collect employment and wage data from
    the universe of establishments of employers who operate more than
    one establishment. This is a joint effort between BLS and the
    SESAs. 4. Survey of Occupational Injuries and Illnesses
    This annual survey is to collect information from a sample of
    businesses on the characteristics of the most serious of nonfatal
    cases (i.e., those involving lost work time) and the traits of
    workers sustaining such injuries and illnesses. Participating
    state agencies collect and process the data from the employers and
    provide it to BLS. (In other states, BLS collects the data
    directly from employers.) 5. Producer Price Index
    This monthly survey, formerly known as the Wholesale Price Index,
    of a sample of businesses is to measure average changes in selling
    prices received by domestic producers for their goods and
    services. 6. Annual Refiling Survey
    This annual survey, formerly known as the Standard Industrial
    Classification (SIC) form, is to collect and update information on
    a sample of employer's worksites, products, or services covering
    one-third of the businesses each year. It is a joint effort
    between BLS and the SESAs. Source: BLS Handbook of Methods (April
    1997) and supplementary documentation to the Form 83-I for each
    survey. All 14 BLS and Census GPS/R surveys were based on
    statutory provisions Underlying Statutes Give
    that either require or authorize the agencies to collect and/or
    report certain BLS and Census Survey
    information. However, all of the underlying statutes give the
    agencies at Design Discretion                             least
    some discretion regarding how the surveys can be designed and
    administered.  For example, the agency may be able to decide
    whether to collect information from all covered businesses or a
    sample of those businesses, or determine the number and subject
    matter of questions asked in the surveys. BLS' annual Occupational
    Injuries and Illnesses Survey is based, in part, on provisions in
    the Occupational Safety and Health Act of 1970 (29 U.S.C. 673).
    The act requires the Secretary of Labor to "develop and maintain
    an effective program of collection, compilation, and analysis of
    occupational safety and health statistics," and to "compile
    accurate statistics on work injuries and illnesses." However, the
    act does not specify the frequency with which the survey should be
    conducted, the number of businesses to be contacted, or the
    particular data elements in the survey. The Occupational Injuries
    and Illnesses Survey is also based on the general requirements in
    29 U.S.C. 2, which says that BLS "shall collect, collate, and
    report at least once each year, or oftener if necessary, full and
    complete statistics on the conditions of labor and the products
    and Page 13
    GAO/GGD-99-169 Federal Paperwork B-281700 distribution of the
    products of the same." Therefore, although BLS must collect the
    information on workplace injuries and illnesses (and other
    "conditions of labor") at least once each year, it could do so
    more frequently if it determined that more frequent collection was
    necessary. However, BLS has not done so. Other statutes underlying
    the surveys are somewhat more specific, but still provide broad
    discretion in survey design and administration. For example, the
    Census Bureau's SED Survey is based on 13 U.S.C. 301, which
    requires the Secretary of Commerce to submit statistics on exports
    from the United States by country and product to certain
    congressional committees on quarterly and cumulative bases, and to
    report monthly and cumulative international trade balances for the
    United States. The statute authorizes the Secretary to collect
    information from all persons exporting from the United States, and
    says the Secretary "shall compile and publish such information
    pertaining to exports . . . as he deems necessary or appropriate."
    Therefore, although required to report the export information by
    country quarterly, the Secretary has substantial discretion
    regarding the specific content of the survey. Also, 13 U.S.C. 131
    says that the Secretary of Commerce "shall take, compile, and
    publish censuses of manufactures, of mineral industries, and of
    other businesses, including the distributive trades, service
    establishments, and transportation . . ." The statute also
    requires that the censuses be taken every fifth year after 1968,
    and that the data "relate to the year immediately preceding the
    taking thereof." Therefore, the Census Bureau has no discretion
    regarding the frequency of the economic censuses, the sectors on
    which information is to be collected, or the year for which the
    data applies. On the other hand, the agency has broad discretion
    regarding what specific information to collect, and whether to
    contact all businesses or a sample of businesses within each
    specified sector. According to information provided by BLS,
    Census, and OMB, the data GPS/R Survey Data Are     collected in
    the 14 large GPS/R surveys are used for many different Used for
    Many Purposes    purposes by both the public and private sectors.
    For example, the Department of Transportation, the Federal
    Emergency Management Agency, the Army Corps of Engineers, and
    other agencies use data from the Census Bureau's Commodity Flow
    Survey to plan transportation infrastructure. Many localities also
    used the data from this survey in responding to requirements in
    the Intermodal Surface Transportation Efficiency Act of 1991.
    Similarly, the data collected in BLS' monthly Producer Price Index
    is used by the Federal Reserve Board's Open Market Page 14
    GAO/GGD-99-169 Federal Paperwork B-281700 Committee in making
    decisions on monetary policy and in developing and evaluating
    monetary and fiscal measures for evaluating the general business
    environment. These survey data are also used in the private sector
    to evaluate escalating contract costs, inventory accounting
    processes, and the condition of markets. In addition, industry,
    academia, legal professionals, the news media, and the general
    public use the data to evaluate businesses, markets, and
    characteristics of firms; prepare business plans; make economic
    decisions; and conduct economic research. Other examples of how
    GPS/R survey data are used include the following. *  The Small
    Business Administration and the Minority Business Development
    Agency use data from the SMOBE/SWOBE Surveys when allocating
    resources for their business assistance programs. Others use data
    from these surveys to evaluate the concentration of minority-owned
    businesses in particular industries or geographic areas. *  DOL's
    Occupational Safety and Health Administration (OSHA) and its
    state-level counterparts use the data from the Occupational
    Injuries and Illnesses Survey to evaluate the effectiveness of
    federal and state programs to improve workplace safety and to
    prioritize federal and state funds. *  BLS uses the data collected
    in the Multiple Worksite Report to develop the Business
    Establishment List, which is used as the sampling frame for
    establishment surveys, such as the Producer Price Index. *
    Various private enterprises, government agencies, and organized
    labor use data from the Current Employment Survey to calculate
    such indexes as the Composite Index of Leading Economic
    Indicators. *  Information from the Annual Refiling Survey is used
    in developing BLS' Business Establishment List to help ensure that
    the correct SIC code, name, and physical location are assigned to
    the employers' worksites for unemployment insurance records and
    labor market information purposes in a particular state. State and
    local officials also use these data for industrial coding
    classification. In addition, state employment security agencies
    use the data to administer their unemployment insurance programs.
*  The SED Survey serves as the official record of export
    transactions for the United States. In addition, Customs and the
    Bureau of Export Administration use the data to enforce export
    regulations. The Census Bureau indicated that survey recipients
    were required to Response to Most Large      respond to all eight
    of its large GPS/R surveys because the underlying GPS/R Surveys
    Was           statutes require Census to collect and/or report the
    information. In Mandatory                   addition, Census
    officials said that they tend to get higher response rates Page 15
    GAO/GGD-99-169 Federal Paperwork B-281700 on surveys with
    mandatory response requirements than on surveys with voluntary
    responses. Census officials said that the mandatory response
    requirement to their surveys is traceable to 13 U.S.C. 224, which
    says [w]hoever, being the owner, official, agent, person in
    charge, or assistant to the person in charge, of any company,
    business, institution, establishment, religious body, or
    organization of any nature whatsoever, neglects or refuses, when
    requested by the Secretary or other authorized officer or employee
    of the Department of Commerce or bureau or agency thereof, to
    answer completely and correctly to the best of his knowledge all
    questions relating to his company, business, institution,
    establishment, religious body, or other organization, or to
    records or statistics in his official custody, contained on any
    census or other schedule or questionnaire prepared and submitted
    to him under the authority of this title, shall be fined not more
    than $500; and if he willfully gives a false answer to any such
    question, he shall be fined not more than $10,000. The survey
    recipients' obligation to respond to these mandatory Census GPS/R
    surveys was communicated on the survey forms themselves and, in
    some cases, in cover letters sent to the recipients. For example,
    the Economic Census of Retail Trade and Food Services noted in
    bolded capital letters that "your response is required by law." At
    the bottom of the form, also in bolded capital letters were the
    words "penalty for failure to report." The SED form required
    shippers to certify that they understand "civil and criminal
    penalties, including forfeiture and sale, may be imposed for
    making false or fraudulent statements herein, failing to provide
    the requested information or for violation of U. S. laws on
    exportation." Although the statute allows for the imposition of
    criminal penalties for failure to respond to a survey, Census
    officials told us that they have never pursued any prosecutions.
    In contrast to the Census surveys, response to five of the six BLS
    GPS/R surveys of businesses was voluntary at the federal level.
    BLS officials said that they prefer to use voluntary surveys, and
    that response rates are generally higher when a survey is
    voluntary. They said one of their surveys, the Occupational
    Injuries and Illnesses Survey, is mandatory because of a specific
    federal statutory provision. Some "voluntary" surveys were
    Although the RISC database indicated that responses to five of the
    six mandatory in certain states      large BLS GPS/R surveys were
    voluntary at the federal level, BLS officials said that some
    states and territories made four of these surveys mandatory during
    fiscal year 1998. Page 16
    GAO/GGD-99-169 Federal Paperwork B-281700 *  Response to the
    Multiple Worksite Report Survey was mandatory in 21 states, Puerto
    Rico, and the Virgin Islands. *  Response to the Annual Refiling
    Survey was mandatory in 19 states, Puerto Rico, and the Virgin
    Islands. *  Response to the Current Employment Survey was
    mandatory in four states and Puerto Rico. *  Response to the
    Occupational Employment Survey was mandatory in North Carolina.
    The officials said that survey recipients' obligation to respond
    varied between levels of government because BLS and the states and
    territories jointly administered these surveys, and states can add
    requirements and make the respondents' obligation mandatory under
    state law. Although the BLS survey instruments usually indicated
    that responses were voluntary,9 at least some of the cover letters
    that the states include with the surveys inform the recipients of
    their mandatory nature. For example, on the first page of the
    Multiple Worksite Report Survey, BLS stated that "[y]our voluntary
    cooperation is needed to make the results of this survey
    comprehensive, accurate, and timely." However, the cover letter
    added by the State of California stated that "completion of this
    form is required by section 320 and 320.5 of the California
    Unemployment Insurance (UI) Code." As noted earlier in this
    report, agencies' annual burden-hour estimates for Large GPS/R
    Surveys       an information collection are a function of three
    elements: (1) the time Impose Substantial        needed to
    complete the information collection (including reviewing the
    Aggregate Paperwork       instructions, searching existing data
    sources, and gathering and Burden On Businesses      maintaining
    the needed data); (2) the number of respondents to the collection;
    and (3) the frequency with which the information is collected from
    the respondents within the specified year. Table 5 shows that,
    although the agencies estimated that each of the 14 large GPS/R
    surveys directed at businesses would require more than 100,000
    burden hours of paperwork, the surveys varied substantially in
    terms of these three components. For example, the agencies
    estimated that some of the surveys would take only a few minutes
    to complete (e.g., the Annual Refiling Survey and the Current
    Employment Survey), but others were expected to take hours (e.g.,
    the Economic Census of Manufactures). Most of the surveys were
    conducted only once during fiscal year 1998, but others were
    administered quarterly (such as the Multiple Worksite Report) or
    monthly (such as the Current Employment Survey and the Producer
    Price Index). Several of the surveys that were administered during
    1998 are not 9BLS indicated in the Current Employment Survey that
    responses were mandatory in four states. Page 17
    GAO/GGD-99-169 Federal Paperwork B-281700 scheduled to be repeated
    for 5 years. The number of survey respondents varied from as few
    as 27,000 (the Producer Price Index) to more than 2 million (the
    Annual Refiling Survey and the SMOBE/SWOBE Survey). In most cases,
    the surveys that the agencies expected to take the longest to
    complete were administered less frequently and to fewer
    respondents than surveys expected to take only a few minutes to
    complete. Table 5:  Burden-Hour Estimates for Selected BLS and
    Census GPS/R Surveys of Businesses in Fiscal Year 1998 Est. survey
    Est. number of                Est. annual completion time
    respondents             burden hours Survey
    (minutes)             Survey frequency                (thousands)
    (thousands) Bureau of Labor Statistics Occupational Injuries and
    Illnesses Survey                         54
    Annually                     230                          207
    Multiple Worksite Report                                    10 to
    60                        Quarterly                     113
    167 Current Employment Survey                                    2
    to 15                         Monthly                      437
    596 Annual Refiling Survey                                       5
    to 15                         Annually                   2,086
    203 Occupational Employment Survey
    45                        Annually                     337
    251 Producer Price Index Initial Visit
    120                           Once                        (6)
    (13) Monthly Repricing                                     5 to 30
    Monthly                        27                         378
    Total
    3,230                     1,802 Bureau of the Census SED
    3 to 11                 Each shipment                        159
    1,316 Economic Census of Manufactures
    120 to 360            Once every 5 years
    210                          762 Economic Census of Wholesale
    Trade Sector                          70           Once every 5
    years                        540                          634
    Economic Census of Transportation/
    Quarterly during a 1-year Commodity Flow Survey
    120          period every 5 years                       100
    805 Economic Census of Retail Trade
    46           Once every 5 years                      1,291
    993 Economic Census of Professionals
    37           Once every 5 years                      1,443
    900 Economic Census of Utilities
    74           Once every 5 years                        625
    766 SMOBE/SWOBE
    10           Once every 5 years                      2,500
    417 Total
    6,868                     6,593 Total for BLS and Census
    10,098                     8,395 Note 1. The estimated annual
    burden hour totals reported by the agencies do not always equal
    the product of the estimated completion time, frequency, and
    number of respondents.  Average completion time is reported as a
    range for some surveys and anticipated response rate may affect
    the estimated number of respondents. Note 2. The Annual Refiling
    Survey is required at 3-year intervals.  BLS annually surveys a
    different third of the universe so that at the end of the 3-year
    period, it has administered this survey to the universe. Unlike
    all the other surveys we reviewed, SED surveys are completed by
    the shippers at the time of the export transaction and sent to
    Census. All of the other surveys are administered to the
    businesses by the agencies on a monthly, quarterly, or annual
    basis. The economic censuses are required every 5 years, which
    happened to fall during the time frame of our review. The Economic
    Census of Transportation/Commodity Flow Survey is required during
    each quarter in the year that the survey is taken. Source: BLS and
    Census. Page 18
    GAO/GGD-99-169 Federal Paperwork B-281700 Federal paperwork can
    also be measured in terms of financial costs Agencies Indicate
    Surveys    imposed on respondents. For businesses, those costs can
    be roughly Impose Substantial           divided into the cost of
    the employees' time to respond to the survey and Aggregate Costs
    other costs, including capital and operating expenses (e.g., the
    purchase of new hardware or software systems to produce the
    required information). Both BLS and Census indicated that the 14
    GPS/R surveys imposed no such "other" costs. BLS and Census
    estimated the cost of employees' time to respond to an information
    collection by multiplying the number of burden hours associated
    with an information collection by the wage rate(s) of the types of
    employees needed to produce the information. According to OMB,
    agencies' estimates of the wage rates for respondents should
    include overhead and fringe benefit costs associated with the
    employee's time (e.g., health insurance and contributions to a
    retirement plan). As shown in table 6, BLS' estimates of the
    respondents' hourly rates for its six largest GPS/R surveys ranged
    from $9.59 to $19.91. Census estimated the respondents' hourly
    wage rates for its eight largest GPS/R surveys at between $12.73
    and $30.00. Multiplying the agencies' estimates of the
    respondents' wage rates times their burden-hour estimates for the
    14 GPS/R surveys of businesses yields a total cost to businesses
    of about $179 million in fiscal year 1998. Page 19
    GAO/GGD-99-169 Federal Paperwork B-281700 Table 6:  Estimated
    Costs Associated
    Respondents'          Annual burden       Total costs to with
    Responding to Surveys in Fiscal
    wage rate per                 hours         businesses Year 1998
    Survey title                                         hour
    (thousands)           (millions) Bureau of Labor Statistics
    Occupational Injuries and Illnesses                $12.50
    207                  $   3 Multiple Worksite Report
    13.72                  167                        2 Current
    Employment Survey                             9.59
    596                        6 Annual Refiling Survey
    13.04                  203                        3 Occupational
    Employment Survey                       19.03                  251
    5 Producer Price Index                                 19.91
    391                        8 Total for BLS
    1,815                  $ 27 Bureau of the Census SED
    a               1,316                  $ 14 Economic Census of
    Manufactures                      12.93                  762
    10 Economic Census of Wholesale Trade Sector
    30.00                  634                       19 Economic
    Census of Transportation/Commodity Flow Survey
    30.00                  805                       24 Economic
    Census of Retail Trade                      30.00
    993                       30 Economic Census of Professionals
    30.00                  900                       27 Economic
    Census of Utilities                         30.00
    766                       23 SMOBE/SWOBE
    12.73                  417                        5 Total for
    Census
    6,593                  $152 Total (BLS and Census)
    8,408                  $179 aThe estimated cost per hour to
    complete a SED Survey ranges from $10 for a paper submission to
    $15 for an electronic submission. Of the 1.3 million burden hours,
    Census estimated that about 1.1 million burden hours were for
    paper responses and about 0.2 million hours were for electronic
    responses. Therefore, the burden-hour cost for the paper responses
    would be $10.7 million and for the electronic responses $3.7
    million. Source: BLS and Census data. It is important to remember
    that these are aggregate burden-hour and cost estimates for all
    businesses in fiscal year 1998. The cost borne by each business
    that receives one of these surveys is significantly less. For
    example, recipients of the Occupational Injuries and Illnesses
    Survey take an estimated 54 minutes (.9 hour) each year filling
    out the form, with an associated wage rate of $12.50 per hour.
    Therefore, the estimated cost to each recipient of the survey is
    $11.25 per year. Although a single business may receive more than
    one of these surveys each year, its aggregate burden may still not
    be substantial. For example, even if a single business received
    all six of the BLS GPS/R surveys and the Census Bureau's
    SMOBE/SWOBE Survey in 1 year, the estimated labor cost to complete
    all of the surveys would be less than $134. Page 20
    GAO/GGD-99-169 Federal Paperwork B-281700 Both BLS and Census have
    taken a number of steps to minimize or reduce Census and BLS Have
    the burden associated with the 14 large GPS/R surveys primarily
    directed Attempted to Minimize toward businesses. These steps
    include (1) designing and administering the survey instruments to
    minimize burden on the survey respondents, (2) Burden of GPS/R
    using various types of information technology to enable businesses
    to Surveys on Businesses respond to surveys electronically, and
    (3) using information found in administrative records in other
    agencies as a substitute or a supplement to surveying the
    businesses directly. However, both agencies said that statutory
    confidentiality restrictions are an obstacle to greater data
    sharing, which could result in further burden reduction. Census
    and BLS have attempted to design and administer their largest
    Survey Design and               GPS/R surveys to minimize the
    burden they impose on businesses, Administration
    particularly small businesses. For example, both agencies have
    developed versions of some of their surveys for small businesses
    with fewer or less complicated questions than versions directed to
    larger businesses. In the Economic Census of Wholesale Trade,
    Census sends establishments with fewer than four paid employees a
    simplified version of the survey form that requests less data than
    the standard form that is sent to larger employers. Similarly,
    businesses with only one owner receive shorter SMOBE and SWOBE
    Surveys than businesses with more than one owner. BLS and Census
    have also used sampling techniques to limit the number of
    businesses to whom surveys are directed. The agencies also use
    special sampling procedures in some surveys that reduce the
    chances of a smaller business being selected compared to larger
    businesses, or of the same businesses repeatedly being selected to
    respond to the survey. For example, in its Occupational Injuries
    and Illnesses Survey, BLS employs a stratified random sampling
    approach in selecting survey recipients.10 As a result of using
    this approach, less than 4 percent of all businesses actually
    receive a survey. Also, small employers (10 or fewer employees)
    are prenotified in the Occupational Injuries and Illnesses Survey
    that they have been selected for the survey and will need to
    maintain records. In the Annual Refiling Survey, BLS annually
    surveys one-third of the universe of the eligible businesses. As a
    result, an individual business can only be selected for the survey
    once every 3 years. Similarly, to ensure that small businesses are
    not unduly burdened by the Producer Price Index Survey, BLS
    changes the sample of companies surveyed every 4 to 7 years. 10In
    stratified random sampling, the researcher ensures that
    appropriate numbers of elements are drawn from homogeneous subsets
    of the population. Page 21
    GAO/GGD-99-169 Federal Paperwork B-281700 The agencies have also
    attempted to minimize the surveys' burden on business by making it
    as easy as possible for the business to respond to the surveys.
    For example, BLS' Occupational Injuries and Illnesses Survey is
    designed to allow employers who have no worker injuries or
    illnesses (many of whom are small employers) to simply check a "no
    injuries/illnesses" box on the survey, provide the number of
    employees and hours worked, and return the survey in a postage
    paid envelope. Similarly, the Census Bureau incorporated several
    changes into its Economic Census of Transportation/Commodity Flow
    Survey to reduce burden on businesses. For example, Census reduced
    the number of establishments that were surveyed by 50 percent;
    reduced the shipping period covered by the questionnaire from 2
    weeks to 1 week; reduced the number of shipments per questionnaire
    that respondents had to report on by 20 percent (from 50 to 40
    shipments); and allowed respondents to use estimates rather than
    actual numbers, particularly when generating the actual numbers
    would require the respondent to incur additional expenses.
    Changing the frequency of administering the surveys is another
    approach that could be used to reduce burden on business. However,
    in some cases the agencies are already administering the survey as
    infrequently as the underlying statutes allow.  Also, BLS and
    Census officials indicated that administering their large GPS/R
    surveys any less frequently could have a negative effect on the
    primary purposes of the data collections. For example, BLS
    officials indicated that federal policymakers and economic
    advisors use the monthly Producer Price Index in forming and
    evaluating monetary and fiscal policy and helping to evaluate the
    general business environment. They indicated that if this
    information were collected less frequently it would take longer to
    recognize and adapt to changes in the economy. Similarly, BLS
    officials said that collecting the quarterly Multiple Worksite
    Report data less frequently (such as annually or semiannually)
    would not provide the data necessary for certain users to produce
    important economic indicators or to administer various programs at
    the local, state, or national levels. In addition, Census
    officials stated that conducting economic censuses less frequently
    than every 5 years would diminish the time lines and usefulness of
    the statistics produced and diminish the usefulness of the
    economic census as a source of comprehensive information for
    economic policymaking, planning, and program administration.
    Likewise, Census officials said it is very important to collect
    export data through SED on an ongoing basis. They said that if SED
    data were collected less frequently, the government would not be
    able to produce monthly statistics on the trade balance. They
    indicated that this information is a valuable Page 22
    GAO/GGD-99-169 Federal Paperwork B-281700 component of the
    merchandise trade balance, which is one of the most sensitive and
    closely watched economic indicators produced by the federal
    government.  Slight movements in the deficit could affect public
    policy, financial decisions, or congressional actions. Census and
    BLS are also using information technology in some of their Use of
    Technology    surveys to minimize the burden imposed on businesses
    through their GPS/R surveys, to increase survey response rates,
    and to obtain more accurate and up-to-date information. One common
    use of technology for these surveys was some type of electronic
    filing of survey responses. For example, BLS collects more than 80
    percent of the responses to its Current Employment Survey
    electronically. Census collects nearly half of the SED responses
    in this manner. Electronic submission of the SED form cut the
    estimated time needed to complete the survey from about 11 minutes
    to 3 minutes. The format of those electronic responses sometimes
    includes a range of techniques for a single survey, including the
    use of computers, telephones, and facsimiles. For example, BLS
    collects responses to the Current Employment Survey via touch-tone
    data entry, computer-assisted telephone interviews, electronic
    data interchange, facsimile collection, and submission of tapes
    and diskettes. The agencies are continuing to develop their uses
    of technology in virtually all of these large GPS/R surveys, and
    have begun to provide technical assistance to businesses in their
    use. For example, in 1987, Census established a National
    Clearinghouse for Exporter Data Processing Services to encourage
    exporters to use technology when filing their SED reports. The
    Clearinghouse provides technological support to exporters, freight
    forwarders, and carriers who want to file electronically but do
    not have the capability to do so. Census officials said that they
    are replacing an antiquated electronic exports reporting system
    with a new state-of-the-art system that includes the filing of
    export documentation over the Internet. When this is completed,
    there is to be one automated export tracking system that would
    allow exporters to enter information into a single common system
    that is used by Census, the Customs Service, and other agencies.
    BLS identified several specific ways it is changing its data
    collection procedures to allow businesses to make better use of
    information technology. For example, employers with worksites in
    more than one state must currently submit their Multiple Worksite
    Reports to the relevant state agencies in each state. BLS is
    encouraging the use of a program to permit these employers to
    electronically submit the reports to a single data collection
    center that would then distribute the report to all of the
    relevant Page 23
    GAO/GGD-99-169 Federal Paperwork B-281700 state agencies. BLS is
    also working with several private organizations that prepare
    businesses' payroll and reports commercially to enable them to
    offer their clients the additional service of electronically
    submitting the Multiple Worksite Reports directly to BLS. Finally,
    BLS officials said the agency is developing a technology-based
    system for administering the Producer Price Index that could
    benefit both businesses and BLS. Under this system, BLS would be
    able to send businesses the price quotation forms via the computer
    and allow the businesses to respond by facsimile. For several of
    the 14 large GPS/R surveys, Census and BLS obtained at Use of Data
    From          least some of the needed data from administrative
    records kept by other Administrative Records    federal agencies,
    by state and local agencies, or by other organizations rather than
    asking the respondents to provide the same information. In at
    least some of these surveys, the use of administrative records
    resulted in less paperwork burden on small businesses. For
    example, the Economic Censuses on Manufactures, Professionals, and
    Utilities exclude most small establishments from the mail surveys,
    relying instead on information from administrative files kept by
    the Internal Revenue Service and the Social Security
    Administration. The Census Bureau obtains more than 40 percent of
    the responses to the Economic Census of Manufactures through
    administrative records and more than one-quarter of the responses
    to the Professionals and Utilities Censuses in this manner. In the
    SED program, the United States and Canada entered into an
    agreement in 1990 to share data on imports and exports, thereby
    reducing the paperwork burden on shippers. In this agreement, the
    Census Bureau uses Canadian data on imports from the United States
    as its measure of American exports to Canada. Similarly, Canada
    uses U.S. data on Canadian imports to replace the same data it had
    collected from its exporters. This data-exchange agreement has
    resulted in the elimination of the requirement for business
    exporting products from the United States to Canada to file SEDs
    for each shipment, thereby saving American exporters from
    preparing an estimated 3.5 to 4.0 million SEDs annually. BLS also
    uses administrative data to reduce the paperwork burden associated
    with some of its GPS/R surveys. For example, businesses typically
    report workplace injuries and illnesses to both BLS through its
    Occupational Injuries and Illnesses Survey and to state
    governments through their workers' compensation programs. However,
    version "W" of the Occupational Injuries and Illnesses Survey
    enables businesses in six participating states to provide less
    detail on their survey response if they have already reported the
    detailed information to their states' workers' Page 24
    GAO/GGD-99-169 Federal Paperwork B-281700 compensation programs.
    BLS obtains this information from the worker compensation agencies
    in those states. In the Information Collection Budget for fiscal
    year 1999, OMB identified Statutory Restrictions on    seven
    categories of initiatives that agencies were undertaking to reduce
    Data Sharing Prevent         paperwork burden. Census and BLS have
    begun or completed initiatives in Further Burden Reduction
    most of these categories. However, Census and BLS are statutorily
    prohibited from fully utilizing one of the seven categories-data
    sharing. The Census Bureau's restrictions on data sharing are in
    13 U.S.C. 9, which states that no officer or employee of the
    Department of Commerce or bureau or agency thereof, may "(1) use
    the information furnished under the provisions of this title for
    any purpose other than the statistical purposes for which it is
    supplied; or (2) make any publication whereby the data furnished
    by any particular establishment or individual under this title can
    be identified; or (3) permit anyone other than the sworn officers
    and employees of the Department or bureau or agency thereof to
    examine the individual reports." This and other statutory
    prohibitions on data sharing were put in place to protect the
    confidentiality of survey respondents. The original PRA in 1980
    gave the director of OMB the authority to direct a statistical
    agency to share information it had collected with another
    statistical agency. However, this authority did not apply to
    information covered by laws prohibiting such disclosures. In the
    early 1980s, the statistical agencies, under OMB's leadership,
    proposed legislation that would have extended their ability to
    share data, but legislation was not enacted. For the past 2
    decades, we and others have urged legislative changes that would
    allow greater sharing of data and information on data sources
    among agencies. In 1979, we recommended that existing law be
    amended to allow the Bureau of the Census to share information on
    business establishments with other government agencies.11 More
    recently, in March 1998, we said that the inability of statistical
    agencies to share data is one of the most significant issues
    facing the statistical system and one of the major factors
    affecting the quality of data, the efficiency of the system, and
    the amount of burden placed on those who provide information to
    the agencies.12 11After Six Years, Legal Obstacles Continue to
    Restrict Government Use of the Standard Statistical Establishment
    List (GAO/GGD-79-17, May 25, 1979). 12Statistical Agencies:
    Proposed Consolidation and Data Sharing Legislation (GAO/T-GGD-98-
    91, Mar. 26, 1998). Page 25
    GAO/GGD-99-169 Federal Paperwork B-281700 Data sharing is still an
    issue of concern in both the executive and legislative branches.
    One of the Clinton administration's key initiatives identified in
    the Statistical Programs of the United States Government: Fiscal
    Year 1999 addresses needed changes in statistical confidentiality
    and data-sharing statutes and practices. Also, legislation has
    been introduced in the 106th Congress (S. 205), which would, among
    other things, permit limited sharing of records among designated
    agencies for statistical purposes under strong safeguards. Both
    Census and BLS support administration proposals to revise the
    confidentiality statutes to allow data sharing among statistical
    agencies. Officials in both agencies said that proposed changes
    permitting greater data sharing offered a major way to reduce
    burden on businesses without compromising the confidentiality or
    the quality of the information collected. However, neither BLS nor
    Census is able to estimate how many burden hours might be reduced
    in general or for specific surveys if data- sharing legislation
    were enacted. A Census official told us that the extent of burden
    reduction would depend on the specific provisions of the
    legislation and on how data sharing was implemented. Among the 14
    surveys that we examined, the official identified instances where
    Census and BLS have duplication or overlap between collections and
    could potentially reduce burden by sharing. However, he said the
    greater potential for reducing burden would come from the agencies
    jointly reengineering survey programs to increase standardization
    and reduce duplication. GPS/R surveys directed at businesses are
    extremely useful to federal Conclusions    agencies, state and
    local governments, and to the businesses themselves. For example,
    the Producer Price Index is used to make decisions on monetary
    policy and to evaluate the condition of markets. Information from
    the Occupational Injuries and Illnesses Survey is used to
    determine the effectiveness of federal and state safety programs.
    The SED data are the official record of export transactions from
    the United States. These surveys also contribute data to national
    economic indicators that have been developed and refined over the
    years to meet the needs of economic policymakers and others for
    timely, accurate, and consistent information on the state of our
    national economy. Although GPS/R surveys of businesses account for
    a very small portion of federal agencies' total paperwork burden,
    the 12.6 million hours that agencies estimated it took all
    businesses to complete GPS/R surveys in fiscal year 1998 is a
    substantial paperwork burden. We estimated that businesses'
    aggregate labor costs to complete these surveys ranged from Page
    26                                          GAO/GGD-99-169 Federal
    Paperwork B-281700 roughly $219 million to about $305 million.
    However, the burden imposed on any one business may be quite
    small. Census and BLS, which account for the bulk of the burden
    hours associated with GPS/R surveys, have taken and continue to
    take steps to minimize or reduce the burden associated with these
    information collections. However, Census, BLS, and other
    statistical agencies are unable to minimize or reduce paperwork
    burden as much as possible because of statutory restrictions on
    sharing data collected for statistical purposes between the
    agencies. We, OMB, and the statistical agencies have supported
    revising those restrictions. Doing so could reduce information
    collection costs to both respondents and the agencies collecting
    the data. We provided a draft of this report to officials
    responsible for GPS/R survey Agency Comments and policy and
    administration in Census, BLS, and OMB for their review and Our
    Evaluation                 comment. We received oral comments from
    Census and BLS on August 20, 1999, and from OMB on August 30,
    1999. Each agency agreed with our overall characterization of
    GPS/R surveys of businesses and our conclusions and each suggested
    technical and editorial changes, which we incorporated into the
    final report as appropriate. We are sending copies of this report
    to Representative William J. Pascrell, Jr., Ranking Minority
    Member, House Committee on Small Business' Subcommittee on
    Regulatory Reform and Paperwork Reduction; the Honorable Jacob J.
    Lew, Director of OMB; the Honorable Alexis M. Herman, Secretary of
    Labor; and the Honorable William M. Daley, Secretary of Commerce.
    We will make copies available to others on request. Page 27
    GAO/GGD-99-169 Federal Paperwork B-281700 If you have any
    questions regarding this report, please contact me on (202) 512-
    8676. Key contributors to this assignment were Curtis Copeland,
    Elizabeth Powell, and Ellen Wineholt. Sincerely yours, L. Nye
    Stevens Director, Federal Management and Workforce Issues Page 28
    GAO/GGD-99-169 Federal Paperwork Ordering Information The first
    copy of each GAO report and testimony is free. Additional copies
    are $2 each. Orders should be sent to the following address,
    accompanied by a check or money order made out to the
    Superintendent of Documents, when necessary. VISA and MasterCard
    credit cards are accepted, also. Orders for 100 or more copies to
    be mailed to a single address are discounted 25 percent. Order by
    mail: U.S. General Accounting Office P.O. Box 37050 Washington, DC
    20013 or visit: Room 1100 700 4th St. NW (corner of 4th and G Sts.
    NW) U.S. General Accounting Office Washington, DC Orders may also
    be placed by calling (202) 512-6000 or by using fax number (202)
    512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of
    newly available reports and testimony. To receive facsimile copies
    of the daily list or any list from the past 30 days, please call
    (202) 512-6000 using a touch-tone phone. A recorded menu will
    provide information on how to obtain these lists. For information
    on how to access GAO reports on the INTERNET, send e-mail message
    with "info" in the body to: [email protected] or visit GAO's World
    Wide Web Home Page at: http://www.gao.gov United States General
    Accounting Office           Bulk Rate Washington, D.C. 20548-0001
    Postage & Fees Paid GAO Permit No. G100 Official Business Penalty
    for Private Use $300 Address Correction Requested (410381)

*** End of document. ***