RESEARCH

TRIANGLE

INSTITUTE

RTI/4131-OOF

THE NATIONAL

SURVEY OF SMALL BUSINESS FINANCES: FINAL METHODOLOGY REPORT

Research

Brenda 6. Cox Triangle Instltute

Gregory E. Elltehausen Federal Reserve Board John D. Wolken Federal Reselvc Board

September 1989

POST

OFFICE

BOX

12194

RESEARCH

TRIANGLE

PARK

NORTH

CAROLINA

27709-2194

TABLE OF CONTENTS

Page
ACKNOWLEDGEMENTS .................................................. LIST OF TABLES .................................................... I
V

1.
2.

OVERVIEW

....................................................

1 4 4 6 9

BACKGROUND AND OBJECTIVES ..................................... 2.1 2.2 Bank Market Deflnltion ................................... Other Survey Objectlves ..................................

3.

SAMPLE SELECTION AND WEIGHTING ............................... 3.1 3.2 3.3 3.4 3.5 Deflnition of the Target Population ...................... Choosing a Frame for NSSBF Sampling ...................... Constructing the DMI Sampling Frame ...................... Coverage of the DMI Sampling Frame ....................... Design of the DMI Sample ................................. DMI Sample Selection ..................................... Constructlon of DMI Analysls Welghts ..................... SBA Sample Selectlon and Welghtlng .......................

3.6 3.7
3.8 4.

QUESTIONNAIRE DEVELOPMENT .....................................

35 35 37 43 53

4.1 4.2 4.3 5.

Questfonnalre Fonnatting and Valldatlon .................. Cognltlve Testlng of the Questfonnalre ................... Pretest Results ..........................................

DATA COLLECTION ............................................... 5.1 5.2 5.3 Screening Results ........................................ Intervlewlng and Response ................................ Use of Worksheets ........................................

:; 60 65 65 69 72 :i 77 79

6.

OATA PROCESSING ............................................... 8:: :*: 6:5 6.6 Computer Asslsted Telephone Intervlewlng ................. Problem Resolutlon Reports ............................... Completeness Checks ...................................... Edlts to Resolve Interviewer Errors ...................... Hatching to Instltutlonal Flles .......................... Final Editlng and Ffle Dellvery ..........................

7.

CONCLUDING REMARKS ..................................... .... ...

ACKNOWLEDGEMENTS

The National Survey of Small to implement in computer-assisted terms of size,

Business Finances was a challenging study complexity of the subject, and use of The aid of many

telephone

interviewing

(CATI).

individuals, both inside and outside needed to complete assistance. The most important Depending the study. The

Research Triangle Institute (RTI) was authors gratefully acknowledge this

contributions upon the last

were

made extent minutes

by of

survey respondents use of financial

themselves.

business's from of 15 the

services, an interview could Without the voluntary

to close to 2 hours.

cooperation

sampled small businesses, this

study would not have been possible. We thank Frederick Yohn and at the Board for their Mcwilltams and Andrew advice members of the Financial Structure Section during the survey. research Gerhard Fries, Ronnie assistance to Board

Steinerman

provided

staff during the course of the for the generous support

survey.

Special recognition is also needed by staff of the Small Business

provided provided

Administration. study planning

Charles Ou and

guidance and consultation throughout Bruce Phillips and Satyapal (hera

implementation.

(Social and Scientific Systems, the DMI portion of the study.

Inc.)

provided sample selection support for

Allen Handel consulted on issues associated

with sampling firms with SBA-guaranteed loans. Others outside RTI who provided the many reviewers who survey questionnaire. provided Special assistance implementing the study were advice is on the content of the to Robert A.

technical

appreciation

extended

Eisenbeis, Wachovia Professor of Banking University of North Carolina at Chapel

and Coordinator of Finance at the Hill, and to Ernest Hylton, a Dr. Eisenbefs for

certified public accountant with Koonce, Wooten and Haywood. reviewed the questionnaire with respect to its

appropriateness

investigatlng the geographic and in bank acquisition and merger

product market assumptions currently used cases. that Mr. Hylton consulted on accounting would minimize respondent burden

issues and questionnaire approaches

and misunderstanding of the questions being asked. The authors appreciate the study. testing Judy Lessler the survey designed efforts the of RTI staff In implementing this evaluatlon approach used in Monroe based formatted upon the draft

cognitive Nancy

instrument. the

questionnaire and testing.

revised

questionnaire

the results of

Don Anderson and

Michael Janella

Weeks provided technical advice during Pantula selected samples for the two

questionnaire development. survey components. materials for the

Lanny Piper study and

and Susan Henderson developed interviewing supervised all aspects of data collection.

Christy Kooyman and Lance Tashet developed questionnaire. Lisa Packer verlfied Viviane the Board's

the CAT1 program for the survey

completeness of survey interviews and Cobb data matched respondent reports of bases of commercial banks, G. G. Frick and Elinor Finally, the authors in the typing of this

calculated analysis weights. finsncial institutions to

savings and loan associations and consnercial banks. Cfftan produced the final deliverable attention data to base.

thank Brenda Porter for her

accuracy

report and the many revistons of the survey questionnaire. Last but not least, the authors Telephone Survey Unit of RTI. The express special thanks to staff of the information being requested in this

survey was considered businesses. The

quite and

personal managers

and of

highly

confidential by sample

owners

small businesses are difficult with much to do. business owners Tact and and to

people to contact with schedules already packed persistence was required to to locate the small

persuade them of their need

participate.

Without their dedication and

persistence, we could not have achieved the response rates and data quality that the study exhibits. This report was prepared by Board of Governors of the Research Triangle Institute (RTI) for the System and the Small Business

Federal

Reserve

Administration under Contract 00010-88. in this document are those of the offlclal position or

Points of view or opinions stated

the authors and do not necessarily represent of the Federal Reserve System or the

policies

Small Business Administration.

-*

-m

TABLE OF CONTENTS (cont.)

Page
REFERENCES ........................................................ APPENDIX A: APPENDIX B: APPENDIX C: NSSBF SCREENING FORM ................................. NSSBF LEAD LETTER MATERIALS .......................... 1987 INCOME TAX RETWW .............................. 81 A-l B-l C-l

LIST OF TABLES

Table 1 2 3
4

NSSBF Frame Counts by Hetropolltan Status and Other Business Characteristics ................................ . NSSBF Frame Counts by Hetropolitan Status, Number of Employees and Total Sales................................ NSSBF Framz Counts by ktropolftan Status, Census Region and Number of Employees...'........‘..................... Sample Allocation Across Domafns Deffned by Census Region, Metropolitan Status, and Fins Size....................... Allocatfon to Strata for the Various Stages of DMI Sampling................................................. NSSBF Pretest Screenfng Results.......................... NSSBF Pretest Intervlew Results.......................... NSSBF Screening Results.................................. NSSBF Interview Response and Eliglbfllty Results......... Response Rates by Reporting Domatns...................... Use of Records When All Sectfons Completed...............

20 21 22 24 27 44 47 54 59 61 62

5 6 7 B 9 10 11

LIST OF EXHIBITS

Exhfbit 1 2 Structure of the NSSBF Questlonnalre .........*.......... . NSSBF Problem Resolution Report..........................

m 36 70

I

1.

OVERVIEW

Uith respect to antitrust activities, the government currently treats banking as a distinct and separate line of cOlllDerce and includes within the market definition of banking only banking services. those firms that offer a full array of banking markets are

In

addition, it

assumes that

i'ostricted the local geographic area. How?ver, ncmetvus chdages illbank to regulatlon, technology and the economy have occurred over the past two now provided, not only by banks, but by

decades. Banking services are

thrift and nondepository institutlons. In addition, many of the firms that provide banklng services operate market. In light of these In a in national, and even lnternatlonal financial services, the prevlous

changes

market definition for banks may no longer be appropriate. There has been little systematic study of bank market. the recent changes In the

In particular, vlrtually no evidence has been gathered on the affected the use of financial services intensifies the move toward

extent to which these changes have by small businesses. As

interstate banking

bank mergers, adequate government deliberation of mergers and acquisitions will depend on the avallabllity of data relevant to the following

questions: . . . . How can economlcally meanlngful banking markets be defined? Wlthln what geographic area do their financial services? small businesses typically obtain

Do small businesses prefer obtaining services from a particular type of flnanclal institutlon, such as a coasnercial bank? Do banks compete for small business credit and deposit services In a single product market or in several separate product markets?

Appropriate application of antitrust regulations iS particularly crucial in rural and small urban areas, because small businesses have fewer service

options in these geographic areas. Small businesses account for about U.S. industry and a majority 1988). Despite the importance a third of employment and sales of

of its growth (Small Business Administration, of small businesses to the economy, little

is known about the financial side of their operatlons. S~IJI-ISOX~ by Administratlon the Federal Reserve by 3oard and the Small Business

(SBA) and

conducted

Research Triangle Institute (RTI), Finances (NSSBF) meets this need for and their finances. The target

the National Survey of Small Business basic data about small businesses included of A 500

population for the survey

all those

for-profit, privately owned small whose principal business activity

businesses with the exception was agriculture or finance. that

small business was defined as an enterprise employees across all of its branches or

employed

fewer

than

establishments.

Dun's Market Identiffers A stratified, the

(DMI) file

was used to construct

the sampling frame.

one-stage

sampling design was used to

select sample businesses, where

strata were deffned using urban/rural

categories, Census region, and number of employees. A total of 3,600 small businesses had together with an addltional 400 The interview

businesses were intervlewed who was by telephone wlth an

SBA-guaranteed mailout

loans.

initial

advising the business of their

selection for the study and the completing the detennlned such interview. as

company

records that would assist them in of the small business were type, principal Detalled

Characterlstlcs location, and

geographic and

organization management

business activities,

ownership

structure.

2

information was collected for thefr last fiscal year on: .
.

their use of financial services for each fnstftutfon used; the characteristics of their loans, leases and credit lines; the extent to business: and which financial the institutions solicited the firm's business shopped around for

. .

the degree to which financial services.

small

Data were also obtained that

allowed

balance sheets and income statements

to be corci.ructed for the ffrar. Several procedures were Implemented to the main survey, the questionnaire was enhance data quality. Prior to

tested using cognitive psychology then In a fornral pretest of 30

techniques in a laboratory envfronment and small businesses. The telephone

Interviews were conducted using computer methodology with more extensfve than Following the

assisted telephone Interviewing

(CATI)

usual edit checks occurring while the interview was ongoing. interview, respondents were asked hard copy documents to RTI. to mail

financial statements and other

These documents were transmitted to the Board Extensive edit checks were also made of Too

for use in verifying data quality.

completed fntervfews to Insure that sufficient data had been obtained.

many missing responses in a completed fntervfew resulted in follow-up calls to obtain the missing data. This Institute Finances. report (RTI) sumnarfzes fn the methodology the National used by Research Trfangle of Small Business

conducting

Survey

3

2.

BACKGROUND AND OBJECTIVES

One of the regulatory functions of determine whether proposed mergers companies and of certain banks are

the

Federal

Reserve Board is to bank This holding requires

or

acquisitions of

anticompetitive.

empirical delineation of the economic markets In which these firms operate. Unfortunately, little data exist on of banking markets. the geographic and product dimensions rescarch and c0ui.t

Current analysis relies on two decades old. that would aid in

decisions that are more than survey was to collect data

A major purpose of this determining whether the

current approach to market definition in bank merger cases is appropriate. Large firms likely operate in It is not clear whether as wide In scope. small natlonal or International financial markets. firms and households face financial markets

Information about household's use of financial services Consumer Finances. One major objective of banking markets should be delineated for for this task also provide Information

is avallable from the Survey of

this survey was to determine how the small business. Data needed

about other issues associated with small business finances. 2.1 Bank Market Definition Historically, the financial system In the US has been characterized by

legal constraints that segmented markets along institutional lines. Thus, conmcrcial banks provided business credit and unions provided consumer loans, and checking services, credft

savings and loan associations provided

mortgages. This market segmentation was recognized in the Supreme Court's decision In the 1963 Philadelphia National Bank case which forms the basis for the current approach to bank market definition. Considering the

evidence available at that time, the court concluded that because customers generally obtained multiple financfal products from one place, only firms offerfng the full array of bank markets (In other commercial bank products should be included in cormoerclal banklng was a dfstfnct and

words,

separate lfne of cwnerce).

Furthermore, the court detennfned that banking

markets were limlted to small geographic areas because the bulk of banking business was conducted wfth local customers. Small busfnesses (and

househclds) were considered to be

constrained in their choice of financial

product supplfers. with indivfdual bank

Subsequent survey research conducted in conjunctfon merger and acqulsftlon cases supported the court's

decfsfon, and In the 1974 Connecticut Natlonal Bank case, the Supreme Court

reafffnned Its earlfer decfsion.l Sweeping regulatory changes, advances in technology, and financial

fnnovations have occurred since the Connecticut declsfon. Most notable are the regulatory changes empowering savlngs Institutions to offer traditional bank products (checkfng accounts, comnercial loans, and consumer credit), the emergence of nondeposltory instltutlons (money market mutual funds) as competftors of deposltory fnstftutfons for household savings accounts, and technologfcal changes reducing depository and (such as electronfc fund transfer technology) of delivering

nondeposltory institutfons' cost

financfal servfces to final users. These changes make the current approach to bank market deftnitlon appear Increasfngly antiquated. Crltlcs argue

that the provfsfon of bank products takes place fn a market that includes

11

For an extended discussion of the Issues, see Federal Reserve Bank of Atlanta (1982).

s

both thrifts and nondepository institutions contend that the geographic market in

as which

well these

as banks. fins

They also operate has

national, if not international,

dlmenslons. of financial

While large businesses have products In a national or

access to a

variety

of

sources Is

international market, there geographic extent of small

little

infonsatlon

about the sources and Except for data

business

financial

dealings.

collected in conjunction with evidence Is linlited to Reserve Bank of Atlanta two

individual surveys of

bank merger applications, recent 1981: a Federal

ro;rducted durlug small

survey

businesses in the Sixth Federal of small businesses In Ohio

Reserve District (Whitehead 1982)

and

survey

sponsored by the Federal Reserve Bank of Cleveland (Watro 1982). The need to study product and geographic markets dictated that NSSBF First, to delineate geographic markets, the geographic location of suppliers of small business. Second, the survey

collect data along two dimensions. the survey obtained information on financial products relative collected an Inventory of obtained from services as a each to the the

different to a

financial products that the firm whether institution (in other the business obtains or purchases them whether banks

supplier from

determine single

'cluster' dtfferent

separately from compete In

institutions

words,

a single product

market or In several separate product markets).

2.2 Other Survey Objectives The bank market definition problem of how small businesses finance their to Is only one component of the study

actlvitles.

In a

modern economy, Finance is

buslness firms invest in real

assets

carry on production.

concerned wlth the questions of how much

the firm should invest and how It Important considerations

should obtain funds to pay for these investments.

for the small business are the its operations.

cost

and avaflabflity of credit to finance

Hotivated in part by concerns about credit avai1abflfty, the

the effects of monetary policy on

federal government sponsored several studies of the Second World War (Bridge 1948; Bridge and Cfaccio 1955; Board of Governors

small business financing after and Holmes 1950: McHugh 1951;

McHugh

1958; Stockwell and Byrnes 1961). dominated all othe: sources of

These surveys showed that bank financing

outside fira~zfng, supporting the view th;i an important detenainant of small business for investment, however, were obtained a broader perspective is needed to

availability of bank credit was Investment. The majority of from owners. This

funds

suggests that

understand small business finances. Unfortunately, most efforts to collect financial data from small businesses were abandoned after the 1950s. Since these early studies, advances in the ability of researchers to economic theory have enhanced

study business finances. Modfglfani and

Hiller's (195B) pathbreaking work, which demonstrated that the value of the

firm is independent of its source of financing in a perfect capital market,
motivated substantial theoretical analysis of imperfections (such as differential tax costs, and asysssetric information) on the effects of market

rules, transactions costs, agency the cost 1988; and availability of

alternative sources of ffnancfng (see Miller market imperfections may have a greater

Weston 1989). These

effect on smaller finns than on

larger firms, providing a basis for of smaller firms differs 1985). from

empirical analysis of how the behavior of larger firms (Pettft and Singer not available for this purpose.

that

However, adequate data

are

Existing surveys generally do not cover all sources of financing (Dennis

7

1985; Dennis and Dunkelberg 1988) or have limited or incomplete coverage of the small business population (Ando 1985; Combs, Pulver, and Shaffer 1979; 1988). A few sources provide sumry

Dennis 1985; Dennis and Dunkelberg statistics of data from prepared

balance sheets (Internal Revenue Service

1989a, 1989b; US Bureau of the Census 1987: Robert Morris Associates
While these sources include all or no information on the sources of

1988).

of financing, they provide little the sources of financing or on

identlfy

demographic characteristics of the fInas. The NSSBF collects a complete of financing to the flrm. and management It balance sheet and identifies the sources also collects information on the ownership of the and fin, extent collateral and guarantor

characteristics

characteristics of the firm's debts,

of use of cash services.

This infonnatlon permits empirical analysis on small business flnanclng

of many of the basic questions

(Du 1986; Pettit and Singer 1985).

21

Ou (1986) discusses prevlous surveys and other data sources on small business finances in greater detall.

8

3.

SAMPLE SELECTION AND WEIGHTING

The sampling plan for

the

National Survey of Small Business Finances

was designed so that study findings could be used to make inferences about small businesses In the United States. Major steps involved in data

collection for the study included: . . . . . . . defining the target population, constructing the sampling frame, designing the sample, selecting sample businesses, screening for eligibility, interviewing eligible persons, and weighting the survey data. to two distinct sample designs. The

Two distinct objectives for NSSBF led first objective was to conduct a

multipurpose national survey of small obtaining data that could be bank market definition in

business finances with special attention to used in reviewing the Board's approach to

antitrust cases. To

satisfy this objective, a sample was selected from A second objective was to obtain the Small Business

Dun's Market Identlfiers‘(DM1) File. detailed data for firms having

loans guaranteed by

Administration (SBA). This

chapter discusses the DMI sample first: then

modifications made for the SBA sample are discussed In the last section. 3.1 Definition of the Target Population The target population for a survey is the entire set of elements about which Inferences will be made using the survey data (Cox and Cohen, 1985,

pp. 20-22). For the DMI sample, the

target population was defined to be

all small business enterprises in the United States that are for-profit and privately owned and whose line of business Is not agricultural or ffnanclal in nature. The rationale behind the rules made In defining the target

population follows. An enterprise was defined as an autonomous operating entity, whfch a firm. While production and level of the firm, financial

includes all subsldlaries and sales occurs at the

branches of branch

subsidiary or a

decl<lons typlcally are made at and subsidfaries of the flnn.

hlgher level and encompass 211 branches This approach is slmflar to the Census

Bureau methods for financial surveys of Statistlcal Methodology 1988). An enterprise was defined to than 500 full-time equivalents In be a u

businesses (Federal Committee on

business if it employed less This definition Is one consnonly

1987.

used by the Small Business Administration In its reports (e.g., Small
Business Adminlstratlon 1984, 1988)z Large firms were excluded from the access to a national market in bank antitrust cases.

target population because they generally have for financial services and are not a

concern

Moreover, a large amount of publicly available data already exist for large busfnesses. Nonffnancial and nonfarm business was and for-profit buslnesses,
excluding

defined as all prlvately owned
groups: (1) agriculture, and (3) real

fndustry

forestry, and flshlng; (2) finance and insurance underwriting;

;/

Total

sales are also commonly used In deflnfng what constitutes a small business. Uslng total sales to define the target population for NSSBF would have been lmpractlcal, however.

10

estate investment trusts.
transportation, and

Some

industry groups (especially utilities,
services) as contain both privately and

educational as well

publicly owned entities

for-profit and not-for-profit

firms.

Ineligible

firms in these industry

groups were identified in the screening

interview, discussed later in this report.
These restrictlons correspond corporate funds to the definitions for the nonfana, in the Federal

noncorporate business and Reserve Board's flow of

nonfinancial (Board

sectors of

accounts

Governors 1980)

It!

additjon to conformance with eliminate from the target

extant statistical programs, the restric ions several types of organizations that

population

dfffer substantially from most businesses. business (financial intermediation) makes of other business finas. for most financial firms. excluded because their Moreover, Publicly

The nature of financial f T-Ills’ their behavior differ from that

detailed financial data art available owned and not-for-profit and firms were

peculiar

objectives than

environment give rise to used by for-profit firms.

different record keeping procedures

those

Agricultural firms, on the other hand, are often small and operate in rural markets where bank antitrust problems art likely to arise. However,

adequate data art routinely collected for agricultural firms (US Department of Agriculture 1988), and the existence of substantial federal programs for agrtculture mitigates concerns about adverse effects on credit availability for this industry. Finns that were no longer than a month at the survey. time of in business, interview bankrupt, or in business less

the
the

were also intligiblt for the

At the target

date,

finances of such firms would not reflect

those of a going concern.

11

3.2 Chooslnq a Frame for NSSBF Samplinq The sampling frame for a survey is the list or mechanism used to The ideal

identify population elements for frame is a list of all

sample

selection purposes.

population members with sufficient data to identify

each member and locate them for Interviewing purposes (Cox and Cohen, 1985, pp. 24-33). Available list frames can be bus+ne+ses repcrt categorized by the underlying reasons why themselves on a continuing basis.

Information about

These underlying reasons impact on

the

completeness and accessibility of

the frame. Several alternate frames were considered before choosing Dun's Market Identiflers (DMI) file as the basis for sampling. Perhaps the most complete of business the Internal Revenue Service (IRS). frames is the tax return file of

All corporations in existence at any

time during the tax year must file a corporate income tax return regardless of whether they had any income (Internal Revenue Service 1987a, p. 1). file a partnershlp tncome had income during the year

With llmited exceptions, all

partnerships must whether they

tax return, agaln regardless of

(Internal Revenue Service, 1987b, pp. l-2). addition to their personal return when

Proprietors file a Schedule C they have business income or

deductlons to report. To be included in IRS tax records, a business must have started

operations before the present tax year. cannot be accessed via thls disobey tax ffllng laws. Completeness issues are a minor accessibility and timeliness. Tax data base, as

Hence, newly formed businesses well as those businesses that

concern for the IRS frame compared to legislation imposes strict regulations

12

on the IRS and Its

release of

confldentlal data.

Tax data was used to

select a high-Income sample for the Comptroller of the Currency malled

1983 Survey of Consumer Finances. The each partlclpant a letter requestlng

their partlclpatlon In the survey. The only names released to the survey's outslde contractor were the nlne that they were wllllng to percent who had notlfled the Comptroller

partlclpate (Avery and Elllehausen,l988). As a returns had to be used In selecting

further completeness Issue, 1980 tax the sample. Businesses go out of per month (Converse and

business at a rate qf abollto:lepercent

Heerlnga, 1984) suggestlng that an IRS tax-based

sample frame might suffer from a 15 to 20 percent obsolescence rate as well as fail to cover new businesses startlng operatlons durlng that tlme of Consumer Finances experience, then, suggests lnfeaslble for use In surveying small

perlod. The 1983 Survey that IRS

tax-based frames are

businesses due to Inherent response and tlmellness problems associated wlth the severe confldentlallty restrlctlons placed on their use. Another frame resultlng from Statistics' Employment and that flnns wlth flve or Wages a legal mandate flle, which Is the Bureau of Labor

results from a requirement

more

employees must file wlthholdlng lnfonsatlon.

The InformatIon in the BLS data base ls confldentlal and the flle cannot be released to outslders. However, the Bureau of Labor Statlstlcs has

selected satnples for other

government agencies and supplled the necessary An example of such

ldentlflcatlon lnfonnatlon for the sample facllltles. Interagency cooperation Is Survey conducted by the furnlshed by

the Natlonal Occupatlonal Hazard Occupatlonal Safety and

Natlonal Instltute for coverage of

Health. Besldes the lack of flle had deflclencles that

very small businesses, the BLS for the NSSBF. Telephone

precluded Its use

13

numbers are not available and Idcntfficatlon of the headquarter's office is difficult because a mafn office is firm has branches. For NSSBF designated for each state In which the use, study planners were advised that have to be obtained from each

permission to use the sample

records would

state and state confldentfallty rules would apply to use of the data. Another alternative frame considered for raeplfng frame. The area-based frame NSSBF was an area-based

is constructed by selecting a sample interviewers list all bvsincsses

of geographic areas and located in these areas.

having field Size

measures needed to select sample areas for obtain. be In additlon, ldentlfying buslnesses a nontrivial problem for this study.

listing would be difficult to

In sampled areas was expected to

NSSBF's target population includes home-based businesses and businesses without an official site at which operations are conducted. As a result,

the coverage achieved by the area-based design in practice would not be as comprehensive as expected. Moreover, listing the indtvidual businesses in an area would be expenslve. Lack of identifying data for the llsted

businesses would mean that all methods with an

interviewing would have to use face-to-face identify out of scope

initial screening interview to

businesses and branch offices. None of the frames described above include telephone numbers, which The expense of a face-to-face

precludes their use in a telephone survey. interview, together with the data bases or implementing an

complexities of gaining access to government area-based approach, led to the decision to

use a consnercially available data base for sampling. 3.3 Constructing the DMI Sampling Frame For this study, Dun's Market the sampling frame. The Identifiers @MI) file was used to build

CM41 file combines records 'derived from the 14

traditional Dun and Bradstreet credit credit or interacting with businesses

rating program -- firms applying for who require credit information --

with business telephone listings. listings file substantially the traditional films, firms program, with few

The addition of firms from the telephone the there and coverage problem associated with is still in undercoverage of new service industries

reduced although employees,

firms On

(Iannacchionne, LaVange and Duffer 1986). listings are updated evt,r;12 used for the NSSBF. Information on the tfi 18

a flow basis,

DMIbusiness

months.

The December 1987 file was

DMI

file

includes

the business address, telephone (SIC)

number, main office/branch status, Standard Industrial Classfffcation code, and the name of the owner or chief would executive officer. have

Without

telephone numbers, a impossible from a

telephone

survey

been difficult if not The other information

cost/data

quality

standpoint.

allowed us to eliminate many out of scope businesses and provided a contact person for mailouts and subsequent telephone calls. The December 1987

DMI File uses

1972 SIC codes instead of 1987 codes.
for NSSBF could be identified by

Lines of businesses that were the following SIC codes:

ineligible

0000 - 0999

Agriculture, Forestry and Fishing 4311 U.S. Postal Service 6000 - 6399 Finance and Insurance Carriers 6724 Unit Investment Trusts 6732 Educational, Religious and Charitable Trusts 6733 Trusts, Except Educational, Religious and Charitable 6798 Real Estate Investment Trusts 8600 - 8699 Membership Organizations 9000 - 9721 Public Admfnistration. all unit records with the above SIC codes

In constructing the sampling frame, were

deleted.

Since

the

sampling

was the entire enterprise rather

15

than

fndfvidual

establishments,

DMI

records

for

branch

offices

and

subsidiary companies were also excluded. Previous RTI experience with the DMI file it suggested that the Total is not defined in terms of

Employees variable was imprecise.

Further, can

full-time equivalents so that a fins less than 500 sampling frame, full-time the equivalents. Employees

have mOre than 500 employees but

Hence,
variable

in constructing the NSSBF was not used to exclude

Total

records from the frame. 3.4 Coverage of the DMI Sampling Frame To evaluate coverage issues further, with we compared frame counts for

establishments on the

DMI file

statistics

on business tax returns

compiled by the Statistics of Income Service 19BBa, 1988b,

Division of the IRS (Internal Revenue comparisons are not possible, but IRS

1987).

Exact

statistics provide some basis for evaluating the DMI file's coverage of the universe of interest for this study. Before discussing our flndfngs, the IRS and DMI lists should be is the business tax return. several conceptual differences between Ftrst, the IRS unit of observation other firms have the option of

noted.

Firms

owning

filing consolidated returns for the enterprise or separate returns for each company. Hence,

IRS

statistics Next, the

overstate

the

number include

of corporate and firms that were

partnership enterprises. active at any time during firms that were in

IRS
year,

statistics while

the NSSBF includes only those point of time, Thus, IRS

business

at

a

specified

estimates of the population of businesses wtll be larger because they count firms that change ownership more than of business before the reference date once and include firms that went out and firms that started operations

16

after the reference date. more than 500 employees.

In
This

addltlon,
is not

IRS estimates include firms with
a large number, however, and it Wtth these caveats, we present

primarlly affects the corporation total.4 some results of our prelimlnary work. The DMI estimate for number of

corporations Indtcate

is not much smaller than that the majority of

IRS est1mate.i

Both

data

sources trade is

corporations were engaged in retail of ccrpo-ations ry industry groups contains proportionately more manufacturing and

and services.

The distribution

also similar, although the DMI list service than the and real estate firms and

somewhat fewer trade firms DMI

IRS

list.

Hence, our

preliminary work suggests that the good. The DMI esttmate of the is substantially lower more number the

file's coverage of corporations is

of proprletorships,

on the other hand, In both lists,

than

IRS
In

estimate.
the service

proprietorships are

concentrated

and construction

industries and less concentrated or corporations. groups.

in manufacturing than either partnerships in all Industry

The DMI list undercovers proprietorshlps a greater

The DMI file contains

proportion of trade firms and a

smaller proportion of services than the IRS list.

21

There Is one further discrepancy. These preltminary comparisons are based on the latest avatlable published statlstlcs -- 1985 for corporations, 1984 for partnerships, and 1983 for proprletorships. The IRS totals for 1987 would be somewhat larger than the numbers we

used.

51

The DMI file does not tdentlfy distribution of fins by ownership results from the NSSBF. 17

ownership form. We estimated the form within industry groups using

A large part of

the

coverage

problem

for proprietorships

appears to Businesses

arise from a lack of coverage without employees are businesses and are mostly not for

of businesses without employees. proprietorships. identified. part-tf=

They are often part-time These firms include, for 1s a secondary

easily whom

example, Individuals occupation.

self-employment

Zero-employee

firms are a large proportion of the total number they probably account for a small share Other than tax return data, the is from household surveys

of businesses in the US, although

of total revenues, assets, or employment. most complete coverage

of

zero-employee fines
Survey

such as the Current Population

or the Survey of Consumer Finances,

which provide data on self-employment.6 The DMI file's coverage of partnerships of corporations but better than undercoverage of partnershlps the undercoverage firms. 3.5 Desion of the DMI Sample To permit inferences from survey of that is not as good as its coverage The most severe Agaln,

of proprietorships.

is found In the real estate lndustry. may be

partnerships

associated with zero-employee

findings to the target population of

small businesses, probabilfty rauqllng was used to select the NSSBF sample. Probability sampling refers to sampling In which every unit on the frame Is given a known, nontero probability of
sample design was a

inclusion in the survey. The NSSBF

stratified random sample with oversampling to insure reporting domains deflned by Census

separate estimation capability for

region, metropolitan status, and number of elnployees.

61

For further discussion of Business Administratlon 1983,

firms without employees, see US Small 1986, 1988a.
18

To construct the sample design, we first obtained frame counts from the DMI file after fntllglblt business types and branch offices and subsidiary companies had been removed from the frame. Reporting domains of interest

to this study included business type, Census region, metropolitan status, age of business, trnploymtnt size and net worth. Tables 1, 2, and 3 present

some of the frame counts computed at this stage of study planning. The analysis plan for the study requires in depth analyses for each nonmetropolitan flms, ard for

Census region, for metropolitan versus small, medium and large sized businesses. in Table 1 revealed the obvious --

The distribution of firms shown

unless nonmttropolltan firms were

included in the survey In flnns, analysts will not

approximately the same numbers as metropolitan be able to make Also indtpth comparisons. of with 33 percent of all

metropolitan versus nonmetropolitan firms. records having missing responses, total stratification. An examination of

salts was

unsuitable for use in

Table 2, however, indicated that number

of employees was sufficiently correlated with total salts so that number of employees could serve as a useful substltutt for stratlficatlon purposes.

The uses to which the data art to be put in antitrust studies required that analysts be possible by slzt of fins by region of the country by

metropolitan status. Table 3

suggested that unless provisions were made, sample with Insuffic1tnt sample sizes

very small ffnm would dominate the

for separate analysts of moderate to large buslntssts. The final conclusions drawn counts for the NSSBF frame was from examining the distribution of fins

that moderate-sized firms (those with 50 to

99 employees) and large-size flnns (those have to be oversampled to for these two fin

with 100 to 500 employees) would

insure adequate analytical estimation capability (Sample size constraints

slzts by metropolitan status.

19

TARLE Frequency dlstrlbution of Sampling Universe (Frame) SBA Task Order by -Constructlon Manufacturing Transportation Communication Public Utilities hholesale Trade Retail Trade Insurance 8 Real Services other Urban 22021 537919 ::fz 9163 B909 Estate 301528 127343s

1. Survey of Small Business Flnsnces

for National EA-AB-52

Urban/Rural

Vat-tables --Rural --

--

Total 40337

--

3%::

16316 166424 82472 4549a 6367 5675 103392 401911 73859 329986

z65oJ:; 16l433 15530 14584 509317 '3fZ 1603421

&u~us

Northeast North Central South Nest

RepLnns

967614 891507 ':t:;ff

172823 387859 Z8Z

1140437 1279366 l::;f;:

S-25 earn Over 1; years 5 Unknown Emplovment ;r; 10-19 :t-;; lOOI249 250-499 Unknown G&%R%OOk 1OOk <= x <250k 25Ok <= x <SOOk 500k <= x < l,OOOk 1,OOOk <= x < 5,OOOk 5.000k <= x < lO,OOOk 10,OOOk <= x C 25,OOOk 25,OOOk <= x < 50,OOOk 50,OOOk <= x < 100,OOOk lOO,OOOk <= x < 250,OOOk 250,OOOkt Unknown Total
5004

429196 2393257 510023 619696 1939123 ;:6J;:r 247237 81980 45440 13069 12586 483209 594352 662797 445695 349546 411607 60427 393a4 11759 4742 2758 1161 1367344 3952172

104836 739474 191143 200447 703366 2K; :x ;31; 1687 120336 255521 :x 97421 'X2 ::2Jsb 554 302 130 359256 1235900

534032 JlJZIJl 701166 820143 2641509 973602 481471 303827 99lO? x: 14273 603547 849a73 912466 595133 446967 512200 74179 47020 t33ll7 5296 3060 1991 (726600 5198072

uZ;n

TABLE 2 Frequency
dlmfrlbufloaa of Sampling (Frame) Unl~~rso SDA Task Order Urban/Rural

for NatIonal EA-66-52
by Sales -Slro

Survey by --

of

Small

lhrrlna~s

Flnancos

Employment

Slzo

Urban

IOOk <= x c250k 250k <= x <500k 500k <= x < 1,OOOk 1,OOOk <= x < S.OOOk 5,OOOk <= x < 10,OOOk 10,OOOk C= x < 2S.OOOk 25,OOOk <= x c 50,OOOk 50,OOOk <= x < 100,OOOk 100,OOOk C= x < 250rOOOk 'u,'~~WJ~~k'

1 <= x C lOOk

?mz --u4 --%h KOSS
493256 207755 92462 %53 ::; -50 :: 550160 127065 94129 4406 1544 265

-=%3 4166
106626
20401 6736 1017 11724 36169

-=%
966 3447 32494 1%: 23:; '623 16635

q 500

149 250 664 10251 6634 'fl!f 1117 295 100'9::

=-3x 5:
9x: 1732 5597 29Jsz': 373 26:: 66 1: 4620:;

91 162lif -Rural

:it 529;: --

1 <= x < 1OOk 100k <= x <250k 25Ok <= x <SOOk 5OOk <= x < 1,OOOk l.OOOk <= x C 5rOOOk 5,OOOk <= x =Z 10,OOOk 10,OOOk <= x < 25,OOOk 25,OOOk <= x < 50,OOOk 50,OOOk <= x < 100,OOOk 100,OOOk C= x * 250,OOOk 250,OOOkt Unknoun

---hi& 'X 25411 11905 414 'Zoo 13 160265 1

-+-i?T 47977 54366 39962 2%t 2fS 10 45396 :

-@=&3 -=b 6648 1237
17114 ii::: 2127 676 S522 9947 2x 2::45 's" 1131;

=@% 45 1:; 394 is: 116 48 713

97%

1216 797 590 1:: 60 6

7: 7
1637:

;
11576:

-1 <='x < tOOk 1OOk <= x <250k 250k <= x GOOk 500k <= x < 1,OOOk 1,OOOk <= x C 5,OOOk 5,OOOk <= x < 10,OOOk 10,OOOk <= x < 25,OOOk 25,OOOk <= x < 50,OOOk 50,OOOk <= x < 100,OOOk 100,OOOk <= x < 250,OOOk 250,OOOkt Unknown* -#h 62x:::': 117893 53469 :'os5; 167 643 7104:: --%h 165729 X81: 116517 5444 1773 319 93 25: 227501

Total

--=% W v fS 109 1066 2126 4470 ::7970 421 56;:

--h

26394 73595 104706 150914 9636 3445 474 141 63

-9% 5403
15246 46136 ?K 10743 116i ::: 642::

594 1261 4504 39257 W 2:f: 203 2,oz

177 304 613 12659 Kaz 3693 1231 3:f 12767

4307 533 345 76 it I: 577712

i:::

1047::

1261fl lOlosl lb9L91 a&i

956s %k?m

xi

OSbOl 98991 UCOSl 2gAff(jr

152912 630SZS 109822 *

-'5662 10162 lb8SC w 90s to11 9bL +tkiicz 0801 IlSftS 5262 w=m 9SS2 1699 ZSSS wrz-

le)ol

--

S6SP Dcrb22 5299) gyf&--Iemu --

29S91 62SOS s1002 =&j-r

0916s 16506 902f9 -L=f&-

019221 69C192 159122 w-r-

lSS2 os9s 012s q&&g=

OLS6 OS021 S2601 wka-r

fSOU1 OOLSZ 2St61 ug$g-

SSOSS K y&-"PqJn --

SlSbO 959Lll LZSbO 4"&gi-

lbOLL1 8SCbSZ SL9191 =H---

l c

3wv.L

did not

allow

for

separate analysis capability by

Census reglons for

moderate to large firms.) was expected to be

In addition, approximately equal sample sizes for small flrms in metropolitan versus

needed

nonmetropolitan areas of the four Census regions. The sample frame was partitioned on the basis of Census region

(Northeast, North Central, South, and West), urban/rural location (firms in KAs are classified urban; all others are classified rural), and fins size

(small = I-49 employees, medium = 50-99 employees, and large = 100 or more employees).: DMI frame counts for these strata and their allocated sample

sizes are presented in Table 4. The choice of strata and allocatlon of completed interviews to strata

was based on consideration of three constraints: . . . the size of the final sample should be 4,000 completed interviews; domain estimates for small firms for each region and urban/rural location should have approximately equal precision: and domain estimates for medium and large firms in urban and rural locations should have approximately equal precision. strata, allocatlng 400 completed interviews by urban/rural location category, in each urban/rural location large firms in each urban/rural firms

The decision was to create 28 to small firms in each tOOcompleted interviews to

region medium

category, and 200 completed interviews to location category. For medium

and large flnns, the completed Interviews

were allocated proportional to population size within each Census region to reduce the effect of unequal welghtlng. Simple random sampling was used to select firms from each stratum.

L!

As mentioned, the DMI file contains records In which the number of employees is missing. For purposes of sample selection, these firms were classified as small. 23

Table 4.

Reglon, HetropolItan Status, and Finn Stze

Sample Allocation Across Domafns Deflned by Census

Census

Region
Northeast

Metropoll tan Status Urban

Size

F1l-W

Count

DMI Frame

Allocated

Sample
Size

Medium Large North Central Urban Small Hedl urn

Small

926,956

20,911 19,747 19,332
17,215

400 z: 400 47 48 400 z: 400 : 400 :E 400 :;

054,960

Large

South

Urban

Small Hedtum Large Small Hedlum Large Small Medium Large Small Medium Large Small Medium Large Small Hedlum Large

1,104,141 23,700

19,921

West

Urban

913,040 18,037 14,212 167,650 2,732 2,441 378,395 5,352 4,112 442,133 6,697 5,793 216,681 2,346 1,568

Northeast

Rural

North Central

Rural

South

Rural

400 ii 400 f:

West

Rural

3.6 DMI Sample Selection
To begin discusslng sample contained an unknown number included: . . . . flnas no longer In business, subsldlaries and branch companies that were not so identified on the DMI file (through error or a change In ownershlp), not-for-profit firms and publically owned Identified via their SIC codes, and firms that could not be of selectlon, lncllgible we should note that the frame flnns. These Ineligible flnns

finTis witr LOO or more full-time equivalent employezs in 19&7. screened presence to of determine frame if they were eligible for the does not bias study

Sample firms were study. results. Not knowing the Hence, the

ineligibles

percent

of

study

ineligibles to

In

the

frame led'to

uncertainty as to how many sample.

selections

make to yield the desired final

To deal wlth thls uncertainty, the

we decided that a much larger than

needed sample would be selected from

DMI files and then subsampled to
The initial sample

yield the required number of cases for NSSBF screening.

selection was done by SBA based upon specifications provided by RTI. With over five million records In sampling method that was both a the DMI frame, we needed an inltial simple and suitable for

computatlonally simple

further subsampllng.

We chose

random sampling method within 24 status and firm size. Frame

strata defined by Census region, metropolitan

counts were first computed by SBA for each of the 24 strata. point was established for each stratum

Then a cutoff

as the desired initial sample site (The desired tnitial sample size was

divided by the stratum frame count. set at approximately study.) A random number the sire was

we antlclpated would be needed for the for each frame record and then

generated

25

compared to its stratum's cutoff

point.

When the random number was less

than the cutoff point, the frame record was included in the Initial sample. When this type of sampling is used, the stratum sample sizes are random varfables rather than ffxed quantities. present a problem since this needed wfth initial Thfs variable sample sfze dfd not sample was intentionally larger than For each stratum, Table 5

an establfshed plan for subsampling.
counts, the desfred

presents the DMI frame

sample sizes, and the sample

sires that were actually selected for the fnitfai sample. The allocated sample sizes completed Interviews. shown In to from terms of the desired number of

To convert these

the number of selectlons to be the December 1987 DMI File, we Further, we

made from the inftial sample selected had to factor fn the effect had to have a plan in

of nonresponse and inelfgibility. place should our sample

size estimates prove

insufficient. The data collectlon strategy that we adopted was this. wave sample size was set so that if response and eligibility as expected, three data collection purposes. The fourth wave was The data waves screened total Data collection rates occurred

would be needed for intervfewfng to serve as a backup should is presented in about 2,000 DMI

supplementation be needed. Table 5. We selected four

sample

screened waves of

collection

lfstings each. The next Issue to be addressed for Is the method that we used in

subsampling the fnftfal DMI sample

allocation

to waves.

For each of

the 24 strata, we sorted the records by the value of the random number that was generated for the record and led to its selectlon for the lnitlal DMI Then usfng the wave sample size for

sample (from lowest to hfghest value).

26

Table 5.

Allocatlon to Strata for the Various Stages of DMI Sampling

Targetted Census Region Northeast Metropolitan Status

In1 tial

Actual

In1 tlal

Screen1 ng
Sample Size

Fll-Ul Size

Sample Size

Sample Size

Interview
Sample Sire

Urban

Small
MdiUlI

Large North Central Urban Small Medium Large Small Hedlum Large Small Medium Large Small Hedlum Large Small Medjum Large Small Medium Large Small Medium Large

1,200 153 168 1,200 141 144 1,200 174 168 1,200

1,210 153 157 1,224 152 162 1,176 140 158

800 ;: 800 128 120
800 156 168 800 56

782 102

110

782 94 94
782 114

South

Urban

110

West

Urban

132 120

1,139 147 121 1,203 t: 1,185 190 187 1,224 263 247 1,212 t:

48

782 86 78 782 ;; 782 125 117 782 153 164 782 ::

Northeast

Rural

1,200 1:: 1,200 189 177 1,200 234 249 1,200 81 69

800 104 112 800 if
800 116 112 800 t:

North Central

Rural

South

Rural

West

Rural

the stratum [call this WSS(h) for
W%(h) records to Wave 1, the

the h-th stratum], WC assigned the first WSS(h) records to Wave 2, the third

second

WSS(h) records to Wave 3, and the fourth W%(h) Midway through the data collection

records to Wave 4.
we evaluated whether we

period,

could complete the required number of interviews with three data collection waves and decided that sample sire. We supplementation was needed to achieve the desired cases from Wave 4 and then

subsampled

additional

designated these selections for interviewing. to sample for interviewing, we

To determine how many cases eligibillty rate

used our then-encountered

and assumed a 75% response rate.

The subsample of the screened sample used

for tntervlewing purposes Is included in the counts glven in Table 5. To select the subsample for intervlewlng purposes, we used the first Within strata, Wave 4 was

three waves and then subsampled the fourth wave. subsampled by first ordering the and then taking the with this procedure, desired records of

by their generated random number supplementary ftnns wlthln records. Note that

number

'Interview

sample" The

each stratum were

selected with equal probabllity. a strattfied simple random sample.

final sample may be characterfzed as

3.7 Constructlon of DMI Analysis Weights To make Inferences about the data, probability sample weights design. The weight of a sample Natlon's were small businesses from the NSSBF

constructed that reflect the sample can be viewed as the number of

business

mall

businesses in the population that

the

sample firm represents. The

sampling neight was constructed as the Inverse of the selection probability of the sample and, hence reflects the differential sampling rates that were

used.

This sampling weight was

adjusted

in

later steps to account for

28

sample businesses who could not be contacted or who refused to participate. The final analysis weights serves to differentially weight the sample

respondents to reflect the level of disproportionality in the final sample relative to the population of interest. The remainder of this section

discusses the development of the analysis weights. The probability of selection for sample record 1 from the h-th stratum

(h - 1, 2, .... 24) of the DMI frame was simply: P(hl) = n(h)/r?(h) where n(h) = the total records selected from the h-th stratum for the intervlew sample, and N(h) = the total DMI records associated with the h-th stratum. The sampling weight for the hi-th selection was calculated as the Inverse

of the probability of selection or Ws(hi) = N(h)/n(h). Note that l&h) Ws(hi) = N(h) all sample buttnesses in stratum h. the sampling weight had been

where ieS(h) denotes sumnatlon over This furnished a check to

insure that

constructed correctly. The next step was to correct for a nonresponse from eligible sample need to discuss how ineligible

businesses. To digress for cases were handled. business firms,

moment, we

First, in

screening we attempted to identlfy out of more than 500 employees, not-for-profit

firms with

businesses, government-owned entities, and subsidiaries and branches -- all of which were InelIgIble for the study. All designated Intervlew Sample

29

cases that were not

positively known to

be ineligible (i.e., screeningbrought into the CAT1 data base

defined eligibles and nonrespondents) were for interview. During the

interview phase, additional ineligibles were

later identified. The first weight that we with Section I respondents. constructed was A an analysls weight for use

weighting class adjustment procedure was The Section I weighting class

used to adjust for nonresponse to Section I.

adjustment for the c-th weighting class was calculated as Al(c) = WS(hi)/ X Whi) I: hi&E(c) hicSRl(c)

where SE(c) is the set of eltgible sample businesses belonging to weighting

class c and SRI(C) is the to weighting class c.

set

of eligible Section I respondents belonglng

Section I respondents in

the c-th

weighting class

were assigned a

weighting class adjusted weight calculated as the product of their Section 1 adjustment factor and their sampling weight or

h&l)

= Al(c) WsW).
I nonrespondents were assigned an

For computation convenience, Section

adjustment factor of zero and an adjusted weight of zero. The next issue is the classes used In making thls Section I nonresponse adjustment. Candidate variables included Census type region, metropolitan

status, fins size, SIC classification, and

of business. Except for

the last vartable, these variables were derived from the DMI file. Type of business (proprietorship,partnershlp or screening and/or during the intervtew. The candidate variables were examined to determine the extent to which they were defined for nonrespondents as 30 well as the extent to which they corporation) was obtained during

cxplatned Section I nonresponsc rates and size, SIC group, metropolitan status and define Section I weighting classes.

responses to data items. FlW Census rcglon were chosen to

Collapsing over Census regions was

needed for medium and large flws due to their smaller sample sizes. The next nelght that we constructed was an analysis wetght for use with

Section

II

respondents.

(By deflnltlon, all

Sectlon

II respondents

responded to Section I).

Again, we calculated welghtlng class adjustments

for the c'-th welghtlng class as Am C YI(hi) / I: hieSRz(c') hlcSRI(c') WI(hl)

where SRz(c') is the set of ellglbles from weightlng class c' who responded to Sectfon II. Section 11 respondents ln the c'th weightlng class were assigned a

weighting class adjusted welght calculated as the product of their Sectlon

II adjustment factor and their Section I analysis weight or
WZ(hl) - AI WI(ht). adjustment factor of zero and a

Section II nonrespondents were asslgned an Sectlon II analysis weight of zero. Candidate variables for forming plus

weighting classes number of organlzatlon

included the

prevfously mentloned variables employees (I-A5 + 0.5 locattons (I_D4).

full-tlme equivalent

I_A6),

type of

(I-Cl) and number of

Ftnn size, SIC group, metropolitan status, Census regton
chosen to define welghtlng classes.

and corporate/noncorporatestatus were

Collapsing of small class occurred for the last two variables, particularly for medium and large sized firms. The last weight that we constructed was an analysls weight for use with Section III-V respondents. (By deflnltlon, all of these had responded to

31

Sectlons I-II.)

The

weighting class

adjustment for the c'-th weighting

class was calculated as Aj(c') = I: W2(hi) / I: W2(hi) hirSR2(c') hitSR3(c')

where SR3(c') is the set of ellglbles from weightlng class cm who responded to Sectlon III-V. Sectlon III-V respondents In the c'-th weightlng class were assigned a

weighting class adjusted welght calculated as the product of their Sect!on 111-J adjustment tdrtor and their Section II analysls weight or W3(h) = A3(c') W2(hi). Nonrespondents were again given adjustment factors and welghts of zero. Candidate variables for previously mentloned plus a fonalng weightlng classes Included those

type of cash of

financial services used recode (cash services and financing, neither cash

services only, financing only, services nor financing). Type

financial services used, firm slze, SIC region were used to define the last three

group, metropolltan status and classes. Collapsing of

Census

small classes occurred for

the

variables, In particular for medium and large size businesses. Three analysis welghts resulted from the weightlng process -- WTANALl, WTANALE and WTANAL3 -which are
I

to

be

used

in

analyzing data from
(or

Section I, Section II

(or Sections

and II data), and Sectlons III-V

Sections I to V data).

DMI data

can be analyzed using stratifled random

sampling procedures, the above weights, and the varlable STRATUM. STRATUM records the stratum membership of the sample business wlth levels l-24

assigned following the ordering glven in Table 4. 3.8 S8A Sample Selectlon and Weightlnq S8A furnished us with thelr file of guaranteed loans dlsbursed in 1986. For comparability to the DMI sample, flrans whose business was agricultural 32

or financial in nature were

excluded as well as not-for-profit businesses,

government entities, and branches and subsidiaries. Having removed ineligible businesses was sorted by Census and and duplicate entries, the frame

region, SIC

code

number of enploytes using a

serpentine ordering and then Chromy's sequential sample selection procedure
was used to select a sample of randomly allocated to 13 data ultimately 975 businesses.

These businesses were then Ten of these waves were

collection waves.

fielded for data collection ?qotes.
a self-weighting the SBA sample. of The probability of the sample can be

This procedure leads to selectfon of the I-th expressed as: P(i) = n/N where firm

from

portion

n N

is the nmber of records contained in the ten fielded data collection waves and is the total SBA frame file records.
I

The sampling weight for SBA sample f'im this probability of selection or Ws(i) = N/n. Nonresponse adjustments

was

calcu lated as the inverse of

were
I,

made

following the

procedure

outlined

previously to create Sectjon

Section sample

II,

and Sectlons the

III-V analysis

weights. Because

of

the

smaller

size

number of weighting

classes that can bc used was necessarily less. were defined using the weighting

Section

I

weighting classes

SIC group
cl asses

and Census

region.

For Sections

II

and

II-V,
and

were

defl ned

using

SIC

group

corporation/noncorporationstatus.

33

Chromy's sequential sampling algorithm, used produces an lntpllcitly stratlfied design analyze these data, we weightlng class with

to select the SEA sawle,

one selection pe.rzone. To WTCLASSl, the Section 1

suggest analysts use deffnt

variable, to

pseudo-strata.

StratIfled random

samplfng procedures would then be used pseudo-strata and the analysis wtlghts. As a final point, we note that be analyzed jointly. There Is an the

to analyze the SBA data using these

DMI data and SBA data should never

unknown amount of overlap betwecr, the the SBA sample. The analysis

target populatlons for the DMI sample and

weights apply only for separate estimation for each of the two samples and do not allow the data to be combtntd.

34

4.

QUESTIONNAIRE DEVELOPMENT

The analytic goals

of

the

NSSBF

required the

collection of large

amounts of financial data.

Accurate responses for the associated questions

depended upon the respondents understanding of terminology from accounting (e.g., nonrecourse loans, trade services providers (e.g., notes, treasury stock) lock and financial

capital

leases,

boxes, banker's

acceptances). Access to

financial records also enhanced the accuracy Jf these response issues was limited so study

responses. Information about

planners decided that a more elaborate than usual questionnaire development actlvlties were in order. Questionnaire development questionnaire created by survey data. the for the NSSBF began with a draft

Board to

satisfy their analytic needs for

This questionnaire was refined using four rounds of cognitive major revisions of the questionnaire at

testing and a formal pretest with each step in the process. are displayed in Exhibit 1. The

topics covered by the final questionnaire

4.1 Questionnaire Formattlng and Validation The first step in the questionnaire development process was to reformat the Board's draft questionnaire to create a survey instrument. Steps in

the process included establishing the

logical ordering of the questions,

specifying skip patterns and defining the temporal basis of the individual questlons. As examples of the latter, some questions had tobe answered as of a particular date (e.g., saving account balances, amounts owed) while period of time (e.g., total sales, Interest

other questions referred to a expenses).

35

Exhibit 1. I.

Structure of the NSSBF Questionnaire

Characterlstlcs of the Firm Screening Informatlon i: SIC Code C. Organlzatlon Type 1. Proprietorships 2. Partnershlps 3. Corporatlons 0. Location of Firm E. Reference Period (time perjod to which data apply) Sources of Financial Services A. Use of Oeposii Services 1. Checking Accounts 2. Savings Accounts 8. Use of Credit and Financing 1. Leases 2. Lines of Credit 3. Mortgages 4. Motor Vehicle Loans 5. Equipment Loans 6. Loans from Partners/Stockholders 7. Other Loans 8. Most Recent Loan from an Instltutfon C. Use of Other Flnanclal Services 0. Relatlonshlps wfth Financfal Institutions 1. Characterfstlcs of Financial Institution Locatlon Used for Noncredlt Services f* Location Used for Credit Services E. Other Financial Instltutfons F. Previous Relationshlps wlth Financial Institutions G. Shopping for Financtal Services H. Solfcltation by Financial Instltutlons Use of Trade Credit :: SBA Borrowing Experience (SBA sample only) Income and Expenses Balance Sheet A. Assets 8. Liabflities and Equtty C. Miscellaneous Closing Remarks (Use of Records)

II.

III. IV.

V.

At this point, our emphasis was on verifying the accuracy and integrity of the survey's questions, not on creating questions that survey

respondents would understand. For accountants to ensure that that the questions were

instance, the document was rtvftwed by

accounting conventions were being followed and appropriate for the diversity of accounting

practices that would be encountered in also reviewed by subject matter

the survey. The questionnaire was verify that all required

specialists to

data for study analysts were btfng collected by the questionnaire. 4.2 Cognitive Testfng of the Qutstfonnafrt Our next step was to focus on the ability of the small business to

provide the rtqufred information.

The survey object1vts required detailtd

fnfonnation on small business ffnancts including an inventory of assets and liabilftfts of the business, a instftutfons used services within description of stwicts, The the types and locations of the clustering of these Involved technical

for

financial
same

and

the

supplier.

questions

concepts and requested confidential information about the ffrm, and in many casts required that answer. To test the workability of address the following issues: . . What records art readily available to What data do they provfde? the small business owner? the Board's draft questionnaire, we had to the small business consult records to supply the

business owners How knowledgeable art about their firm's relationships with ffnancfal institutions and alternate sources of financing? To what extent do business owners understand accounting concepts embodied in the survey qutstfonnafrt? For what date/time period, can fines respond most accurately about their financial operations?

. .

37

.

What impact will advance copies quality and response?

of

survey questions have on data of the business

. .

What are the specific confidentiality concerns owners and how can these concerns be addressed?

How can we motivate businesses to participate in the survey? to these broad issues, we also needed to access the

In

addltion

respondent's understanding of each individual questlon. To obtain data to answer these basic testing. in the the questions about the interview Prior-to each rcund, curvcy local area to Identify 5-g process. Businesses of and

process, we conducted four rocnds of specialists contacted small businesses who would agree differing types to participate in

testlng

(e.g.,

services,

manufacturing,

transportatlon)

differing organizations were recruited respondents. in an

(e.g., proprietorshtps, attempt to

partnerships, corporations) the diversity of study

characterize had to

Considerable effort

be expended to identify wllling to spend the necessary time

participants as many businesses were

hesitant

for testlng or to answer the kind of questions that Businesses who agreed to place of work. 'Think-aloud' talk out participate were

would

be asked.

visited in person at their

Intervlews were conducted where respondents loud about how they were interpreting the with coming up wlth an

were encouraged to

questlons and their thought answer.

processes

associated

Reluctance to respond was also observed during the testing process respondents us off initially when but we agreeing arrived. to to participate and then Others responded to the provide income statement or

itself wtth some

refusing or putting

majority of he questionnaire balance sheet Information.

refused

With the respondent's permission, This allowed the interviewer to

we

recorded the testing Interviews.

focus on encouraging respondents to report 38

what they were

thinking or

doing.

The

recordings were

rev1twtd by

questionnaire dtvtloptrt to determine the

strength and weaknesses of the

draft questionnaire. Based upon the results, the questionnaire was revised in preparation for the next round of testing. showed the questionnaire was longer than rtvlsions also involvtd ~dfffcatfons
time. Study plans called for mailfng worksheets to sample businesses prior to the lntervfew and urging them to complete them in advance of the inttrvltw.

Testing results conststtntly desired 30 mlnutts so these

the

designed to cut the adminfstratfon

Thus, part

of

the

response

process

for

the

study

occurs

before the

inttrvltw itself.

This process was

observed

in Round 3 of testing where

businesses were shown draft worksheets to Information and asked about their

be used for recording financial

inttrprttation of

the

form and the

records that they might use to

complete ft.

Prior to Round 4, we mailed to participate. somt

the worksheets to

the

buslntssts who they

agrttd

businesses refused after

received the

worksheets, citing their

reluctance to answer qutstlons about their business finances. As should already be clear, the testing process Indicated that the

subject matter of the survey was expected to be reluctant to testing were as . follows:

highly sensitive and businesses could be the study. Other results of

partlcipatt In

Some questions were

best answered by accountants/bookkeepersand WIthin logical sequences of questions, both others by owners. types sometimes occurred.

.

Even for accountants/comptrollers there was confusion about the meaning of ttnns describing financial strvlcts (e.g., banker's acceptances, zero balance accounts, lock box). Rounded figures were often given for amounts even when the respondent was looking at records with exact figures.

.

39

.

There was a tendency to describe personal finances even though emphasized. For the smallest business-only finances was
businesses, co-n. comingling of personal and business finances was

.

you mean by ..: a definition. .

Balance sheet and income questions frequently provoked a, 'What do

response that put testers on the spot to provide 'I'd have to look up that number.'

Many questions provoked a, response.

.

list all accurately Respondents not financial could A better approach appeared to institutions/credit sources used. be to ask if the firm had any use of the service and then to ask for specific tnstltutions/sourctsused. Some concepts remained problematic such as the number of full time equivalent employees, use of trade credit, and the distinction between capital and operating leases. the questionnalre were addressed by Important issues not identified as well as skip pattern problems, repetitive questlons, etc. Sensitive questions were highlighted with unexpected results at times. For instance, respondents sometimes gave amount owed for individual loans but refused to give aggregated total. Other respondents stated that their company did not give out balance sheet data to anyone but when asked the questions would answer yes/no questions and give some amounts. The recent tax law affected the reporting period by creating partial years for firms forced to convert to a calendar year basis. In revising the questionnalrt after each round

.

.

.

.

This information was used

and in preparation for the pretest. For instance, respondents with accounting backgrounds insisted on short questlons based on accounting terminology while less knowledgeable

respondents needed longer questions with solve this problem, we

definitions embedded in them. To

created short questions that only knowledgeable

respondents could answer. followed up with a probe

If
that

not

imntdiattly answered, the inttrvltwer

contained a definition. Instructlons were

40

also given

to

provide guidance in glossary of

separating business and

personal

finances. In addition, a use in interpreting the

terms was provided for interviewer answering questions from

questionnaire and

respondents. Another problem was to develop a structure for collecting reasonably in the sample. Several obstacles

consistent financial data from all firms were encountered.

Accounting practices, developed for controlling the

operations of business firms rather than reporting econcmic value, are not uniform. Larger firms tend needs and thus more to have greater coordination and monitoring smaller firms. Similarly,

complex records that

partnerships and corporations tend to

have more sophisticated records than

proprietorshlps. Indeed, a proprietor's business and personal finances are not normally separated. Federal tax returns provided a nearly cormnonbasis for reporting balance from sheet the items tax for partnerships and corporations. Identlfled for each item

Appropriate lines

forms were

requested in the worksheets and questionnaire.! Proprletorships, however, do not normally prepare balance sheets. how to construct a balance sheet. Some respondents were unwilling to report aggregate debts or cash They were given instructions on

holdings in a balance sheet, even

though they had reported dollar amounts

of indlvidual debts and accounts. The order of questions was based on this experience. Questions about general characteristics of the firm came flrst and then questions about the sources of financial services. The Income Thus, for some firms, partial

statement and balance sheet were at the end. balance sheet information would be

available from responses to questions

about financial service use even if balance sheet questions were refused.

s f

Respondents sometlmes had to add two or more lines or to disaggregate Items reported on a slngle line on the tax form. 41

4.3 Pretest Results A formal pretest was conducted for sample businesses selected from the

DMIand SBA frames. The same target population definition was used for the
pretest as planned for the main study. Our testing experience prior to the pretest indicated that response patterns and difficulty in responding

differed by organization type. samples evenly across the

For this reason, we distributed the pretest

three organization types -- proprietorships,

partnerships and corporations. The SBA frame choose five of indicator was had an organization type for screening. indicator which we used to For the

each

type

DMIframe, no such
the fins's

available.

Instead, we

used

the

title of

representative taking all records with

the

title of partner or proprietor

and choosing a representative mix from the rest to get a total sample of 59 firms from the DMI frame. The interviewtng planned for the main An initial telephone call was they were eligible for the made study to study was tested in the pretest. sample businesses to determine if to verify name a lead and address

and

information. containing .

Eligible businesses were

mailed

letter package

a letter from Federal Reserve Board chairman Alan Greenspan, and Introducing the study and urging another. letter from RTI participation, a brochure wlth answers to questions about the study, a worksheet for recording income statement, balance sheet and financial services data prior to the interview, a postage-pald envelope for returning the worksheets, and a return post card so that the business could advise us of the best time and place for Interview.

. .

. .

42

Results of the screening are given To insure equal spread of the a

in Table 6 by type of organization. intewlews type across the three

pretest

organization types, we added

question on

of organization to the

screener. For completed screeners this response was used to categorize the business. All of the SBA firms were eligible for the study. Screening Interviews were completed with 73.3 percent or parttal data, one flnn eleven firms, with two firms providing one firm with no answer. The

refusing and

organization type variable on the SBA frame appeared to be out of date with more firms listed as proprietorships and partnerships than reported

themselves to be corporations. Two of the DMI firms were not-for-profit and the other ineligible types of businesses -- one was was a subsidiary. Another 13.6 percent of For this latter

the DMI firms (eight firms) had

disconnected numbers.

group, we implemented extensive telephone tracing attempting to get a new number for the firm or its representative through directory assistance, The 13.6 percent that remained

contacting the Chamber of Consnerce, etc. untraced after this procedure are most treated as ineligible bringing the percent.

likely out of business and can be

total percentage of ineligibles to 16.9

Screening interviews were completed with firms (39 fines) with Interview. Of the an

79.6 percent of the eligible

additional 4.1 percent partially completing the had two refusals, one callback four wtth no answer

remaining firms, we

appointment never kept, one after repeated tries. For

answering machlne, and

the

main study, callbacks, answering machines,

and ring no answers would be worked longer than we did for the pretest (one day) and so the response results are somewhat atypical. 43

Table 6.

NSSBF Pretest Screening Results

Screening Component

CM1 Sample

SBA Sample 15 0 0 11 : 7 2 1 0 0 1

Overall Total 74 8 2 50 :z 23 4 3
1

Total Firms Screened No Telephone Lfstlng Complete: Complete: Ineligible Finns Eligible Firms

59 8 2 39 15 1: 2 2 1 1 4

Proprietorships Partnershlps Corporations Parttally Completed Refusal Callback Appointments Answering Machine No Answer

1 5

Based upon the pretest experience, WC we would not attempt to convert a
mail

decided for the main study that
.

refusal to screening. Instead we would convert the fin, and if

then the lead

letter packet, attempt to to

successful we would then screen prior behind this approach is that

the interview itself. The idea

reluctant firms may become more receptive

after they receive the lead letter packet and hence should not be followed up too strenuously during screening. A: the result of the pretest, r;e found that hence thr name arm'address asked first. We

information was least

sensitive and

should be

deliberately chose to screen whoever answered the telephone and not ask for the owner In the pretest. employees and on We discovered that the questions on number of

not-for-profit, public ownership, and subsidiary status the person answering the telephone. This

could not always be answered by was the explanation for the

partial interviews. Based upon this finding,

we decided that interviewers should ask to speak to the owner (If possible) in the main study screening, verifying name and address data first and then asking the eligibility questions. After screening was complete, we identified the firms we wished to mail to for the pretest interview. Using titles of corporate officers and the 51 DMI firms and 9 SBA firms, numbers for the three

SBA organization type variable, we trylng to come as close as

chose

possible to

equal

organization types.

Lead letter packets were mailed to these businesses. the lead letter packet number to so that the firm call. Only two

Postcards were included in could indicate the postcards were best

time,

person and by a

returned, one

firm that

ultimately refused to

participate. This experience suggested that

it would not be worthwhile to

include return postcards in the main study lead letter packet. 45

The lead letter gave

the

RTI

Project

Director's name and telephone

number to call if the respondent had questions. Two respondents called -one to request another copy of the worksheets. With both firms it appeared that their true purpose was to set if we were legitimate. (Roth

participated and were easy interviews.) president, George Herbert, sign RTI's

For lead

the main study we had RTI’ s The letter gave the

letter.

name of a survey specialist to call if they have questlons.

Sample firms were mailout. As a part

called approximately 2-3 of the introduction, the
letter packet.

weeks

after tre pretest

interviewer asked if the
Packets were remailed to

respondent had received the lead firms who indicated that

they had

not

received the packet of materials.
it but threw it away. the package Several

Some firms indicted that they had

received remail

of these firms

insisted

that

we

before they would

complete the interview.

The control

sheets show nine remailings, two with

ultimate refusals, five with lntervlews completed, and two that were still being worked when data collectfon ceased.
The experience with the that firms should be called lead soon letter after package in the pretest suggested they receive the package and then on general characteristics and to establish rapport with the

the interview schedul td later. ownership could respondent. be asked at

Questtons that time

With this approach, the respondent can be notlfied in the lead

letter that we

will

call

In

'the next Note

few

days"

and

may retain the

materials until he/she talks to us.

that it does not make sense to
it

enclose a postcard with this approach since call. Table 7 presents the pretest

will not arrive before the

Interview results.

It should be noted

that we chose the most convenient stopping point after we had reached 30 46

.

Table 7.

NSSBF Pretest Intcrvlcw Results

Interview Component

OMI
Sample

Sample

SBA

Overall Total

Total Hallouts Out of Business Unable to Locate rompletcZ Intervirus Partnerships Proprietorships Corporations Partfal Interviews Refusals Answering Machine Not Ffnallred

51 4 3 26 6 I3 2 13 1 3

9 0 0 : 1 3 0 2 0 1

60 2 3 33 : 16 2 15 1 4

completed interviews.

Therefore,

we

had

some

put

Offs,

refusal and

breakoff cases that had not been finalized yet. From screened firms with SEA guaranteed interview. Six firms completed the loans, we chose nine firms for

interview, two refused and one was not One of the completed

finalized at the time data collection was terminated.

interviews was conducted with a firm that initially refused to participate. Uailouts were made to 51 firms from locate three firms and another four flnas the DMI sample. We were unable to These

had gone out of business.

cases had been deemed problematic in screening. confirm our belief that firms the company were ineligible. without For two

We followed up on them to

directory listings for the owner or firms, the designation of 'out of

business' was chosen because the lead letter was returned as undeliverable. For the other two we received For the confirmation during interviewing that they unable us Most their to locate firms, we had two firms but the owner did not contact us likely these are firms that had mail forwarded to an owner. that address The

were out of business.

where the mailout was not returned to and the firm could not be traced. have

gone out of business and now uncertainty that existed here should be sent out marked

suggested

the main study lead letters requested'. (This was an

-forwarding

oversight for the pretest.) Responses were obtained from 59 percent of the identified or 26 firms. Six of these fim

DMI eligibles

had initlally refused to participate, of identlfied eligibles) some of six firms (13 percent of

Remaining were 13 refusals which could have been

(29

percent For

converted.

another

eligibles), we had partial interviews, otherwise had not finalized them.

answering machine contacts only, or of the latter were put-off type

Some

refusals but still might have been converted. 48

An extraordlnary amount of

telephone calls was

needed to establish

contact and attempt an interview. of contacts was 6.4 and

For the SBA sample, the average number to 16. For the DMI sample, the from 1 to 17. Note that

ranged from 2 and

average number of calls was also 6.4

ranged

these calls do not include the calls made to complete the screener. We also Implemented two mlni-experiments as a part of the pretest. We

were concerned that our normal interviewers might not be able to master the business terminology being used or cope with the questlons that respondents mlght ask. To try this out, we recrufted an intervlewer wlth a business an intervlewer), one inexperienced

background (but inexperienced as

interviewer wlth no business background, and two experienced interviewers. Basically, we found during the pretest that the glossary of terms and

the deflnitfons embedded in the questions were sufficient in explaining the accounting terms although some business background was helpful. However, the dlfflculty of converting reluctant respondents and cajoling other

respondents to resolve mlstakes convinced us that our originally proposed strategy is the best. That is, that the main study should concentrate

efforts on hiring experienced Interviewers (to the extent possible) and also interviewers good at nonresponse conversion. To fill the cadre of

Interviewers, some Inexperienced interviewers might have to be hired. Here the search should be for a business background and good Interviewer traits. Another experiment that we proposed plan to ask trled durlng mail the pretest concerned the the worksheets. Some

respondents to

back

uncertainty was felt by survey specialists as were going to request the

to whether knowing that we

worksheets might make respondents reluctant to

49

partlclpatc In the

telephone Intcrvlcw.

For

this

reason, we enclosed

postage-pald envelopes In 25 of the that they were for the return of end of the Intervlew. For the

mallouts and noted In our lnstructlons

the worksheets and other materials at the remalnlng mallouts, we dld not enclose an returned to

envelope or indicate that materials should be conclusion of each business' Intervlew Indicated In the lead letter), we

us.

At the

(regardless of whether mallback was

requested that worksheets and other

records be returned to us. It dld not seem to matter whether we requested the mallback or not.

In

all,

we got records back from three respondents, one of whom had not been
Three respondents refused to return their all or promised to return them

sent a postage pald envelope.

records. Most respondents sald nothlng at and then dld not. to a declslon to

Knowing that advance warnlng would not hurt response led beef up the lead letter and the main the closing script to

encourage return of the

worksheets In

study and to enclose a

postage-pald envelope. The

pretest experience made clear, however, that

the majority of respondents would not return their worksheets. Some parts of the pretest Intervlew went well and some parts remalned

problematic. We examlned all of the pretest Intervlews and generally found that the lntervlew worked. However, we had difficulty obtalnlng accurate of flnanclal services and for the Income We also found durlng the pretest

responses about the flnn's use

statement and balance sheet questlons. that some respondents were sheet sections. For the refusing to

complete the Income and balance

maln

study, extensive converslon attempts were

clearly lndlcated to complete as many of these cases as possible. The major problem In complete the worksheets. the pretest was that many respondents dld not

Hence, dollar amounts were often rounded and 50

inaccurate. taken for the

As a result main study

of to

the pretest experience, several measures were increase use of the worksheets. First,

respondents were contacted much sooner after the mailout packages were sent to reduce the number of lost and discarded worksheets. worksheets to firms that lost, discarded recontacted respondents after or Second, we remailed

did not receive them and then Finally, when interviewers

remalling.

discovered that respondents had

not completed worksheets,

respondents were

encouraged to do so, and the interview rescheduled for a Iater date. With respect to use of financial services we had respondents giving us Clearly their memory was faulty

estimates and not checking their records. at times and even more disturbing, interest and hence missed some and

they neglected to discuss all events of institutions. data collection For the main study, we

redesigned the questlonnalre this underreporting. Two problems statement. appeared to

procedures to minimize

be

happening dfd

wlth

respect

to the Income

First, some respondents so a taxes left

not have their taxes completed at of the questions blank. (This

the time of the pretest and would not be expected to be all finas will have filed

many

problem by

fn the main study since presumably October.) Another problem that this section was

late

occurred was obvious erroneous responses. so trim that there caused the problems. were not sufficient

Unfortunately, redundancies

to figure out what

To correct

the problem, we built in more definitions

of terms in the main study questionnaire and added checks to spot erroneous responses as they were reported. the respondent to correct the The Interviewers then attempted to get (e.g., profit unequal to

obvious discrepancy

total sales minus cost of goods sold).

51

The problems with the A number of the balance

balance sheets

sheet questions were more approachable. did not add up because liabilities plus In other

equity was given instead of the cases, a misunderstanding adopted for the main study identify the discrepancy section and the take the each answer. had was and

total clearly

liabilities we requested. occurred. The

approach that we

to include strategic questions designed to extensive CAT1 edits to spot the problem

respondent

through the entire section confirming

The most disturbing result of the pretest was the length of time needed to conduct the pretest interview. was 52.3 minutes and ranged from sample, the average minutes. time was 37 43.2 For the SBA sample, the average length to and 67 minutes. ranged from For the DMI 20 to 124

minutes minutes

Only 9 of the 25 DMI

firms with interview times (1 firm had time 30 minutes or less. These results do Another disturbing

missing) had an interview length of

not include the time required to complete the screener. event observed in the pretest was that

the length of time to complete the This suggests that

interview kept increasing as data collection continued. if we had continued the pretest the until average we

obtained the 75% response rate intervlew ength would have been

desired for the main study, even greater.

Prior to the main study, we the problems described above. of the Board and SEA on the

revised the pretest

uestionnaire to solve

We also added a few questions at the request characteristics of the firm's most recent loan

and a few other miscellaneous questions.

52

5.

DATA COLLECTION

Based upon the results of the four rounds of testing and the pretest, a data collection strategy to evolved. The study approach eligibility began and with a short

screening interview mailing address. the business.

determine

to confirm the

A lead letter After a ten

package and worksheets were then mailed to day delay, the business was contacted by

telephone and questionnaire

data

collected including worksheet responses.

At the conclusion of the interview, the interviewer asked the respondent to mail the worksheet and the records they used in answering the questions.

This section discusses the results of this data collection approach. 5.1 Screening Results Screening interviews were conducted determine eligibility and to telephone number, and address. Appendix A. Because the eligibility rate waves. was unknown, screening was scheduled in correct The for DMI and SBA sample firms to

inaccurate information on ownership, form used for screening is given in

Each wave was a random sample.

Thus, estimates of eligibility from

the early waves could be used to select a sufficient number of firms in the fourth wave to yield the desired number of eligible firms. Many telephone calls were needed to complete the screener even though An average of 4.3 calls was made

any knowledgeable person could respond. for each business selected for screening screenings, 4.9 calls contact firms. for

with 3.1 calls made for completed and 9.6 calls for unable to

nonrespondents,

53

A total of 8,017 firms from the

DMI frame were screened (Table 8).

of

these firms, 66 percent were determined percent remained indeterminable. Virtually all of the ineligible categories.

to be eligible and an additional 5

firms

from

the

OMI

frame can be

classified into one of two ineligible types of

The first category consisting of 12 percent of screened firms.

businesses,

contains

About 57 percent of these

ineligible firms were not-for-profit or publicly office of the firm, 9 percent had more

owned, 28 percent were not the main

than 500 employees, and 6 percent were subsidiary companies. The second category contains About 17 category. percent For 11 of the OMI of firms sample the that firms, not that was were no longer in business.

in this no-longer-in-business finns, was the owner or another out of business. For

percent

sampled the no be firm

knowledgeable person reported another 6 percent of sampled company and the business could executive, the Better

directory listing existed for the traced the through calls to their top Chamber of Commerce, local

Business

Bureau,

libraries, and the telephone company. Similar results were obtained for the SBA sample. from the S8A frame were screened (Table 8). A total of 675 firms

Of these firms, 83 percent

were determined to be eligible fur the indeterminable. 5.2 Interviewing and Response After completion of screening, lead screened eligibles and to

study with an addition 3 percent as

letter packages were mailed to all in the data collection waves

indeterminables More

designated for interview. to be needed.

businesses wee screened than were expected

As discussed in Chapter 3, the screening sample for each

54

Table 8.

NSSBF Screening Results

Category

DMI Sample Sample Percent Size (X1 8,017 5,669 2,348 2,348 969 494 2:: z; 100.00 70.71 29.29 100.00 41.29 21.04 2.47 11.58 2.47 3.71 39.22 19.25 0.26

SBA Sample Sample Percent Size (X) 675 581 94 94 19 5 1: 0 0 41 34 0 100.00 86.07 13.93 100.00 20.21 5.32 1.06 13.83 0.00 0.00 43.62 36.17 0.00

Total Firms Screened Eligible or Indeterminable Ineligible Firms Ineligible Firms Ineligible Business Type Not-for-Profit Publically Owned Branch Offlce Subsidiary 500 or More Employees Confirmed Out of Business No Listing, Cannot Be Traced Other Ineligible

921 452 6

Eligible or Indeterminable Screening Completed Partial Screener Refusal Language Barrier Listed Number, UTC Other Eligible

5,669 5,280 79 125 15 17 151

100.00 93.14 1.39 2.20 0.26 0.30 2.66

581 562 7 3 2 0 7

100.00 96.73 1.20 0.52 0.34 0.00 1.20

survey component that can be

(DMI and SEA) was as a

subsampled to yield an interview sample random sample from the respective

viewed

stratified

sampling frame.

The packages contained letters from Alan Greenspan and the the survey, a question and answer for the survey, and the as Appendix 6. The

president of RTI urging participation in pamphlet explaining the purposes worksheets. Copies of these and

procedures are

materials

included

worksheets used 1987 income tax returns as a guide for answering the income statement and balance provided in Appendlx C. Ten days after respondents. packages had were mailed, interviewers began telephoning that most respondents could not sheet quest ons. Copies of these tax forms are

The pretest

ind cated

complete the interview in one session. interviewer ask the general and then break off the worksheets.

Initially, we planned to have the

questions about business characteristics first if the respondent had not completed the indtcated that the majority of sample respond worksheets who or difficult to reach. In

interview

Early results, however, reluctant to

businesses were very

addition, not having completed the "put off" technique by many firms questions about interviewers were business

was clearly being used as a

were not even answering the general To alleviate only this problem, when they were Otherwise,

characteristtcs. to allow to a

instructed

breakoff

convinced the respondent intended the interviewer was to encourage and continue the interview. Obtaining response was a

complete the worksheets. to

respondents

get any records they had

continuing for

problem throughout the interview (1) respondents did

process with two primary sources

the difficulty:

not want to respond to a survey that asked for such sensitive, confidential

56

data and (2) establishing contact with the owner and finding an appropriate time for interview was often problematic. was This difficulty in in establishing

contact and gaining cooperation

reflected interview.

the number of contacts

needed to complete screening and the 11.6 calls were made on made for totally average for

For the interview itself,

each sample business with 10.4 calls 11.6 to 20.3 calls for partial

complete

interviews,

interviews, 10.9 calls for total to-contact businesses. made during screening. These

nonrespondents do

and 19.2 calls to unable-

figures

not include the telephone calls

Since obtaining response was so

problematic

for this study, we set up the first for

four separate response status indicators for Section I where business characteristics

the interview:

were

discussed, the second for

Section II where the financial services the third for Sections III-V were obtained and the fourth where for

used by the firm were inventoried,

the income statement and balance sheet the questionnaire as a whole. For

completed questionnaires

and work in

progress, we evaluated the individual by tabulating the frequency of missing a percentage of the items in a

section response status indicators data in the data base. When too

large

completed section were missing, the respondent was telephoned to obtain the missing data. reopened cases. The data collection goal was to achieve a 75% response rate among study eligibles. Early in the data collection process, it became clear that Data were obtailied for approximately 60 percent of these

achieving this goal might not be possible. survey, respondents demonstrated extreme study. As an example, during the first

From the beginning days of the reluctance to participate in the

two weeks of data collection, we

57

averaged less than one completed 8-hour shift). Drastic measures reassignment Production The bulk and 50

interview were of

per interviewer dam worked (an

taken to boost production including changes in working hours,

staff retraining, incentives, etc.

personnel, markedly was

increased of to time

peaking at 3-4 completed

interviews per day. contact with

spent in trying to establish them to this participate. The

respondents

convince in

interview itself averaged

minutes data

length;

time to complete

together with the end-of-year difficult to obtain. Table 9 presents results

collection period made every interview

of

the

data

collection

process.

For this

study, a completed interview was defined to II were completed (the bulk of the had

be one in which Sections I and When all possible we achieved a 71

questionnaire). been exploited,

nonresponse follow-up

approaches DMI

percent response rate for the for the SBA sample. cases and 97 Of those of the

sample

and an 81 percent response rate

completing interviews, 92 percent of the DMI SBA cases completed all survey sections Breakoffs were

percent

including the income statement and more frequent than typically found

balance sheet sections. in

sample surveys and they occurred at To illustrate, 12 percent of OMI

all points in the survey questionnaire. eligibles and 8 percent of least some Section I data SBA on

eligibles stopped after having provided at characteristics of their business. Response

rates tended to be higher for the SBA sample than for the DMI sample.

In spite of having screened for
fielded DMI ineligible. sample These and survey four

eligibility, we had six percent of the of the SBA sample classified as

percent

ineligibles

included ineligible business types

(e.g., farms) with incorrect SIC codes, subsidiary companies and large

58

Table 9.

NSSBF Interview Response and Eligibility Results

Response Status

DMI Sample Percent Number

SBA Sample Number Percent 100.00 3.65 96.35

Total Interview Sample Ineligible Firms Eligible Finns

5,547 357 5,190

100.00 6.44 93.56

520 5;;

Ineligible Firms Out of Business Ineligible Business Type Ineligible New Business Ineligile Bad SIC Code Other Ineligible Firm Eligible Firms Section I Incomplete Refusal Breakoff Unable to Contact Language Barrier Other Nonresponse Section II Incomplete Breakoff Unable to Contact Language Barrier Other Nonresponse Sections III-V Incomplete Breakoff Unable to Contact Language Barrier Other Nonresponse All Sections Completed

357 213 120 5 13 6 5,190 1,153 903 53 166 17 14 358 292 65 0 1 282 268 13 0 1

100.00 59.66 33.61 1.40 3.64 1.68 100.00 22.22 17.40 1.02 3.20 0.33 0.27 6.90 5.63 1.25 o.oc 0.02 5.43 5.16 0.25 0.00 0.02 65.45

19 12 3 3 1 0 501 68 57 2 6 3 0 29 26 3 0 0 14 13 ; 0 390

100.00 63.16 15.79 15.79 5.26 0.00 100.00 13.57 11.38 0.40 1.20 0.60 0.00 5.79 5.19 0.60 0.00 0.00 2.79 2.59 0.20 0.00 0.00 77.84

3,397

businesses not identified during screening, and companies that had gone out of business since screening or were not truly in business. The latter

source of ineligtbility often was identified after the interview was nearly complete. In recording income, in the these interviewer would encounter no sales cases in the usually past determined that the but was now totally

and no expenses. business had inactive. Rather been

Probing actively

pursued

surprising

were 10).

the

similarities As an

in

response

rates across

reporting domains (Table

example,

the lowest DMI regional

response rate for the financial statement questions (Sections III-V) was 62 percent for the Northeast, as compared to 68 percent for the South. The

largest difference was found for metropolltan status . financial statement data 62 percent of of the time for rural firms. the lowest response for large sized businesses (68%). the

Urban firms provided

time as compared to 70 percent mpact on response with

Firm size had little businesses showed

(64%) and the highest for medium more variation with the lowest

Industry

resoonse for business

transportation, for businesses were found

conmiunication, and public utilities engaged in mining and manufacturing for the other two sections of the

(61%)and
(70%) .

the highest

Similar patterns

quest ionnaire and for the SBA sample. 5.3 Use of Worksheets Despite efforts to encourage only about 10 percent of returned DMI return of worksheets, over both samples or 32 tax records as requested. percent of SBA respondents (Table 11). We expected

worksheets and

About 16 percent

respondents worksheets,

reported having completed

however

worksheets to be more useful to proprietorships

than partnerships or

60

Table 10.

Response Rates by Reporting Domains

Domain

Level

Sample Eligibles SRA DHI

Business Characteristics Described SBA DMI 82% 84 94 85 --86 100 100 92 88 78 85 90 67 81 86

Financial Services Inventoried SEA GM1

Financial Statements Provided S&i DMI

Reglon:

Northeast North Central South West Urban Rural Small Medium Large Mining, Manufacturing Constructton Transportatlon, Comnuntcatlon, Public Utilities Wholesale Trade Retail Trade Insurance, Real Estate Services

1,328 1,292 1,367 1,203 2,613 2,577 4,258 533 399 623 668 191 504 1,446 281 I,477 5,190

79 148 111 163 --486 11 4 :: 18 52 172 3 154 501

75% 79 81 76 74 82 78 80 77 82 79 77 80 76 80 76 78

68% 72 74 69 67 75 71 74 71 ii: 68 75 69 73 69 71

78x 80 89 76 --80 91 100 84 84 78 81 85 67 74 81

62% 67 68 65 62 70 66 68 64 70 67 61 68 64 69 64 66

76% 79 86 72 --78 82 100 82 84 78 75 83 67 70 78

Metropolttan Status*: Firm Size:

Industry:

Total

*Metropolitan status was not defined for the SEA sample.

Table

11.

Use of Records When All Sections Comoleted

Domain Proprietorships Partnerships S Corporations Corporations Fewer than 5 employees 5 to 9 employees 10 to 49 employees 50 to 99 employees 100 or more employees Mining, Manufacturing Construction Transportation, communication, Public Utilities Wholesale Trade Retail Trade Insurance, Real Estate Services TOTAL

Worksheets Prepared DMI SBA 16% 15 18 22 14 21 f: 26 28 16 15 17 17 :: 19

Records Used For Financial Income Balance Services Statement Sheet

32% 32 28 27 33 24 29 24 80 24 31 50

25% 30 36 36 24 32 38 42 47 44 29 38

52% 27 47 51 47 44 54 41

37% 42 :f 38 48 51 z: 60 :: 47 44 46 42 46

65% 50 60 68 67 57 :“ 9 80 62 81 100 69 60 100 64 65

28% 39 53 54 31 45 52 60 67 59 39 48 45 41 44 38 43

57% 35 60 68 59 56 70 53 80 63 76 100 71 55 100 59 62

a0
51 62 79 50 44 100 49 50

fE __ 33 29

:5 36 28 32

corporations because statements.

proprietorships opposite

normally result:

do

not

prepare financial

We found the

corporations were more likely Among the

than other organization

forms to have

completed the worksheets.

other groups examined, larger firms and manufacturers smaller firms and firms in However, there were no other

were more likely than

industries to have completed worksheets. differences in worksheet use by Census

apparent

region or urban/rural location. Firms were more willing About 32 percent of DMI to use records than to complete worksheets.

respondents

reported consulting records to answer

questions about financial institution relatlonships, and 43-46 percent used records to answer the income and statement and balance sheet questions.

Corporations,

larger fins,

manufacturers

were more likely than other

groups of firms to consult records. In all groups, accounting records were the most connnonly used records Use of accounting

for questlons about ftnancial institution relationshfps. records increased with firm used by partnerships size, and and

nearly three-fifths of the records were accounting records.

corporations

Nevertheless, large proportions of records used tax records. In contrast, tax

in all groups were bank or

records were the most conunonly used records balance sheet, probably because the worksheet which were line of income tax forms each item

for Lhe income statement and and questionnaire could be found. fndlcated on

Proprietorships the Tax

more likely than other organizatfon statement. However, that is not

forms to use tax records for true for the balance sheet. because proprietorshlp tax records are

income

records are less useful for this purpose not some include a balance sheet, although
i terns (especially

tax forms do for

necessary

inventory

and

63

property) that appear on a

balance

sheet.

For both the income statement

and balance sheet, the use of tax records declined with firm size. A few additional preliminary worksheets and records are worth difficulty in reporting about amounts were the most difficul dollar amounts here but later and balance sheet items. is troublesome, but most Major items estimates. appear to Smal of be observations mentioning. about responses and use of

Most respondents had little Dollar

financial institution relationships. questions in this section.

Many estimated

consulted records to report income statement firms' failure to use worksheets or records these firms appear to have simple finances. dollar amounts are often

reported, prepayments,

although

Items such as

deposits, and accrued expenses and However, these items tend to

taxes payable are likely to be underreported. be relatively small amounts.

64

6.

DATA PROCESSING

With respect to data processing, edited data file of survey

RTI's

role was to produce a cleaned, The Board will perform all

observations.

additional logical editing and produce a user data activities. file.

statistical imputation activities needed to Thts chapter describes RTI's data processing

6.1 Computer Assisted Telephone Interviewing NSSBF (CATI). was conducted using using a computer-assisted telephone interviewing the CAT1 interviewer After

Rather than

printed questionnaire,

read questions as they were displayed the interviewer recorded the

on a computer viewing screen. answer, the next

respondent's prior As the system answers

question

consistent with that answer and the process was repeated.

appeared on the screen and was conducted and the

interview

respondent's data keyed, the CAT1 computer-readable file. control errors.

entered the data directly into a

CAT1 gave greater reducing transcription

over

the Because

interview skip

process

and aided in were computer

patterns

controlled rather than interviewer controlled, inconsistent data was reduced. Editing

the incidence of missing or were included in the codes and other

procedures

CAT1 program so that the data were checked for out-of-range invalid responses as the data were entered.

The CATI system required that

invalid responses be corrected before the interview could proceed.

In addition to the range
CATI, the NSSBF included

and

valid response checks normally done with checks deslgned to insure that known

cross-item

relationshtps were not violated.

Relationships that were checked Include:

65

.

The year the business was founded year as it was purchased/acquired. Across all percent. partners, the

should be before or

in the same

.

ownership

shares

should

add

to loo

.

The combined ownership share of stockholders owning 10 percent or more of the firm should be no less than ten times the number of such stockholders. The number of branch offlces in the same area as the headquarters should be no more than the total number of branch offices. The amount owed on a credit line line. should be no more than the total

. .

.

The characteristic least important to the firm in conducting its financial business should be different from the most important characteristic. The instttution from which the firm took business should be different from the institution to which they transferred the business. total equal Gross profit should sales (less allowances) minus cost of goods or services sold. Interest expenses should administratlve expenses. be no more than total returns and

.

. . .

selling and

The book value of the firm's three largest other assets should be no more than the book value of all other assets. The total amount of mortgages, notes and bonds due within one year should be no more than the total amount owed. The amount of the firm's three largest other liabilities should be no more than the total amount owed for other liabilities. are listed in the questlonnaire after

. .

The specific checks being performed the affected items. In addltion to these fairly

simple that

cross-item

checks, an elaborate

cross-item check was made to was within + 5% of the

insure

the calculated total for assets If not,

calculated

total liabilities plus equity.

the checking algorithm determined whether liabilities plus equity responses by

the fault lay with assets and/or the calculated amounts to

comparing

66

the totals reported by the business.9

Data in the faulty section(s) were The review

reviewed with the respondent and corrections made as indicated. process may or may not have corrected the discrepancy

since a second From

consistency check was not implemented past surveys, we have found that one

after

the data were reviewed.

review is helpful as respondents can If discrepancies remain after

spot obvious discrepancies and correct them.

this review, it is unlikely that the respondent can clear up the problem.10 Other special algorithms primary purposes. were and built into the CAT1 program for

two

First, space

burden limitations led us to restrict in Section II to no more than

the detailed suppliers

financial

inventory

four

for each service.
for

Second, some

the number of suppliers of financial businesses. obtained To limit the associated for six institutions and

services was quite large

burden, detailed characteristics

were

more limited location data for another six institutions. Section finn.

II collected an inventory of the financial services used by the

For each financial service, a yes-no question was asked to determine If at the answer was "yes", the CAT1 program institutions the (or sources) could

if the firm used the service. had the interviewer were record

which As

these
add

services

obtained.:

needed,

interviewer

91
El 111

The exact check that was performed is specified in the questionnaire. The Board will use both sets of data cleaned, imputation-revised data base. in their production of a

Financial services, particularly financing, may be obtained from Individuals as well as organizations. For convenience, we use "institutions" as a general term to cover all of these suppliers of financial services. 67

institutions to

the

roster. was

Having the

noted CAT1

all

institutions where the had the interviewer

particular service choose up to four the service.

obtained, for

program

institutions the all

collecting detailed information about had were four fewer than five institutions for discussed. When the business

Usually so

business providers than the

each service and

obtained a service from more interviewer record data for

institutions, the program had the largest or most important

three

institutions and then record data for "All Other Institutions." The "other loan" portion of the inventory posed even more burden

problems as it had a repeating for other loans. For each

section nested within each institution used

institution where the business had other loans, loans the business had and then asked A further limit of no institutions was also

the interviewer determined how many questions for up to four loans at more than imposed. As noted earlier, the ten loans discussed

each institution. across all four

questionnaire of financial institutfons order

only

allowed

for

detailed

discussions of six supplfers another six. To decide

services and location data for to discuss, the CAT1 program levels of this priority

which in

listed the instftutlons

roster

within

ordering established by the Board: (1) (2) the primary institution, institutions where multiple service categories were used (arrayed by the number of categories used where the categories were checking, savings, financing, and other services), institutions with checking only, the first institution (in roster order) with savings only, institutions with lines of credft only,

(3) (4) (5)

68

(6)
(7) (8)

institutions with financing only types of financing used),

(arrayed

by number of different

the remaining institutions with savings only, and institutions with other services only.

Having ordered the institutions, detailed data were collected for the first six institutions and location ordered list. data for the the first next six institutions in the business used more than 12

In

a

few

instances, only the

institutions: in which case, were discussed. 6.2 Problem Resolution Reoorts A routine component of Report (Exhibit 2). passing information

12 in the priority ordering

RTI's

CAT1 surveys

is the Problem Resolution

These reports are a way of documenting problems and/or along to the

CAT1 programmer.
report

For

instance, an

interviewer would complete such a which CAT1 would not allow them

if the respondent gave an answer Typically, the source of the of

to

enter.

problem is a keying error made

by the interviewer or a misunderstanding

meaning of the associated question(s).

In the early days of interviewing,

the source of the problem is sometimes a flaw in the CAT1 program such as a contingency that the program Resolution Report, the needed and then CAT1 cannot handle. After receiving the Problem the CAT1 and data base if

prograimner corrects caused the

comments

what

problem

how it might be (1) it

avoided in the future.

Thus, the

report serves multiple purposes:

allows the correction of data that in error, (2) it serves as the problems they are having

may be unintentionally missing or keyed

a training tool for interviewers by identifying and providlng solutions, and (3) it provides

an early warning if changes are needed in the CAT1 program.

69

Exhibit 2

Cdti No.:
I

Question No.: Supervisor:

NSSBF

I 1 Problem Resolution Report

Date: Resolved by:

11. Brief Description of Problem:

. Resolution By:

Date:

. Notes:

An extraordinary number of Problem Resolution Reports were generated in NSSBF interviewing. The nature of the led questionnaire together with the

built-in checks and special algorithms some interviewers had difficulty during the first few experience Frequently, in weeks of

to a complex CAT1 program that These problems were evident As the interv ewers gained two or new prob ems emerged. additional In

mastering. interviewing.

administering

the

interview, their minds

respondents

changed

remembered

services/institutions

long after the relevant asked the had

section had been passed.

these cases, the interviewer copy questionnaire, Problem Resolution

necessary happened

questions using a hard and any new data on the was for the

recording what Report.

Another

coernon occurrence

respondent and/or the interviewing issues.

circumstances to raise rather technical

For instance, an automobile dealer might have a "floor loan" which they sell. This situation might prompt Should I

they use to purchase the vehicles

the interviewer to ask, "How do you want

us to handle this loan?

classify it as a line of credit, a motor vehicle loan or as an other loan?" As needed, we responded to the Problem Resolution Reports changing the data in the CAT1 data base if indicated and recording actions taken on the

report and any instructions to the interviewing staff. The cross-item checks and for the interviewer. give answers for gmss For range check it were was not also resulted in some problems cormnon for the respondent to equal to total sales (less When this

instance, that

profit

returns and allowances) minus costs of

goods and services sold.

discrepancy could not be resolved, the interviewer would usually enter "OK' for "don't know" and proceed with the interview. After the interview was

complete, the interviewer submitted a

Problem Resolution Report citing the

71

respondent‘s answers.

The CAT1

programmer manually entered the discrepant

data for resolutlon later by the Board. Other checks that frequently balance plus resulted sheet in items. zero Problem Resolution Reports For items such as total was not an allowable

were the range checks for assets and total

liabilities These

equity, did

response in CATI.

responses In the

occur, however, usually for very

small service businesses. in order to proceed with

these situations , the interviewer entered $1 interview. We examined all such cases to

determine whether the busines s was truly in bus iness and hence eligible for the study. At the conclusion of Sect ion I, interviewer was given an opportunity into an open-ended comment file. II and V and at the menu screen, the

to write comments about the interview These conmtents were intended to be That's

technical in nature (e.g., "They are a real estate holding company. the reason for all the zeros in the

income statement.") but often were (e.g., "Never in until late the comment fields to record

observations about the interviewing afternoon"). Some interviewers CAT1 data when

situatlon used

also base. the

changes needed in the interview was

Typically, these were after the remembered an answer or

concluded

respondent

realized that they had answered a question the survey, all comments were base as appropriate. A examined and

in error.

At the conclusion of

changes made in the CAT1 data

Problem

Resolution Report was again generated for

the case to reflect any chang e s made. 6.3 Completeness Checks As noted in Chapter 3, al labeled as "complete" were cases ndeed were checked to verify that sectlons complete and that sectlons that were

72

labeled as "incomplete" were and the source(s) of the

incomplete.

Discrepant cases were examined Usua ly, the problem was In this case, the incomplete

discrepancy

resolved.

that the record had too few data the response status variable

to be considered camp ete. was for reset to indicate

sect on(s) and the case returned the nterviewer labeled an

reinterviewing. as incomplete

In a few instances, when only one or two

interview

questions remained to be asked. the remaining questions to "don't

With all data essentially present, we set know" responses and changed the response

status code to "complete" for the sect on. The completeness checks resulted being examined. This examination n several hundred discrepant records occasionally identified interviewer check. An example

errors other than those associated wit h the completeness

of this type of error would be a "yes" response to the other assets lead-in question and then answers that clearly indicated that the firm had no other assets.

In

some

cases

the

error the

made

a

completed have

interview appear missing data for a

incomplete.

In other

cases,

record

might

particular institution and/or type of

financial service but an interviewer

comment elsewhere might indicate the service was never obtained or that the institution was not used on were resolved on a case by the case relevant basis date. with These types of problems a Problem Resolution Sheet

completed to record the actions taken. 6.4 Edits to Resolve Interviewer Errors Our examination of these discrepant errors being made by the interviewers. of records also identifled some basic When such errors were identified, the problem. As an example, we the section on most

we searched all records for occurrence discovered that many interviewers were

misinterpreting

73

recent loans.

Added after

the

pretest

and

hence never tested, the most

recent loan sectlon was a series of questions to be answered about the most recent loan the person had or other organization. This obtained from a financial institution, business sectlon was not did to have been completed for not make sense then). If

loans received from persons (the all of the businesses' loans

questions from

were

persons,

the interviewer was to

record this fact so that CAT1 could skip the rest of the section. We had two types of Loan section. about a loan First, from a interviewer for some errors occurring for the Most Recent

businesses, the interviewe 'r recorded data When all the business'
1oans were from

person.

persons, the error was not loans had come from persons asked in error. When one

particularly and or blanked more of

serious: we simply noted that a.11 out the the questions that had been business' loans came from

institutions, however, the interviewer error the entire section. These cases had

resulted

in loss of data for

the lead-in question set to a special and for the remaining data items set to sectlon was to have the

"Interviewer Error" consistency code missing. A second source of error

this

interviewer record that all loans financing from an institution. missing data for the entire

came Again

from persons when the business had thls Here left error resulted in completely again, we set the lead-in

section. and

question to "Interviewer

Error"

the remaining questions blank.

Over both types of errors, we had 117 cases coded as Interviewer errors. Two other activities. incorrectly interviewer First,
1 abeled

errors discovered "another

were that

detected some

as

a

result firms

of these had been

we as

business Since these

source." of

miscellaneous

"another sourc eU types

of

suppliers

financial

services did not have

74

detailed characteristics

collected

for of all

them,

this

mistake

resulted in To

missing data on the characteristics identify these errors, we checked

the associated business firms. suppliers

classified as "another We corrected erroneous

source" to verify if they had been

coded in error.

type of supplier data as encountered and recontacted the business to obtain the missing data if possible. Second, in the yes-no questions balance were asked In sheet to a section, determine number we if had amounts where several the firm had assets or

liabilities of that type.

of cases, the respondents had an

amount but did not know the specific it was associated. In this

type of asset or liability with which

OK” situation, the interviewer usually keyed ‘ and then keyed "Yes" so that CATI As a result when of this problem, the there are several

for all yes-no questtons except the last would let them record the amount. may be

individual yes-no

questlons

unreliable To of

associated with a particular yes-no recodes amounts. for the

amount. set

solve thls problem, we created questions associated with such

entire

6.5 Matchinq to Institutional Files RTI also matched cormnercial banks, savings banks to data bases types of Institutions. number and county FIPS NSSBF-reported For The code maintalned result of the of savings and loan associations, and by the Federal Government on these this match was the identification record matched to the

Institutional

Institution. purposes, the Board for provided the most current in 1986.

matching

institutional data base, which was With NSSBF data collected for

fnstltutlons In operation thfs

1987-88,

ttrl\c frame difference can be

75

expected to produce mismatches have changed.

for

new

businesses

and those whose names

For instance, it is common for savings and loan associations

to convert to savings banks. Our first step in NSSBF-reported matching was to verify geographic fields to location of the

institutions, editing Very few

these

correct discrepant

state codes and ZIP Codes.

changes were needed for state codes that made sure the ZIP Code was versa). of For some cases, the

since CATI had included a cross-item check appropriate for the reported state (and respondent did not know interviewer put purposes, we set in the last marks few or

vice digits

their ZIP Code and the For processing

question

dashes to zero

instead.

these

missing

digits

after verifying that no

legitimate ZIP Codes used zeros in these positions. In the process of examining these locatlon were data, we identified

institutions in foreign countries, which

not eligible for the match.

In addition, we identified cases

with

city and state missing, which could

not be matched due to insufficient information. Our next step was to examine the institution names as supplied by NSSBF businesses. For matching purposes, the names were modified to correct

obvious misspellings

and other differences In ;ome cases Such

from the government's data base

r?nditlon of the name. the nw of the

the respondent had refused to give cases could not be matched by

institution.

definition. Only conanerclal banks, savings and were to be matched. Type of loan associations and savings banks was recorded by the NSSEF

Institution

interviewer as reported by the respondent. errors could result in misspeciffed

As respondent errors and keying types, we searched for

Institution

76

occurrences of the terms

"bank,"

'savings," cases bank,

'thrift'

and "trust" in the

institution names, playing out all type was not given savings bank. as commercial the type

with these words appearing whose savings and loan association or

Errors in

of institution variable were corrected In some cases,the entire set of institution, and hence it was not

prior to performing the computer

match. of

data was missing, including the type

known whether or not the institution was eligible for the match. On a state by state basis, we matched the two data bases, using several consecutive matches of the data base. of To be "eligible" for attempting

matching, the institution had to be and location data given. full institution name and eligible institutions were The of first the

the appropriate type and have name pass required an exact match of the Twenty-four percent of all pass required an exact

Zip

Code. The

matched. and

second

match of the full institution name eight percent of the eligible

the city and matched an additlonal instltutlons. The next two

NSSBF-reported

passes were similar except we matched

using the first 13 characters of the In these two passes, 17 and 6 Our final step was

institution name instead of the full name.

percent of the institutions were matched, respectively. a hand match of the remaining were matched by hand with but not location. Over all records.

Thirty-one percent of the records

another steps

four percent having Institution matched of the matching process, we matched 90

percent of the eligible records. 6.6 Final Edltlnq and File Delivery Editing software was developed to perform project-specific checks on

the CAT1 data base prior to delivery. we verified the integrlty of all skip

Using the CATI program as a guide, patterns In the CATI data base.

77

Dfscrepant cases were played out the discrepancy and to resolve the pattern errors tended to result the

and

examined

to determine the source of

problem. from

Few in occurrence, these skip backing up to change

interviewers

responses keyed in error and

CAT1 software not erasing backed-up over

data items when new responses caused these data items to be skipped. In addition to these skfp pattern checks, we examined the frequency and range of the study responses were variables out to for validity of the the responses. best way to Erroneous resolve the

played

determine

discrepancy and then appropriate changes were made. The final data files incorporate the questionnaire the sampling design variables, analysis weights and responses as well as the results of the codebook

match to the financial institutfon data bases. was also delivered that provided as a

A machine-readable for

description appropriate.

each variable and a

range or frequency dfstributlon

The RTI data base will

form the starting point for editfng

and Imputation activities by the Board

to produce a final NSSBF user data base.

78

7.

CONCLUDING REMARKS

The National Survey of Sma 11 Business large gap in knowledge of Our initial results the

Finances was designed to close a

financial behavior of small business firms. that the survey goes a long way towards

suggests

accomplishing this objective. It appears that reasonably good coverage of firms with employees was tax-based sample

obtained using the DMI sampling frame.

More comprehensive

frames may have provided theoretically better coverage of the population of small businesses, but restrictions on their use would have made their

actual coverage much worse than that obtained using the DMI file. Firms requested were induced and to cooperate in thls voluntary survey which

difficult

confidential

flnanclal

informat ion.

However,,

substantially greater resources were

required

to obtain an interview that

is normally required for a consumer survey. interviewers were able to interviewers, a interview. achieve amount adequate of

Only experienced and "natural" yield was rates. required Even for these to obtain an

substantial

time

The time required to complete an interview was dominated by the

time spent to establlsh contact, to gain cooperation, and to resume contact following breakoffs. more than three On average, the total time spent for an interview was
times

the

average

admlnistration

time

for

the

questlonnalre. Despite these dlfflcu 1t fes, obtained for the NSSBF, which is financial surveys. For an acceptable overall response rate was

comparable to those obtained for consumer interviews, item nonresponse does not

completed

appear to be a problem.

Cornnon knowledge questlons had very low rates of

79

missing data.

Even

for

dollar

amounts,

missing

data

rates were

10~.

However, the use of worksheets or records was lower than hoped and reported dollar amounts were frequently estimated. Future work by the Board will

focus on data consistency, the extent of use of estimates, and the accuracy of reported data.

80

REFERENCES

Ando, Faith H. Access to Capital by Subcategories of Small Businesses." Report prepared for the Small Business Administration under Contract No. SBA-6061-OA-82. Fort Washington, PA: JACA Corporation, 1985. Avery, Robert 8. and Gregory E. Elliehausen. finances: Technical Manual and Codebook." Governors of the Federal Reserve System, Statistics, 1988. "1983 Survey of Consumer Washington, DC: Board of Division of Research and

Board of Governors of the Federal Reserve System. Financing Small Business. Report by the Federal Reserve System to the Cormnittees on Banking and Currency and the Select Committees on Small Business. Washington, DC: US Government Printing Office, 1958. Board of Governors of the Federal Reserve System. u. Washington, DC: Board of Governors System, 1980. Bridge, Lawrence. "Capital Requirements Current Business 28 (December 1948): of Introduction to Flow of of the federal Reserve Trade Finns." Survey of

18-24.

New

Bridge, Lawrence and Lois E. Holmes. 'Capital Requirements of New Manufacturing Firms." Survey of Current Business 30 (April 1950): ll8. Combs, Robert, G. Pulver, and Ron Shaffer. "Financing New Small Business in Wisconsin.' University of Wisconsin, Department of Agricultural Economics, 1979. Converse, Muriel and Steven G. Hecringa. "An Evaluation of the Accuracy and Current Utility of the 1981 Master Establishment List (MEL).' Report prepared for the Small Business Administration under Contract No. SBA-6067-OA-82. Ann Arbor, MI: Instltute for Social Research, 1984. Cox, Brenda G. and Steven B. Cohen. Methodoloaical Issues for Health Care SUWCYS. New York, NY: Marcel Dekker Inc., 1985. Dennis, William J. Jr. Credit, Banks, and Small Businesses: 1980-1984. Washington, DC: national Federation of Small Businesses, 1985. Dennis, William J., Jr., and William C. Dunkelberg. Credlt, Banks and Small Buslness: The Unlted States. Washington, DC: Nationar Federation of Small Businesses, 1988. Federal Comnitee on Statistical Methodology. Surveys, Working Paper 15, Washlngton, DC: Budget, 1988. Federal Reserve Bank of Atlanta. Quality in Establishment Offi ce of Management and

Economic Review 67 (April 1982). 81

Iannacchione, Vincent G., Lisa M. LaVange, and Allen P. Duffer. "The Pesticide Use Survey of Food Processors." RTI Final Report No. RTIResearch Triangle Park, NC: Research Triangle Institute, 3493-02-OlF.

1986.
Internal Revenue Service. Statistics of Income, 1986: Corooration Income Tax Returns. Washington, DC: US Government Printing Office, 1989a. Statistics Washington, DC: of Income, 1986: Partnershio Income Tax Returns. US Government Printing Office, 1989b.

"Financing Small Business in the Postware Period." McHugh, Loughlin F. Survey of Current Business 31 (November 1951): 17-24. McHugh, Loughlin F. and Jack N. CiaCCiO. "External Financing of Small- and Medium-Size Business." Survey of Current Business 35 (October 1955): 15-22. Miller, Merton ii. "The Modigliani-Miller Propositions after Thirty Years." Journal of Economic Persoectlves 2 (Fall 1988): 99-120. Hlller. Modigliani, France, and Merton H. "The Cost of Capital, Corporation Finance, and the Theory of Investment." American Economic Review 48 (June 1958): 267-97. 'Small Business Financing: A Review of Data Collection Ou, Charles. Paper presented at the Eastern Finance Assoctation Meettng, Efforts.“ Nashville, TN, April 1986. Pettit, R. Rlchardson and Ronald F. Singer. "Small Business Finance: Financial Management 14 (Autumn 1985): 47-60. Research Agenda." Robert Morris Associates. Annual Statement Studies, 1987. PA: Robert Morris Associates, 1988. A

Philadelphia,

Stockwell. Elanor J. and James C. Evmes. 'Small Business Financino: Federal -Reserve Bulletin 47 (January 1961j: Corporate Manufacturers." 8-22.

us Bureau

of the Census. Quarterly Financial Report for Manufacturfne Fourth Quarter, 1987, Series (QFR-87Mining, and Trade Corporations. 4. Washfngton, DC: US Government Prlnting Office, 1987. Financial Characteristics of US Farms, of Agriculture. January 1. 1988. Washington, DC: US Government Printing Office, 1988.

us Department us Small

Business Administration. The State of Small Business: A Report of the President. Washington, DC: US Government Printing Office, 1984. A Report of State of Small Business: DC US Government Printing Office, 1988. the President. Economic

The Washington, Watro, Paul. Commentary,

'Financial Services and Small Businesses.' Federal Reserve Bank of Cleveland, January 11,

1982.

82

Weston, J. Fred. "What MM 1989): 29-39.

Have Wrought."

Financial Management

18 (Sumner

Whitehead, David 0. "The Sixth District Survey of Small Business Credit." Economic Review, Federal Reserve Bank of Atlanta 47 (April 1982): 428. Geographic Market Dellneatlon: Wolken, John D. Literature. Staff Studies 140. Washington, DC: the Federal Reserve System, 1984. A Review of the Board of Governors of

a3

APPENDIX A NSSBF SCREENING FORM

UllU THE NATIONAL SURVEY I-OF SHALL FORM BUSINESS FINANCES BATCH : I.AS’ I’ SCREENING

110.

/

IUU

U:

J:,

I~l:SIJLT:

L. CONTACT INFORHATION

ORIGINAL INFORMATION

CORRECTED INFORMATIOti --.----. 1. 2. 3.
4.

/. CASE ID NO. ...... !. FIRM NAME ........ I. ADDRESS .......... ,. CITY, STATE, ZIP.. 1. EXECUTIVE, TITLE . 8. TELEPHONE NO. ....

__-

5. 6.(_____~_ )
AC

TELEPHONE NUMBER
TYPE: NAME: I-Prol,. Z-Part.

EXT.

ORGANlZATlON RESPONDENT I. WHEN

3-Corp. 4-Di
_.-_

SAMPLE CUtlPANYIS CONTACT&D, SAT: Hello, my name is . I'm cnlllng from the Research Triangle Institute Carolina on behalf of the Federal Reserve Board in Washington, DC. Hay 1 Speak to ((EXECIJ'I'lVEj/the owner or top executive of your firm)?
in North
l 1. YES, THIS IS HE/SHE + 2. YES, EXECUTIVE COHES TO PHONE NO, UNAVAILABLE 3. 4. (EXECUTIVE) NO LONGER WITH FIRII ,*

CONTINUE SCREENING REPEAT INTRODUCTION WHEN EXECUTIVE COMES TO PHONE. SCREEN PERSON WHO ANSWERED THE PHONE ASK FOR CURRENT OWNER OR TOP EXECUTIVE AND 601.1.~~

CONTINIJE DIRECTIONS

SCREENING 1N I, 2,
OR :

We

would like to meil your company information about e research study the Federal Reserve Boald IS spolbsurJ,,g. ‘ I II, callIn& to veclfy the informetion we hevc on our recocde.
1.

Is your company*s mu address (FIRM NAME, ADDRESS, CITY, STATE, ZIP)? YES ... 1 NO ... 2 + PROMPT FOR ADDRESS CHANGES AND RECORD CORRECTIONS ON LINES 2-4
a.

IN

“CORRECTED

INFORNATION” ON LINE

SECTIt

We would like to eddress this informetion to (you/the owner or top executive). VERlFY NAME SAtit AS LINE 5 .......... 1 DIFFERENT FRDtiLINE 5 ... 2 + PROHPT FOR CORRECT NAME AND RECORD ON LINE 5 IN “CORRECTED
Whet

5 ABOVE. SECTIC

INFORMATION”

b.

is (your/[EXECUTIVE]'s])title or position in the company? SAM AS LINE 5 .......... 1 DIFFERENT FROM LINE 5 ... 2 + RECORD CORRECT TITLE ON LINE 5 IN THE "CORRECTED INFORMATION" SECTION

c.

Can (you/(EXECUTIVE])best be reeched at (TELEPHONE NO.)? YES ... 1 NO ... 2 l PROMPT FOR TELEPHONE NUMBER AND RECORD ON LINE 6 IN THE "CORRECTED INFORMATION" SECTION
--

2.

Is

this

the YES

headquarters .. .

or you

main for

office your

oC time.

(FIRM NAME)?
This is a survey

1 for main offices only.

NO 3. As of

. . . 2 + Thnnk

did (FIRH NAME) heve December 31, 1987, YES (500 oc fewer) . . . 1 + (CO TO 4) NO (more thnn 500) ... 2 DON’ KNOW . . . . . . . . . . . 8 + (GO TO 4) T owncra who worked full-time for

500

or

fewer

full-time

employees?

S.

Including
NUHBER

the

firm,

what

was

the

average _ number -----

of

full -----

time

employees

in

198)

b.

Including NUMBER

owners

who worked

part-time

for

the

firm,

what

was

the

average

number

of’ part

time

employees

in

1987

CALCULATE 1987
NUMBER:

FULL-TIME EQUIVALENTS AS (Q.3a + (0.5 x Q.3b)) AND RECORD BELOW. l IF 500 OR MORE, SAY: Thank you for your time. Thla is a survey

of

small

businesses

only.

Goodbye.

4.

Is

the

firm a not-for-profit YES . ..T* Thank NO . . . 2

you

organization? for your time.

This

is

a survey

of

for-profit

businesses.

5.

Is

the

or Federel government agency? firm owned by A local, State, , YES . . . 1 • Thank you for your time. This ia a survey of privetely
NO . . . 2

owned

businesses.

6.

Does

another YES NO

...

company own mote than 50% of (FIRM NAME)? 1 + Thank you for your time. This is a survey 2

of

independently

operated

businesses.

.. .

firm

coneidered

to

be

.

.. .

Proprletorahip Pertncrehlp
Corpocstion DON’ T KNOW

. . . 1 . . .. .. 2
.. . .. . .. . . .. 3 fl

CIRCLE RESPONSE HERE AND IN “ CORRECTED INFORMATION” SECTION ON PACE ,.

ASK IF RESPONDENT IS

(EXECUTIVE) you About l ny of this information, would you give o\e your name?

In case

I need

to

recontect

ENTER RESPONDENT NAME IN “ CORRECTED INFORHATION” SECTION ON FRONT PAGE.

Think you very much for from the Pederel Peeervo you within l week after

(You/Plosee your help. Board within the next
rcceivlng the psckege.

that inform (EXECUTIVE] few weeks. Pleesc retein Goodbye.

(he/she)) the package.

stmuld Our

be

~ecelvtng

an

important

packrrgti

irlterviewers

will

be

calling

C.

RECORD OF CALLS

k:t‘l

Date 1 Time ['WIT0

RESULT

HIIMBEK IPIIOtIE

CAI.l.EI

-1-_-I

I

I

2 3

71
8 9

I

I

I

I

CONTACT RESULTS :odt Rttult -01
02 glmy

TEHPORARY CODES ' Code Result -08
09

Code Result -20 21 24 25

FINAL __-__- CODES Code -I__ Result 26 27 28 29 30 Language barrier Listed Number, WC No phone listing, UTC Otlher (Specify above) Partial Screening

Ring no tntwer No result from ditl Wrong number Double wrong connection
Non-working number Computer modem

03 04 05 06
07

10 11
12

13

Answering scrvice/mtchint Call btck tppointmtnt Directory Attietenct cell Chamber of Commerce ctll Better Buoinett Buretu ccl 1 Other Trtcing cell

Complete ( eligible bualntss) Complete ( inellgible business) Confirmed, no longer in business Refuse1

:OWlENTS:
__ _~__... .-_

_

______

APPENDIX B NSSBF LEAD LETTER MATERIALS

SOAR0

OF

CavEuNaRs
SYSTEM
‘ ,. 20551 0.

OF TNE

FEDERAL

RESERVE

NASHINGTON.

Oecemoer

19, 1988

The Federal Reserve Board is concerned with how recent changes in financial markets are affecting the availability of credit and other financial services to small businesses. These changes, a result of both innovation by financial institutions and a relaxation of altered the reguiatory restrictions by government, have dramatically zanqe of financial options avarlable to small businesses.

A surpey of business people Like yourself is the only way to obtain the information that we need for understanding the financial needs of small businesses. For this reason, the Board, together with the Small Business Administration, is sponsoring the National Survey of Small Business Finances, which is described in the accompanying materials from Research Triangle Institute of North Carolina. The information collected in the surrrey will help to evaluate public policy in areas such as the cost and availability of financing to small businesses, bank deregulation, and bank holding company acquisitions. I am writing to urge you to participate in this surrrey. Your and others were statistically selected to represent a cross section of small businesses in the United States. Participation in the survey is, of course, voluntary. However, your cooperation is vital for the survey to reflect the views and needs of small businesses. Any information you provide will be held in strictest confidence.
firm

k stated in the attached letter, a representative of the Research Triangle Institute will be contacting you within the nest few days about the sumey. I would like to thank you in advance for your cooperation.

Sincerely,

Attachment

RESEARCH

TRIANGLE

INSTITUTE

October

10, 1988

Dear Sir or Madam: is Pleased to be conducting the National (NSSBF) for the Federal Reserve Board and the Small Business Administration. The purpose of the survey is to help these and evaluate public policies agencies understand affecting small business finances. An RTI telephone interviewer Will call you in the next few days to conduct the interview. As federal Reserve Board Chainnan Alan Greenspan stated in his letter, the of eacn small business is critical to the success of the survey. As the operator of a business you understand the need for complete and accurate information when making a business decision. The Board will use the information collected in this survey to assist decision making and to evaluate policies that affect small businesses like yours. Without your participation, the Board's decisions will have to be made in the absence of complete information about firms like yours.
participation Research Triangle Institute (RTI) Survey of Small Business Finances

Enclosed are worksheets to assist you in locating and organizing business records needed to answer questions asked in the interview. To increase the accuracy of the survey data, it is important that you complete the worksheets prior to the interview. In addition, completing these worksheets will greatly reduce the interview time. commonly asked, we have enclosed a brochure entitled "Questions and Answers About the NSSBF." If you have additional questions about the survey, please call MS. Susan Henderson, toil-free, Monday through Friday, from 9:OO am to 5:OO pm EDT at (800) 334-8571. If you are in North Carolina, call collect at (919) 541-6000. At the end of the study, participating businesses will receive a summary report prepared by the Federal Reserve Board. As described in the attached brochum., your responses will be kept canfidential and your participation is voluntary. The survey is of vital importance to all small businesses. I join Chairman Greenspan in expressing advance appreciation for your cooperation. Sincerely, t.
To answer questions

44 GRH:pb Enclosures FbstOfficeBox12194
.

7h

Researcn Triangle Park, North Carolma

Telephone 919 541-6436

National Survey of Small Business Finances

Questions and Answers

:

u
1
There

Why

is this survey being done?

u
2
from small obtain needs

Why did
a list of Slales. lo represent you

has been lillle study of how recent changes in financial markets and the

YOUI business
the Clniled designed information

regulation of commercial banks and other financial institutions have affected the cost and availabilily of financial services used by small businesses In particular, the government’ s evaluation of policies designed to ensure adequate credit and other services for small businesses depends on answers lo questions such as these: I. What types of financial services do

businesses an accurate of small

cl
3

Will my

small busiqesses use and where do they gel these services? 2. What sources do small businesses use IO meet their financing needs? Do small businesses purchase their financial services from a single source or do they use mulliple sources?
,

Yes, most assuredly

be kept confide&al. who records confidenMity him/her frorn lion IO anyone projed slaff of Lule and the Other than these information that will be released government. cooperating businesses only as aggregate
lhal no single

3.

4. How far away from Ihe firm are the financial institutions that small businesses use? 5. Whal types of financial institutions to get

solicil firms in an attempt their business?

6. What types of inslilulions do small businesses consider when looking for
financial services?

a
4

How long

The National Survey of Small Business Finances is being conducted to answer
these

quest ions.

The interview minutes. Characteristics may cause the less time. However, reduced by completing worksheets prior

0
8

Am I requlred by law to participate?

No. Your participation is voluntary and you may refuse to answer any question IO which you object. However, the participation of each selected business is vital to the success of the survey.

cl
9

What will I get out of this?

The information you provide will help the Federal Reserve Board and the Small Business Administration IO understand the needs of small businesses like yours and to evaluate public policy affecting the cost and availability of financial services to small businesses. All participating businesses will receive a report prepared by the Federal Reserve Board summarizing the basic findings of the survey.

For frrrlher inlormalion,

please conlacl:

Ms. Susan Henderson Research Triangle Institute Post Office Box 12194 Research Triangle Park, NC 27709-2194 (800) 334-857 I In North Carolina, call collect (919) 541-6000

OMB NO 7100-0234 EXO w3i/aa

National Survey of Small Business Finances

Worksheets

the Small

Business Administration

INSTRUCTIONS

Comoletlng these worksheets will reauce the time needed for the interview and prowoe us with the accuraIe aata we neeo. These worksheets ~111 use0 during the IntervIew 10 answer questIons. Recoras Ihat ~111 be be Jseful In compretIng the worksheets are:
1 bank StaIemenrs

2 Income tax returns 3. balance sneers 4 income statements.
Choosing the rlgnt time Denod for answenng cnoose the appropnare time perloa:

these questions IS cruaal. The following rules ~111 allow you 10

ftrms rnat flied 1987 taxes under one Of the currenr owners should use IheIr 1987 tax year in answering :he auestions unless Ilnanclal records are avaIlable for a more recent fiscal year. Use the beginning aaIe ana enomg aate of the 1987 lax year or 1988 fiscal year as the BEGINNING DATE and END DATE for reoonlng firms Inat did not file 1987 taxes under one of the current owners but who began operanons pnor to 1988 snouId use June 30. 1988 as the END OATE for reporting and the date they began operations as the BEGINNING DATE.

Firms that began operations in 1988 should use the end of the prevtous month as the END DATE far reponing and the date they began operattons as the BEGINNING DATE. Enter the BEGINNING DATE and END DATE for the ttme penocl here: BEGINNING DATE: END DATE: Month: Month: Day: Day: Year: Year:

Please be sure that all answers refleCt this time perrod and please provtde ewd amounts in dollars whenever possible. Note that some Items request that you sum over all aCCOUntS the same nstitution. Also note that some at Items may not be applicable for the firm. If not apsicable. lust enter “ or ChaCk the “ 0” NONE” or “ NOT USED” line and go to the next item. Firms that use more than four instItutIons for a particular financtal service should report for the three where they have the m dollar amounts and then aggregate across all other Institutions. Firms that have more than one OffiCe.Plant or store and firms that own subsidiary companies should provide data for the entIre enterpnse. At the ena of the Interview. you will be asked to mail the completed worksheets to RTl in the enclosed postage-paid envelope. We will use the documents for quality control purposes. All information will be kept confidential. Thank you for your cooperation.
L

!

.-

A.

Use of Financial

Institutions

and Other Sources

of Financing*

(As of END DATE) Institution Item No. and Description 1. Checkmg Accounts:
If none. check here _

or Source of Financma 1 Third Fourthor All Others

First

/

Second

ana go to 2.

Name of lnstwtlon Balance of checkmg account(s) 2. Savings and Investments:
If none. check here _ Name of Instrtution Balance of sawngs. money manet deposlt and share accounts, money marnet mutua funds. cemficates of deposit (CDs) and other time aeposlts (except trusts. and go 10 3.

retw’ ementaccounts. and oenslon funds) 3. Capital Leases:
l

* and go to 4.

If none, check here _

Name of instltutlon or source Number of capital leases Total amount of remalnmg obligations
4. Lines of Credit/Revolving

s

S

.

Credit and go to 5.

Agreements: If none. check here _ Name of institution Number of credit lines/ revolving agrwments Total amount of credit line(s)/ revolving credit Total amount owed on credit line(s)/ revolving credit S S 0

s

l Finrncirl Institutions wlude commerctslbank sswngs and loan associations savings banks. credit unions. finance companies. lnsuranca companies. brokerage and mutual fund companies. lessmgcompanies. and mortgagebanks. Other sources of finrncinq Include owners.family and other mdividuals; venture cqxtal and other busmess hrms: the Small Business Admwwatlon (SEAI and other government agencies; and any other source. “CaPits 1. the 2. the 3. the 4 Ihe Leases satisfyone of the folfowmg conditions: lease transfers ownenhip of the asset at the end of the lease term. lease contams an opwn for a bsrgsln purchase of the asset. lease term extends over most of the economic life of the assst, or lease requwes rental payments that appruxlmate the fair marlce(value of the leased ~zx&. 1

.

A. Use of Financial Institutions and Other Sources of Financing (contcnueal
(As of END DATE) Institution or Source of Financing
I

tern No. and Oescrcption 5. Mortgages on Business If none. cneck here _
Real Estate:

First

Second

I

i

Third

Fourthor All Othen

and go to 6

Name 01 lnstltutton or source
I Numoer oi mortgages

I
Amount ot prlnclpal mongaged owea on

s

/s

:s

1

6. Automotwe and Other Vehicle Loans:
if none. cneck here _
dfld 90 10 7

Name or lnstltutlon or source Number of motor vehicle loans Amount of pnnclpal owed on motor vehcle loan(s) 7. Loans Secured by Equipment and installment Purchases of Equipment: If none. check here _ and go to 8.

Name of lnstltutlon or source Number of equtpment loans Amount of pnncipal owed on equipment loan(s) 8. Loans from Partners/Stockholders: If none. check here _ and go to 9. ALL LOANS FROM PARTNERS/ SrOCKHOLDERS ARE TO BE COMBINED AND ENTERED IN THE FIRST COLUMN.
l

rb

Number of Bane from partners/ stockholders Amount of pnncipal owed on loans from partners/stockholders

. Item 8 IS 10be completed by partnenhlps. S corporations. and corporations only. Sole propnetorshlps should check “ NONE” and record any personal borrowtng for busmess purposes m Item 9.

r

A. Use of Financial Institutions and Other Sources of Financing (contlnueo)
(Aa of END DATE)

institution or Source of Financmg
Item No. and Oescrtption 9. Other Loans: m
If none. cneck here _

First

I 1

Second

I

/

Third

All

Fourth or Others

and go to 10.

Name of nsrltutron or source Number ot other loans Largest Loan horn Institution/Source: Onglnal term of largest loan
/ !

I

M-s.:
Mos:

/ Yn: 1 Mos:

Yn: Mos: %

iyrs Mos:
I $

Amount owed on largest loan Second Largest Loan from Instftut~on/Source: Ongfnal term of second largest loan

S

$

Yn: Moe:

Yn: Moe: 0

Yn: Mos: S

Yn: Mos: S

Amount wed on second largest loan Third Largest Loan from Institutran/Source: Origfnal term of third largest loan

%

Yn: Mos:

Yn: Mos: S

Yn: Mos: S

Yn: Mos: S
.

Amount owed an third larger3 loan Fourth Largest Loan from rnstrtutlon/Source: Original term d fourth largest loan

S

Yfs: MIX:

Yrs: MCX: S

Yn: Mt3.S: S

Yn: Mos: S

Amount owed on fourth largest loan

%

*Forexample. loans from famky. other Individuals. and other busrnesa firms shoct-term loans, working capttal loans.
unsecured term loans. demand loans and so on. Sole propnetonhtps should aleo include any personal bcvrowng for busfness purposes.

A. Use of Financial

institutions
(AS

and Other Sources
of

of Financing

(connnuea)

END DATE)

10. Other Services:
If

the

fvm

used

the followlng

services

at an Inslltutlon.

enter

the name

or names

of the Instltutrons.

a.

Supply of paper money and coins for use in operations. Name(s)
of Instltutlons:

NOT USED: 1

I

b. Cash management

services

such as sweep and zero balance accounts.

NOT USED: 1

Name(s) of Institutions: c. Processing Name(s) of of credit card receipts.
Instltutlons:

NOT USED: z

d. Night depository. Name(s) of Instltutlons: e. Brokerage services. Name(s) of Institutions: 1. Letters of credit or banker’ s Name(s) of Instltutlons: g. Other services such as pension funds, business trust services. lock boxes, safekeeping for securities, factoring, sales financing, wire transfers, etc. Do not include insumnce. Name(s) of Instdutions:
SCCeptSnCeS.

NmUSED:

c!

NOT USED: 2

NUT USED: c

Nm USED: z

I

4

--

A. Use of Financial Institutions and Other Sources of Financing (continued)
(As of END DATE) II. FINANCIAL INSTlTUTlON INFORMATION: For each instWtion you listed, please prowde the following Information. ZIP Code of off Ice used most often for noncradit services (if any)

Full Name of Institution

ZIP Code of off ice where applications made most often for credit or financing (if any)

1. 2. 3. 4. 5.

II I I I I [ I I I I l I I
1l I I I i I I I I I I

7.

8.

9.

I I 1I I I I I I I I
I

I I I I I I I I I I I I I I I I Cl I I I I [I I I I I rlllll I I I I Cl

I I I 1 I I I I I I I I l I I I I I I I

1’ . 12.

1 Cl I l ll

5

6. income and Expenses
Items l-5 on the opposite page may be obtained from 1997 income tax returns. as follows: Item
1

Type of Business Proprietonhip Partnership S Corporation Corporation

Form 1040 1065 11ms 1120 1120A 1040 1065 1120s 1120 1120A 1040 1066 1120s 1120 112OA

I

Schedule Schedule C ----Schedule C ----Schedule C ----Schedule C ----Schedule C -----

Line Number(s)

jili!

2

Pmprietorship Partnership S Corporation Corporation

i!i

1

Proprietorship Partnership S Corporation Corporation

Line 3 Line 3 Line 3 Line 3 Line 3 tine tine Line tine tine 30 20 20 27 23

Proprietorship Partnership S Corporation Corporation _

1040 1066 1120s 1120 112ClA 1040 1066 1120s 1120 112OA

5

Pmprtetorship Partnership S Corporation Corporation

Lines 17a + 17b Line 12 line 13 tine 16 Line 16

Item 6 on the opposite page may be obtained from 1966 income tax returns, as follows: 6 Proprietorship Partnership s Corpomtion Corporstlon 1040 1066 1120s 1120 112oA Schedule C ----Line Line Line Line Line lc lc lc lc lc

B. income

and Expenses

(from BEGINNING DATE to END DATE) Item No. and DescripUon 1. Total sales net of
amOUnts

Amount for returned merchandise and allowances for bad debts. $

[Gross receipts or sales less returns of merchandise and allowances for bad debts. Also called “ net sales.7

2. Costs of goods or services sold. [The cost of prcducrng or purchasmg the product sold to customers. Also called “ cost of sales.“ ]

$

3. Firm’ gross profit. (Item 1 - Item 2) s [Net sales menus the cost of goods or services sold.]

$

4. Firm’ total selling and administrative expenses, interest expense, and other s deductions. [These are the costs that are subtracted from gross profit and other income to determine the firm’ net profit/loss (ordinary income from busmess actwities).] s 5. Total dollar amount of interest expense. [Interest paid on loans, mortgages, notes and bonds. Interest expense is one of the expenses recorded in Item 4.)

%

S

(For Year Prior to BEGINNING DATE) 6. For previous year, total sales net of amounts for returned merchandise and allowances for bad debts. $

6

.

C. Balance Sheet -

Assets

Items l-9 on the oppodte page may be obtained from 1997 Income tax returns, es followe: Item 1

2

I

Type ot Business

1
I

Foml 1065 1120s 1120 112OA 1065 1120s 1120 112OA 1040 1065
1120s

Schedule Schedule L Schedule L Schedule L Pall II Schedule L Schedule L Schedule L Part II Schedule Schedule Schedule Schedule Pall II C L L L

Line Number(s) Line Line Line Line l(d) l(d) l(d) l(b)

partnership S Corporation Corporation

partnership S Corporation Corporation

Line 2a(d) Line 2a(d) Line 2a(d) Lines 2(b) - 2a(b) Part Ill Line 3(d) Line 3(d) Line 3(d) Line 3(b) Lines 4(d) Lines 4(d) tines 4(d) tines 4(b) Line 7

3

Proprietorship Partnenhip S Corporation ‘ Corporation
I

1120 1120A 1065 1120s
1120

4

Partnership S Corporation Cofporatlon

112oA

Schedule L Schedule L Schedule L Pall II

+ + + +

5(d) 5(d) 5(d) S(b)

5

Partnership S Corporation Corporation

1065 1120s 1120 112OA 1065 1120s 1120 112Cb4

Schedule L Schedule L Schedule L Part II Schedule L Schedule L Schedule L Pall II

tines 6(d) tines 6(d) Lines 6(d) tines 6(b)

+ + + +

7(d) 7(d) + 8(d) 7(d) + 8(d) 7(b) + 10(d) + lla(d) + 11(d) + 12a(d) + 11(d) + 12a(d) 9(b)

partnership S Corporation Corporation

Lines 6a(d) + 9a(d) Lines 9a(d) + lOa Lines 9a(d) + lOa Lines 8(b) - aa +

.

Paitnenhip S Corporation Corporation

1065 1120s 1120 112OA

Schedule L Schedule L Schedule L Part II

Line Line Line Line

12(d) 13(d) 13(d) 10(b)

a

Schedule L Schedule L Schedule L part II

Line Line Line Line

13(d) 14(d) 14(d) 11(b)

C. Balance Sheet - Assets (As of END DATE)
Item No. and Dercriptfon
1. Cash holdings.

Amount %

[Total cash on hand and in checking, savings. money market deposit accounts, money market mutual funds. certificates of deposrt (CDS). and other time deposlts.]

2. Trade notes and accounts receivable less allowance for bad debts. [Money owed to the firm by customers who bought goods or Services on account.]

S

3. Value of tirm’ inventories. s IMerchandIse held for sale or materials for productlon such as raw materials. work In progress, or finished goods.]

rg

4. Amount of firm’ other current assets. s [Consists of Federal, State or local government bonds: corporate and other bonds: stocks held for short-term Investment (do not include stocks held for ownenhip control or long term Investment); prepaid expenses such as advance payments for insurance, rent, taxes, advertising and operating supplies; and any other current assets which may be converted to cash or sold or used up within a year through normal operations of the business.] 5. Total value of firm’ investments. s [Loans made to stockholders, money owed Q the firm for mortgages or real estate loans, investments in-mat estate not used in business operations, and stocks held for ownership control or long-term investment.1 6. Book value of land and depreciable, depletable, and intangible assets net of accumulated depreciation, depletion, and amortization. [Acquisition cost of land used by the firm for business putposes; depreciable assets such as buildings and equipment; depletable assets such as mines, timberlands. and oil wells: and intangible assets such as patents. trademarks, copytights, franchises, and good will, less accumulated depreciation, depletion, and amortization.1 7. Book value of any assets owned by the firm.

S

S

%

-%

Type and book value of three largest other assets: a. b.
C. 8.

S 6
%

TOTAL ASSETS. (Should equal sum of last column)

.6

7

C. Balance Sheet -

Liabilities

Items 1-5 on the oppoeite page may be obtained from 1997 income tax returns, as follows: Item ) Type of Bueinees
/ / I Form ) /

Schedule Schedule L Schedule L Schedule L Part II

)

Line Number(s) Lines 15(d) + 17(d) + 16(d) Lines 16(d)+ 16(d)+ 19(d) Lines 16(d) + 16(d) + 19(d) Lines 14(b) + 15(b) + loans. mortgages, notes, and bonds listed on aftached schedule for Line 13(b) Line IS(d) Line 16(d) Line 16(d) LOanS. mongages, notes, and bonds listed on atfached schedule for Line 13(b) Line Line tine Line 14(d) 15(d) 15(d) Mb)

la

Partnership S Corporation Corporation

1065 1120s 1120 112OA

lb

Partnership S Corporation Corporation

1065 1120s 1120 1120A

Schedule L Schedule L Schedule L Part II

2

’Partnership

S Corporation Corporation

1065 1120s 1120 112OA 1065 1120s 1120 112QA

Schedule L Schedule L Schedule L Part II Schedule L Schedule L Schedule L Part II

3

partnership S Corporation Corporation
I

tine 16(d) Line 17(d) tine 17(d) From attached schedule for tine 13(b) excluding loans, mortgages. notes. and bonds

4

partnership S Corporation Corporation

1065 1120s 1120 112OA

Schedule L Schedule L Schedule L Parf II

tine 19(d) tine 20(d) tine 20(d) Line 16(b)

5

partnership S Corporation Corporation

1065 1120s 1120 112OA

Schedule L Schedule L Schedule L Part II

Sum Sum Sum Sum

of Lines of Lines of tines of Lines

14fd) to 13(d) 15(d) to 2Ofd) 15(d) to 20(d) 12(b) to 16(b)

C. Balance Sheet -

Liabilities

(As of END DATE)
Item No. and Description

Amount

la. Total amount owed on all loans, mortgages, notes, and bonds.
[Combtned short and long term loans. mortgages, notes and bonds: loans from stockholders (CORPORATIONS); and nonrecourse loans (PARTNERSHIPS).]

lb. Total amount of Item la due within the next year . . . . . . S

2.

Total amount of accounts payable. [Money owed to supplters for purchases the firm made on account.]

$

3.

Total amount of other current liabilities. [Accrued expenses, taxes payable, prepayments, deposits and advances and other current liabilities (not Included In item lb above) which are obligations that are due wlth!n one year.1

$

4. Other debts or liabilities. [All other debts and liabilities, including obligations under capital leases.1 Type and amount of three largest other liabilities: a. b.
C.

S

% t
S

5. TOTAL LIABILITIES. (Should equal sum of last column)

C

Balance Sheet -

Equity

Item0 2-6 on the opposIt0 page may be obtained from 1967 Income tax returns, as followe: Item 1 Type of Bueineee Form Schedule Line NumberM

2

Partnership

1065

Schedule L

Line 20(d)

3a

Corporation

1120

1120A 3b S Corporetlon Corporation 1120s 1120 112OA 1120s 1120 112oA 1120s 1120 112oA 1120s 1120 112oA

Schedule L Part II Schedule L Schedule L part II Schedule L Schedule L Pall II Schedule L Schedule L part II Schedule L Schedule L Pall II

Line Zlafc) Line 17(b) Aggregated with common stock Line 2lfd) Line 21Mc) Line 17(b) Aggregated with preferred stock tine 22(d) tine 22(d) tine 18(b) Line 27(d) Lines n(d) + 24(d) Line 19(b) Line 26(d) Line 25(d) Line 20(b)

3c

.

S Corporation Corporation

3d

S Corporation Corporation

3e

S Corporation Corporation

6

Partnership S Ccwporetion Cofponrtion

1065 1120s 1120 112OA

Schedule L Schedule L Schedule L Part II

Line 21 tine 29(d) Line 26(d) tine 21(b)

C. Balance Sheet -

Equity

(As of END DATE)
Item No. and .Ducrtption Equity in firm:
1. PROPRIETORSHIP: [Total assets minus total liabdities) $ [

Amount
L: .,./.;. ;I

2. PARTNERSHIP: [Total of all partners’ equlty] 3. CORPORATION: a. Par value of preferred stock

f

$ .

b. Par or stated value of common or capital stock

$

c. Paid-in capital or capital surplus

$

d. Retained earnings

.

. $

e. LessTreasurystock

_.

$ (

1. Total Equity @a + 3b + 3c + 3d - 38)

.

$

4. TOTAL EQUITY IN FIRM. (1 or 2 or 30 5. TOTAL LIABILITIES OF FIRM. (Item 5 from page 8) 6. TOTAL LIABILITIES PUJS TOTAL EQUITY. (Item 4 + Item 5) (Should equal total assets)

9

APPENDIX C 1987 INCOME TAX RETURNS

APPE’ iC!

X

C

12 f 2 j : i i 13a 15 14 16 17

C;-oe~s.3;
Rcmrs

0”

0’

o’ :,cefs

cSc2eau.e
__ _._.

i)
b L~SI to01 :‘ec~l _

12 3aiancc b *13c 14 15 16 17

Sa’ aw ano .Gjer

Baa aeo:s (see ~ns!w:.ans)
Relts Taxes Interest

ia
19 ,20 21 ‘ 22 23 24 25 26 27 28 29 30 31 32 b

ia
(set mstructmns
(attacn form clalmea 4562) in Schedule A ana elsewhere on return

1
: J z f j g ‘ ; 2 2

Contrlouttons
Deoreclatlon Less aeveclatlon Deoletmn Advertwng Pensw. Emoloyee Other

for 10% Ilmltatlan)
20 2111 : /

19

21b I 22 1

’ 23
prollt.snarmg, neneflt OCTAL etc mans ~24 25 schedule) lines 12 tnrougn 26 ana ana erter soeclal nere aeauctlons 29a, (line 11 less line 27) . 26 orograms (attacn aeaucttons-Ada

I
:

ceauctlons

/ 27,
, 28
S?%. i

Taxrole !ncome belore net owratmg loss aeauctlon Less: a Net oDeratIng loss aeductlon (see mstructtonsl b Soeclal Taxaole Payments: 1987 mcome TOTAL aeauctlons (Ilne

(Schedule
J) credltea

C)

29bl

29c 1

20 less fme 29c) overpayment tar payments to 1387

TAX (Schedule

VI T

a 1986 estvnated

: (
I (attacn Form 2439) b E

Less 1987 refund Tar aewslted
l

awlted for on Form 4466
wttn Form 7004 companies Investment

a

0 Crea~tforFedcraltarongasol~ncaodsoec~aIfuels(attachForm4136) c. f z 33 Enter any PENALTY for underpayment of estlmatea
t! .3b 35 36 TAX DUE-If tne total of ltnes 31 and 33 IS larger fme 32 IS larger than OVERPAYMENT-If

Crealt

from

regulated

’ I
11 Form enter 2220 IS attached OWED AMOUNTOVERPAID Refunded b i 33 1 34 1

tax-cneck tnan tar

line 32. enter AMOUNT b

the total

of hnes 31 and 33.

Enter amount of hne 35 you want Creditrd ,I*. j bClllf 8, 0%

to 19gg estimated klrrrrlon

Please

unocr oenu0~101 0erw-f

correcr. ano cornw1e

I mcur~ tmt fifurt ~m*mwa mr mtum, mum*

01 DltDIler (otncr !“a” tl.w”er)

xccomomrq mCd~4es *nG stmmcnts. ma 80tnr oc*t01 9
II wsw an *II lnlWmltlon 01 I”0 veDIrc,

I

i ( 35 1 1 36 I
nrs l“” k”orwase
l

I

no-mg~ ma

2 Purcnases 3 Cost of labor 4a b Aodltlonal Other sectjon 263Acosts schedule) 4b ooeratlons-Lone closq 9’ ::i’ jcczs
used a-c

.~

2 3 4a 4b 5 6

(see ~nstruc:~cns)

costs (artach

-

5 -xal--tda 6 ~nwlfo’ ~ ga Crec* I)
,/‘ _J

lines 1 through a: ena of year

7 Cos: of goods sola andfor aii me!noos = = _ Cosr
:‘ i:~tecown

5 less :ffie C E-‘ e, ‘ r.er:o.‘ . :’ -:‘ *+I ai ces:.
a::a:l

ne’ e

a-o

or ‘ 1-e 2

-age

!

7

used for vaiulng ;.I)
of

5
melnO

ioher

3s Icsc’ ?ea 7 se<_
i

3ec

G+jd~2:rO”s

sect.9”

i

27:

2 iSee

lrgi’ _i’

~7;)

wonorfnal

ations
l

se::.cn

i ‘ 7

1.2(c)

(see

instruct~on~j Form 970)

vv:
b Crecx

Other

(Specify

e~Didna:,or

‘ tne LIFO mventory f _ CO ,n.entory nverlcry cocpoutea

me!nOd

was a ci3:ec LIFL?

:P,s :a,

*ear

‘ oraly

gcocs

(e! cnecked. ior a-ountsj

attacn of

c

!I :-e c cs(rg

rnevn”c
(wi!n

.vas usec unoer
rewec!

‘ ‘ s :Sz rear :’ ‘ :.:;P’ 1 :. ;r;C_;+:

e-rer
GI

oe’ ceq:age

ac
z”es --..c

d Co:-e es~1 ,L: iec!jon 263A

ac:~,rec ‘ ‘esale) c’

aooly to tie coroora[lon’

2 3ebt 3 Cer:ajr 4 Fowgn

finawec orerevea

stock of domestIc and cenaln 1 through suolect from foreign foreign alvldenas

ana foretgn iSI j,o!ec!

ccrooratlocs to sectton

(sec:!on 245

246A)

s:ock of ounl~c u!‘ l~f~es deductron ant (cl) !: .‘ “04 _ aeauctlon (rec:lons 24k3) (sectton wOoan 243(a)(3)) in ltnes 4 and 5 i (attach Forms 5471) 246(a))

c3roorations lines groups

5 Nholli a*neo fore’ wos~a16ries ana FjCi i~iec! q~ 6 Total-AQd 7 Afflllatea to tne 100%

5 See mstruc!lons deductton corooratlons corporations 78) not mcluaea

for l!mltatlon

a
IO 11
12

Other Foreign IC-DISC
Other

awdenas dlwcena

not included under

9 Income

from controlled or former
cwaecas

gross.up(sectlon DISC

in Ilnes

1 and/or

2 (section

13 14 15 m

Oed.;:,on

‘ ofvlaenas patd on certain Drefcrlea stocx 01 wwc 01

ut~l~l~ef mrtruct~onr) iree .
1 1.) 10. of page 1. Form 1120)
, (1) Amount Of Eomoe”la:lO”

Total aivldenas-Ada Total aeauctlons-Add

lines 1 through 12 Ever here ana on lme 4 oage 1 Imes 6. 7, and 13 Enter neie and on line 29b. page

Compensation of Officers (See mstructtans far hne 12. page Comotete Scnedule E only If total recelots (Ilne la. DIUS lines A througn
(I, Name Of ocfcn fb, sa!al ,K”I~1” ““nlDw tme aswxca aullnerl

are 5!50

000

or F-ore

cc, Percent 01

, 10 1

Percent 01 coIwrI,Ion llo.3 mnca t*)Common 1 f.)Prelorea

, I

961 %I %I %’

%i %I %I %! %I %I %I

%I 961 41 %I %I %I %I

I I I Total comoensatlon
of

I

%I %I %I

officers

-Enter

here

ana on lme 12. page

1

1

FVrn

,::01:9an
Tax Computation (See ~nsrrucnons )
group

;>gc 3

1 2

C’ leck 11you are a memoer
If lhne i IS checked. 0:newse. complete

of a conrrollea

(see sec:lons 1561 ana 1563)
InCludeS June 30. 1987. Complete both a ana b below

-,

see InStruCtIOnS only b.

If your tax year

a,iS
b,l %

. . . . . .._.___.__ (11s _. ____ . __. . . . . . f~~0 s . . . _. __. _. . . . . .+I ) .._. _....D~) s . . .._.............
:ax (see instructions
6 less) rax creait cr,< for :.ec,t Check (attacn lartacn

s

. . __ _. _
Scneoule D

3
oa b c

-izme
:.-#:-rLer

10 hgure
Scheaue Form Form from

tne :a~ enter
0 57351 b _

this :ax or alternative

!ax

from

If from (attach

i :‘ gn e’ Z.z-an

1118)
Form 6765) a nonconventional source (see inare

4a b

‘ :ssesS~ons

:ax creo’ t

a
e

C,ead
S~‘ ,i:iOnSi General 3lacnec F---

r-em Droaucea

bLs)ness crec!l En!er here and check wnfch forms :orn 3800 1 Form 3468 1 Form 5884 _ 5;;s _ 'ormtj76j 3526 Cy‘ 17eS 4a tnrougn be tar (artacn Scneoule c’ ealt PH (Form (at:acn Form 1120)) Form 4626) oage J 1 4255)

e

5

T-z-_;ac Derso’ al Tjr :rom C’ :erra::ve

6 7 8 9a 10

L -e 2 ,ess iune 5 noiolnq recc~w:~lg rrln~murn comoany 7 przor year rax(see Investmen:

a
9a 9b I” R&r 10 IlIt 1111I” the ~nslrutt~ons tno slate the o,,nc,oa, w b Busmess acwuy Yes Ho

~nsrruc:~ons-atracn ne’ e

b Ew~ocme”:al Tzld~ :a@---loa lnlormatlon

tax (see Instructtons-attacn lines 6 rhrougn Vb Enter (See lnstructlon F )

iorm 4626) an0 on Ime 31.

Addltvxal H

YcslNot

DIO :?e :xorw?

C~NI I aeauctlon Ior es~enscs conneclea wllh i

_.
. a ll S sharenolcu o! any conrroliea . .

. : 1’ 1 lowgn

rtsorl ranch. tic )’ (1) An c~!trtatnmenr lacilily (boal. LI” accomnloo1tlo”s (trcwt L(PoIoyccs0” buslness~’ irl~ (2)

hJuclorstwlcc
K
Was me corrmra11on coroorauon’

(3) (4)

Emo~oyccsarrcmnp Amtrcan II Yes

conv~n~~onsor

metlmgsouts~ae

Ihl North

area'(SK
we

seWon 274(h) )

Ccc sections 951 ana 957 ) If “ Yes.” attach form 5411 for Cacn Such corooratlon have an m!eresI accoun1 m a ICCOUIII or or other authorlry (such account)’ Ior form 73 F 9” ::
l

Eno~ovees lamfles

al convcnl~ons or mectlngs?
or meetqs oulr~dc the 274(h) 1 or

At any Ilmt durmg IhC tas year did IhC coroorallon
m Or a Slpoaturc

any of these convtnl~ons

over a hnanclal swrwzs

North Amcrlcan arca) (See sechon (5) E~JIOVH or lamely vacabons *c#rec!!~ :;.5oraifon’ , ;a dsi YeI wow iG’ , anacn or now a scn(auI(

fOrt@n counlry olhtr fmanual

as a bank account.

nol reported on form W-2’ 01 a donwstlc aaowsf. mom ana or

1 (1) Did ihe tor~rdf~on II the cna 01 IhC 111 year own olrectly
01 IhC vOfmg IlOck snowq owned ataucllons (a) name For cuxs 01 aftnbu::on Ib) pwtcnragc set sectloo 267(c) ) (cl Iaraak

(See mstrucl~on f ana f111ngreaw.nenn
II “Yts ” tnttr name 01 lorcqn country Was the corooratlon which If “Yu.” Durln( tr!Sltd durmg coroarallon thn tne granlor IhC current mlcrtst

Ij _

q

of 3’ “ aweror lax year 10 11’

!o a rorc,~q ‘ .;s: whelner or no1 Ih( ; ,,’ ‘ $,,,

cclrNIv~ng numow

has any Lwnclmal Iht corwrallon Iar yof.

NO1 ana sowal

of such corooral~on lor

may have 10 llt( forms 3520 3520 A or 926 did Ihe corporaldon m cxcnanp gay c,:,arnas (olhw man lor s1ock) in e,cCss 01

1% ial year cnolng ullh or wlhtn your tar year. (d) hqhul dmour~l owa 01 Inc coroora11oo 10 such corwral~on durmp IhC
,I~, :~‘ x:a:m alia le, qntst awng amoum owed lo lh+ Larporatlon 101 par. oawwshlo. I rotq corporalton. wale. or trusl at 50% or more see (a) or md~rwlly. (a) through by such

i,‘ ‘ y?$,,j $ , T ;, * 1,.

Ilockdwdtnds and dlstrlbullons
corwratlon II “Yes.” Ior och During secr~ons 30 I and 3 16
hereIorpaw*1

m

I 3,4;p,< t’ ;,’

s currenl aoa accumulaita

earnings anc clol~ts) (:c( rerun answtr Scnedulc.

)
ana on form 851. AllMlalloos mamlam rystcm’

(2) S,c anY malwaual.
01 :1c cormation rccrlon la) (3) 267(c)

fllc Form 5452. If lhls IS a Consol~daMo corcmral!on Subsldlarv lhls 1as Year did the corporalloo records on a compuMrutd a Accrual __ _.___________

:?e cno ot lhe lax yuc mvn. dwclly

stock? (for uks

ol alhbubon.

1If “Yes.” compklt
thowml

any part 01 11s

Alrach a scncdult

name. address ana ldcnllfymg b _____________ stock a wrson dher than a Not@: II ‘ Yrs.’

accountmp/lar (1) (310 mstrumcnts It so. 0

pctccnla~towncd number Enrcr Was Iht owns 01 such volq U S. wrson? corawt~on (Set may h&t amount amount

Check method of accounlmg Cash (2)

~nslruct~ons.)

fht

Ofher(sgcclfy), wlh orqmal

_______ debt -

to Ilk form 5472 owed by Ihe COrpOrallOn 10 such

Check UI~S bar 11 the corooral~on

~nued oubl~cly otlcwd

If “Yes.
(c) fntcr

tntcr owner counlry b _ _ _ _ _ _ _ _ _ _ _ _ _ _ I
hqherl hqncsr

ISIUC d~scounl

owntrawnemcycar (d) Enltr
Note:

,.._~...~._._~..._.~_
owed IO Iht coroorallon by such

owner durmg the pear b

Iht corporal~on may have lo lII( form 8281. fntcr Ihc amount of Ias+scmol mtercst recued or accrued durq Iht lax ycrcb . . . . . . . . . . ..I.........
If you are a mtmbtr 01 a coolrolled group p (row. enter the amount of taxable mcome lor the entut

__ __ _ _ _

for wpsts 01 l(I) and l(2). “hrghuf amount owed rnclodcs IWnS and JCCOUnh recervable/Davable.

. _ _. _ __ . _. _ _.

,,

1 Casn 2 Trade notes and accounts recelvaole a LeSS allowance tar baa Oebts 3 Invevanes 4 F&era! apa staw governme,,! iirl jarIons 5 O:*e, :_“ enr assets fanacn

scneouie).

6 Loans to Stocknolders
7 Mongage 8 Othe, an0 real estate uans ia::acn SC-ec~:e) asse[s ~nvesnnents

9 Buil01ngs an0 otne, oecreciaofe a Less accumulatea aeoreclatfon

10 Deolefaole

a

assers Less acc~-_ are< oeofetlon

11 Lane (ler 01 ari a*Qr~rar~on) 12 In:arqoIe assets (amor,lzaoIe only) a Less ~CCL?‘ +s --.. :311on 13 Other assets car:acl 14 Yolal asse!s
15 Account 16 Moqages oayaole TOWS DWCSoayao~e in ws man I year

sL-ezuie)

17 Other current

llablijlles farracn s:-eoule) 18 Loans from s:ockwoers saydole !n I year or more scneQuIe)

19 Morrgages “ores >o-:j
20 Other l~aowt\es (arracn

21 Caolral stock

a b

preferred stock common stock

22 Pata-ln or capital surplus 23 Retalneo earnings-PoDroprtaled (anach deoule) 24 RetaIned earnings-Unaoproprlaced

1 Net Income oec nooks

2 Federal fncome tax 3 Excess of carxtal losses over capital gams 4 Income sublect to tax not recoraed on books
this .-. Year WmW recorded ._____.___ on books ____._____.__ _.-.. this year not . . . . .._..._.__._._._..~~.~....~ I” this return(Ittmw)

7 Income recorded on beaks this year not I mcluded m thts return (Itemize)
a Taxexempt mterest S .__.___.___.__ 7-, ._._----__--.____________________._ .__.-~-___--..____.____._.._______ 8 thzductlons on thts tar return not cha,ged agamst bwk mcome this year (ItemIre) a Otpreclatlon

.

5 Expenses aeclucted

a Deorectatlon

s.. _.___ _-...

_ _...-

s......_.... j b Ccntrlbutlons carryover S _ __ __ _ _ __ _
.____.______.____~_.______________ .________________.________________

b Co~t~~but~O~SCaf~O~~~.................. ___.___.___.....__..~___I________________

9 Totalof kna7and8

complete

ttus schedule I( the total assets on lme 14. column(d).

of Schedule

L are less than $25.000.

1 Balance at begInnIng of year 2 Net mcome per books

5 Dtstrdutwvs: a Cash
b Stock c 6 Other decreases Froputy (itemue) ___________

3 Otherincreases(Itemlre)

._________________

4 Totalofbnes

.._._~_____~...____.~~....~~~~~--.~~...~. .-~__.~._._-~__.____.~-..~~____----~~---..~_.~~.___-.__.._-.____________--.~~-~--.._._--_.__-__..__-_-.~~--.----...~-----1.2.and3

I

___________-_____--_______________ ____-._____--____----__--..~~~---~ 7 TotafofiinesSand6 . 8 Balanct at end of year (Ime 4 & line 7)

1120s

i

U.S. income Tax Return for an S Corporation

1 1; z
: 2

2 C;s: 01ipxs
3 4
5 6

soia ana,or 0De’ at~ons iSc-ecuie A Ihne 7) Zr:ss orotit (sdctrac: iine -7 “ :n line lc) :;e-i~ *aIn 10’ 05s) from iorn ‘ 797 Ilne :a isee .nstruct~ons)
s:reculeJ *es j f ana 5 arc enter mere ,I

3 4
5 6
7

CT-er ,lcome rsee ~nstruc:~o” S--a!tecr TSiAL income ,loss.:;;-z,-e
:_-Densa:.on or 5tt1ce1s

z 2

7

;;i
E

8a Law dna wer 9 Pr-a,rs
10 11 kc Revs

b Lessiooscrealt

__

_.

__

___

Balance

b

8c
( 9 11 : I 12 I

aeots (see ~nsrruct~ons)

= h

;;

._
Decuc:N31e Interest exoense not cialmec or *eooneQ elsewhere on return (see instructions) ,14al
1 14b I

$, 13 7
k, 2’

13

1

z, 14a Deoreclatlon
b Deoreclatlon
C

from Form A562 reDonea

(artacn

Fork

4562). on return

on Sct?eode

A ana elsewhere

k,

Suotracr
Deoletlon

:lne l4b from line 14a (Do not deduct OII and gas depletion. etc plans See mstruct!ons

-z I5 5,16 g 17 .$ 18
‘ 19 ;,

)

,:;~df+

15 I 16 I 17 I 18 I 19 1 I201

Aadertismg Pension oraflt.sharmg.

;
a u?
z

20 21 22

E-otoyee oeneftt orograms C’ ,.>. ‘ I-:_ C!‘ OPS (attach scneaule, TCTAL aeciuctlons-Add lmes 7 tnrough 19 and enter nere 2~: narv lncorre (loss) from traae or ouslness actwty(ies)-Subtract Tax
a Excess net passwe b Tax from Scneaule

..
line 20 from lme 6

/ 21 I

mcome

tax(aftaCn

Schedule)

D (Form 1120s) c Aad lmes 22a ana 22b E, %’ 23 Payments 0. a Tax beoostted wtn Fornl 7CJ4 :;’ fn b Crew for Federal tar on gasohneana soeclal fuels(attach c Aca lines 23a and 23b : + 24

Form A1361 for Pay!

25

TAX DUE (subtract lme 23c from lme 220 See mstructtons OVERPAYMENT(rubtract hne 22c from lme 23c).

Please
Sign Here
Paid

De141 C.correct. mnacomDlctc otcllrrtfon It II ~~“
;b

Llnaer cmunnotaw.

. . 25 I I arcun?t~t I hJ” *cummeatnlr mwrn ~nc~“ a~ngacc~m~n” m~Xn~“ ,M,tatemm,,. anaW,“ aes, In”snovdcbge M C 01 rna
0, vepdr.?r (otncr llu” IlwayCo I, naI.ra on ,I, l” fonlwlOn
0‘ WIlS”

Drcpwer nrr I” knorw.a(e ”

SlgMt” 01ofllcer re

Date

1

TIIIR

USC Only
*II

h ; Ilgnarwc Prwrcr’ i Firm s I rumc,or

Pwomf S

041.

cha

sellall.

Ii

Pfemrcf I ,oclll1ec” r~l” “ m&r “

; “ If s.?l‘ our3 -cmo~ea) 1 ’ andadare3,

ww@O 1 E.I.N.3 l I ZlPcme &
form 1120s
(19871

..^ . ..“.s

.3E“

:

.;A

2

Cost

of Goods

Sold and/or

Operattons

ijee

~*s;ruct~ons

ior Scneaule

A )

Addltlonal

lnformat!on

Required 5:‘ ~ or more 01 tne vor~ng s:occ ct a aomestfc corporation’

Yes

N0

J -D? ,oti 1: .-* If (1) (2)
K

e+3 2’ :-e rar year own c’ ,ec’ , 1’ -: ‘ ec::.

vej Y2-e

3::3c- 3 r;-ec,le i” 5nlng :z;.~Is 2-c err3ioyef ,oer:~hca11on -.-‘ :t’ :.vnea

(3) p <“ es1 arrounl OweG Dv vou !o sue? coIporat!on during the year. and (4) H,glest amount owea to you ny sue:: corooratjon aurlng the year

Pe,:evaje

(Note: For ?mx_xes of J(3) drd Refer to :*a us:,ng 01 busmeSs acltvq

-~?esr dmount owed" mchaes loans dnd accounts recervable/payable.) J(Q) codes al :ne ena of the Instructions for Form 1120s ana state your prmclpal

Product or serwce . . . . ..__...___.__..______...___ Suslnw 3:: b’ ä :. ‘ . L VJere you a mevoer Ola COntrOlledgroup suD(eC: t0 tne PrOVlSlOnSOf SeCtIOn 1561? M Did you c’ aim a aeauctlon for expenses connec!ea wlfh (1) En:ertalnmev facllltles (boat. resort. rancn erc 1’ (2) (3) (4) (5)
ouslness)' L . Iqaccc--oaat~ons(erce~tforemo~oyeeson E-Z :.elis r!!er,cmg conventlons 01 meer ‘ ;i:sIce the North Amencan 45 C-“ d r.ryS _ ” ‘ amflies at conventlons or meerlrjs’

area’ (See sewon

274(h)

) )

:.ere any o( these conventions or reetings outsde 11 yes ETolovee o, family vacations not reported on Form W.Z!

the North Amencan

area’ (See sectlon 274(h)

N At a?‘ !‘ ” me au’ lng :ne tax year. ala you have an Interest I” or a slgnafure or other authority over a fmanctat account m a I ‘ orelgn country (sucn as a bank account. SeCurltleS account. or other flnamal account)? (See lnstrucnons for exceptlons ana l111ngreaulrements for form TO F 90-22 1 ) if Yes +“ ... . . _ _. _. . _ _ _ . . . . _. _ . . i.%%i~ lef ine rme of the lOfew cOuw c . . . _. 0 Were you the grantor of, or tranSferOr to. a fOrelgn trust wmch emted durmg the current tax year, whether or not you nave an, nenetlclal Interest cn It? If “Yes.” you may nave to file Forms 3520.3520.A. or 926
During tnls tax year did you mamtam any Pan of your accountmgjtar records on a computerized system? (2)2 (1)a Cash Accrual (3) a Other (spwly) w . . . _. . _ _ _ _ _. . . _ . . . P Check memodot aCCWW!g: R Chew tnls 00x II the S corporation has flied or IS reaulred to file Form 8264. Appkatlon for Reglstratlon of a Tax Shelter s Check tnls box 11the CorporatlOn Issued PubMy Offered debt instruments wth ongmal Issue discount It so !he corooratlon may have to ttle Form 8281. lntormatlon Return for Publicly Offered Ortgmal Issue Dtscount instruments

:.,:

P

.

T

II sectton 1374 fnew bu1lt.m gams tar) appks aeftnea I” sectlon 1374(d)( 1) (see tnstructlons)

to the corporatton. w

enter the corporatton’ s

net unreahzed

budt-m gam as

Designation of Tax Matters Person The followmg sharenolder IS hereby designated as the tax matters person (TMP) Name of aeslgnatea TMP )

for the tar year for wh#Ch this tax return IS flled ldentlfying number of TMP 1

Aaaress 01 ceslgnatea TMP 1

;,rn

I,:“ :

:9:-l

3 Sharenolde_ls_?hares of Income, Credits. DeductIons. .,a, D,str,butlve share vfcms etc. 15% ~?s:r~_:,cq: _____ --_ -.__ --_
101 Tow amount

Income

~Larscr)

and

Deductions

1 2a b c 3a b c 4 a b c d e I 5 6 7 8 9

Ordmary mcorne (loss) from trace or Dusmess act~w!‘ ~l~es) (Dage ! ,a 2 1) 2a Gross mcome lrom rental real estate actwry(~esl tb Minus emenses(atiacn scnecule) Bafance ne! tccye (loss) from rental real +‘ 3:c act,. tji~esl Gross income rrom other renral actlwtyfles)
Mtnus ewenses

tattacn

scneoule)

3b

Balance ne! income (loss1 t’ 0:Per rental a,::~, rvc esi om PorTtoll ~“co-‘ ossI ~‘
Interest Dwaena Royalty Net snort income income income rc’ m ca?‘ :a ga ;SSI (SC-~:, 1 ; ‘ Form 1 IZCSli ! :_‘ I: ,

3c :
4a Ob 4c

,1,,

i

4d
de Of

31 Ne! lorg ret- cz? ‘ salr Vossi ISclecL~e C :C*,n
Other pont0180 ixov (loss)
set: on

tattacn

scneoulei

Ne! gain UOSSI dnce,

.231 lorrer ‘ -a- cl,e :o :asualty or t?ett)

5
6

O:ler sn:oce ~551 (attacl 5cIeauIe) C”aritaose :z”.’ out’ 0~5 !di:acn sc-eaulel Sec:~on : 79 erswse cea,c: on (attacn sc*ec_lel Exoenses re a:x :o 3onrol’ wome (loss1 cat:acn scnecule) (see instructions) o ^

7

a
9 .^ lla llb
IlC

lla

b c d e
12

crec : .~::ai? Form 5884) Low-Income polsq crecst (attacn Form 8586) Quallllec re?aoll~tat~on eroencltures relatec to rental real estate actlvQ(les) (attach schedule) Crealts relatec to rental real estate actwty(les) other Inan on lines 1 lb and 1 lc (attacn scnedule) CreClt(s) retalec to rental actlwty(les) other man on lmes 1 lb. 1 lc. and lld (attach scnedule) Otner crealts (attach
ScheCule)

Joos

lld lle 12
; 13a

I

Tax Preference
131

and Adiustrnent

Items ! 13b 13c 13d 13e(l) 13e(2) 131 ; 14r 1 :14b(l)l ‘ 14b(2)1

Acceleratea oeoreclat!on of real property placed in sewce before 1987 b Accelerated depreclatlon of leasec personal oropeny placed in serwe before 1987 c Deoreclatlon ac1Lstment on orooew placec rn serwce after 1986 d Decletlon (other :?a” 011 ana gas) l (1) Gross inccrr, tram 011, gas. or geothermal DropeTlles (2) Gross aecuc!ions allocable to 011. gas. or geothermal propenles f Otner Items (attach scnecule) Investment Interest

14a Interest expense on Investment debts b (1) Investment ~ncorne mcluded on lmes4a through 41. Schedule (2) Investment eroenses mcluded on Ime 9. Schedule K Foreign Faxes 15.3 b c d

K

Tyoeotincome.................................................. ‘ .~..~~~~.~‘ .~~~.~. Name of forelgn country or U.S. possessnon . . . . . . . . . . . . . . . .._____........_.______....___ Total gross mcome from sources outslde the U S (attach schedule) Total appkable deductloos and losses (attach schedule) l Total foreign tares (check one): b 0 Pald E Accrued f Reduction m tares available for credit (attach schedule) P Othercattach schedule) Total property dlstrlbutlons (mcludmg cash) other than dlvldend dlstnbutlons reported on lme 18. .,. ScheduleK Other Items and amounts not Included m lanes 1 through 16. Schedule K. that are required to be reported separately to shareholders (attach schedule). Total dwldend dwlbutlons pa!d from accumulated earnmgs and proflts contamed In other retamed earnmqs (Ime 26 of Schedule L)

! 15c 1 15d 1 1% 151 lIza

I
i

I I

16 17 16 nn7

Assets
1

,a,

CD,

,CI --

IO! -_

casn
and accounts recewable
i..
i ,,

2 Trace notes

a mess allowance tar bad debts
3 lnvellorles 4 F.geral and sta:e governrrb-: 5 :.,-or ~‘ re?t
/ ,I

i

oOl@atlons

,,

asse!s (attacn scheaule)

-

-

6 COJVS io snarenolders 7 r.!oc;age ano real estate loans 8 O:ne, snbestme-7s (dttacn scheaule)
9 al,, J -;s a iess ana orwr aeoreciaole assets

-_

accimulareo assets

aeoreclat10n

10 De3 t’ a3Ie 11 b-3
12 a

a irss acttimwa:ec awletlon -e: 2’an. a~orr12ar10n)
Ir:a-g 3 e assets ?j;s iamonlzanle only) .?ss act-muta:ec

-__ : ,,1,

._

amoflizat\on
: .,

13 o:,-;.

1x2~7 scneaufe)
I’ ,
I,, ;

14 TOT.5 d5ibIS

Liabllltles

and Shareholders’

Equity

15 Accs_rrr oavao’ e 16 Mcc’ gig+i -z!es DOW savaale in less than 1 year 17 Ot:ev currev IB~DII~IIPS(attacn scnedule)

18 Lows :*om sna’enolders
19 Mor!gages ro:es oonds oayable ,n I year or more 20 Ot-er IlaDllltles fattacn scneaule) 21 Caolral stack 22 Pass sn or caoltal surplus 23 Accti.rulated aalustments account account 24 Otne, aalustments

25 Sharemolders’ tina,strlbuted taxable Income 5wwo.s’ . raKea 26 a’ .-__ ‘ c: ea,nsngs (see Instructions) 1, _ .e;. :“s 3or I !ne corooration has sub. :‘ .axer 2 earnlnes an0 orotlts at the tlose of :-e :aA year w 1 (see Instructtons) 27 To:~I ‘ e!dlvu tarnmgr MI ocorr-Comb~nr amounts on ,!“er :3 rnrol@ 26 ~OIUlll”5 ana (0 (Kc InrtructIonsl (1, 28 Less cost of treasury
stock.
i;

i _. -.p!” ’
I

-.

‘ ,,<>
I) ,f

I”

i

‘, 1

I,,;,

Analysis of Accumulated Adjustments Account. Other Adjustments Account, and Shareholders’ Undistributed Taxable Income Previously Taxed (If Schedule I.. column (c). amounts for lmes 23. 24. or 25 are not the same as corresponding amounts on hne 9 of Schedule M. attach a schedule explalnq any differences. %e instructions.)

1 Balanceat begtnnmgof year 2 Oralnary mcomefrom page Llme21 3 Other addltlons .
4 Totalofllnes1.2.and3

5 O!strlbut!ons other than dwdend
6 Loss from page 1. lme 21

dtstrtbutlons

7 Other reductions . . 8 Aodlmes6.6.and7 9 Balance at end of tax year-Subtract he 8 ‘ ram lme A

2

20 21

_-is c‘ :‘ s: rrearury stock - ‘ 63,’ ves rna s~ocknolacrs -?Q” 3, lfY

, : ,l,’ j (

Reconcrllatlon

of Income

per Books With Income

oer Return

I Must be completed

by all filers)

!I

il

11 11 il tl

tl

121212131313131313131405(51515156i6~6i7i~:

11 tl

1212121212(3(3)313)31313(4)4)5)6)6/717)7)7)7171Rt9(1nlttr

la iross retmls or sales f _ _ _ _ _ _ . . lb Mmus returns ana allowances I 2 Cost of gooas sold and/or ooerat~ons (Scneaule A he 7) 3 Gross profit (subtract Ime 2 from line lc)
E s 2 4 5 6 Oralnary Income (loss) from otner partnersh!os and fduclarles
bttach Schedule)

_ _ _ _ _ _ _ Balance b

1 lc 1 2 3 1
/ 4 5 6
17’

I

/

7 8

Net farm proflt (loss) (attach Schedule f (Form 1040)) Ner gam (loss) (Form 4797. lme 18) Other income (IO%) TOTAL fncome (loss) (combme lmes 3 through 7) Salawsana wagcr(otherthan tooartnen) Zuaranteea payments to panners ?eqt Zecuctlole Taxes Baa debts Reoalrs Deoreclatlon from Form 4562 (attach Form 45621 ~~a~meaonSchedulcAandelscwhereonreturn
Deoletlon (Do not deduct oiland ps depletion.)

a
I

S.

gc _ _ _ _ _ _ _ _ _ 9b Minus 100sCrMlt s _ _ _ _ _ _ _ _ _ _ Balanceb i 10 1 /
e#sew*nere on return (see lnstructlons)

Interest exoense not clalmea

S _ ___ _ _ _ _ __ _ _ _ _ _ _ _ 16b Mmus acpreclabon s ..___..~......._.._._~~_._8alance~

) 17
tiaat

I

Prtlrement

p~dns. etc.

Employee benefit programs. Other oeauctii(mttach schedule)

j labI I 19 !

1
2 3

Inventory at oeglnnlngor ,ear
Purchases m!nus cost of Items section 263A costs lb) !:ne 6 from we 5, E-ter ne’ e ano or cage : .pe 2 witndrawn !see for oersonal use Cost of labor Instructions)

1 2 3 ‘ %a 4b 5

4a Adaltlonal
b Other 5 6 Total (add invenrory Check (I) (Ii) (Iii) z 1 -

costs (attach

schedule,

lines i through at ena of year

6
7 clcslng inventory in regulattons fn section ! 2: L 2 regulartons semen 1 47 1.2(c)

7
8a

Cost of gooas sola (subtract
all me!‘ ocs cost Lower

u5eo ‘ iamurg or

of cost or mar-et of ‘ subnormal

as oescnoeo gooas

Wrlteaown

as aescnoea

(iv) ’ Otherrsoeclfv metnoa used and anacn exolanatlonr + was aaoo:eo !fr,s tar vear ror any gooas (of cneckea b Check ,f tne LIFC .nven(oc’ -emoa f
c Do tne rules 01 sectton 263A (wtn resoect :o orooeny DrooLcea or acculrea for resale) quantltles.

..__
at:acn apvy Form 5701 Detween Yes 1 No to the oartlersruo’

Yes z Yo

d Was there ar‘v cparge
ooenlnp ulncorne 1 In !ne soace P,ooeny Prooeny A 6 tncome orwoea ano c!osjng

corler ‘ -an ‘ Secnon 263A OuWOSeS)m or ~n~e”~orv) If ‘ Yes attacn erolanatlon oemo*. wow tne kvw anC .ocat’ or 3’ eacl

aetemlnlrg

Cost. or ValuatiOnS

rental

orooeflv

Attacn . .

a scneaule .

11 Tore ..__.

soace

s neeaea

_.
ProDe7kes A B C 2 Totals lAda attacned columns A 6 C

P*ooenv C Rental Real Estate
and amounts from any
schedule) 2 Gross Income 2; ..

Aavenlsmg Auto

and travel
ana mamtenance

Cleanmg Commlsslons Insurance Lega, fees Interest Reoavs Taxes Utlhttes Wages

ana

c:. ef

votessfo:;31

ewense

an0 salartes from Form 4562

Depreclaoon

17

through 15 Net income (loss) from rental real estate actlwty(les) Subtract lme 16 from Ime 2.

all properties Schedule K. hne 2

pale 3 Partners’ Shares of Income. Credits. share Deducttons. Items coaee etc. fb) Total amount -

(a) Distributive

1
2

Oralnary Income (loss) from traae or busyness actlvltv(lesI
Net Income (lass) from rental real estate actlvlt@s) b Minus emenses c Balarcr (attach (toss) schedule)

(Scireaule

I ‘ 2 ii ce F l\ne ! 7)

1

3a Gross Income from other rental actwnty(les)
net Income (loss1 from other rental actlvity(~esI . 3c

z

6 i
E 2 5

4

Porff0110 (“ come

I ~,“,/,6 ,I : “, .I ,
4a

a Interest
b Olv,aena c Rovalty

Income Income Income caoital gal” (loss) (Scheaule 3 line 4) D Me 9) -

d Net ShOFte’ m
l

44 4t 41 5 6 7

Net long.term oontoho

capital gal” (loss) (Scheaule income (loss) (attach

f Ot’ ler 5 6 7 ;, iz w P a ;. 8

scneau:e)

Guaranteea

oavments ~231 (other!nan Ilst) oro0eny (section !79) Invesirrent r:e-est ex0ense) cue to casuai:y or men)

Net gain IIOSS) under sectIon Other (attach schedule) Charltaole Eroense Otrer contriout~ons ceauctfon

ianacn

a
9

9
10

for recovery

Decuctlons la::ach

related to oontollo scneaule)

Income (Co not ,wu0e

10
11
12a

11
lta

-

Creolt for income tax wltnheld b Low.lncome nous~ng crealt (attach repaollltatlon Form 8586) relatea to rental real estate actwty(fes) (attacn

12b
i(~$,,~,~+,,,~ lzc

u

Y % z

c QuaIlflea
scnedule) d Creoit(S) schedule)

exoenaltures

i

related

to

rental

real estate

actwty(!eS)

otner

tha‘ n

lib

and

12~ (attach schedule)

~&jjfi

c Creait(s) related
13 -5 zz 1 1: p\ s Other (attach

/
to rental actwty(les) other than 12b. 12~. and 126 (attach

12d
12~ 13

i

schedule)

5

, 141 Net earmngs (loss) from self.employment
b Gross farmlng c Gross nonfarm 1% Acceleratea or hshlng mcome Income of real propeny of leased on propeny placed in serwe oefore 1, 1187 beiore ! ‘ 87 !,‘

~ 14r
/ 14b 14c

/ !

I
personal oroperty placed in sewce

deoreclat~on aeorectatlon adjustment

b Accelerated

ZE’

5,

15a 15b
/ 15c

I

c Deoreclatlon
d Depletion

placed m serwce after

12. 3,. d6

,i=
: k

(other than 011and gas) Dropertles propemes

I

c (1) Gross mcome from 011.gas, and geothermal (2) DeductIons
allocable

15d l%(l)

I I

to 011.gas. ana geothermal

f Other (attacn schedule) Ti 16a interest expense on mvestment debts & $? b (1) Investment Income Included on lmes 4a through 41. Schedule K 5 u-’ 5 EL: (2) Investmentexpenses mcluded on Ime 10. Schedule K
17a Type of mcome -_.--.---~~.---..-....--...~.-........~._~....~_~.._.

, l%(2) j 151 I 16a ’16b(l)

I ’ ;
;

-

j
.5 ?! 9

i ’
;

bForelgncountryocU.S.posrcsslon ____ ____ __..._.________ c Total gross IncOme from sources OUtSIde the U S. (attach Schedule)
d Total applrubk deductlons and losses (attach b schedule) 13 Accrued schedule)

__..____________

c Total forergn taxes (check one):
f Reduction VI taxes available schedule) g Other (anacn

q

Pard

for credit (attach

Anach

schedule

for other Items and amounts

not reported

above. See lnstructrons

Assets

.

1 2 3 4 5 6 7

CaSn

Traae notes and accounts recewable
a Mms allowance for bad debts

ln”e?rorIes
iecefal Otrer Or-er Bullatngs a Minus Depletable MoCgage an0 currenr aw ant InvesITews otner accumulated assets aeoletlon amonlzatlon) (amonrzable amonlzarlon only) ary accurrulatea (ner
0:

s;are

gOvernment

ObhgatlOnS

assets

(attach scneaufe)
loans sc-eaule) assets larracn

real estate

a

aepreclable aepreclatton

9 10 11 12

a hl~nus
Lana a Minus inranglole

assets accurulatea

Crlerasse!s

,a:tach scneaule)

13 14

TOTAL assex Liabilities and Capital
Accounts payaole 1 year

15 16 17

pI!c:g1~5 --*es bonas payable m less man ‘ IaDl18tles farracn scneaule) C:-er cJire-: All nonrecourse Mortgages. Otner llabllltles loans notes. bonds payable m 1 year

ia
19 20

or more

(attach scneaule) Paf ners’ caoital accounfs

I
Designation
The followmq Name 01 aeslgnated

I

I

.

of Tax Matters

Partner
return IS fIlea

Rene’ oartner 1snerebv Oestgnatea as the tax matters oafrner (TMP) for the tax year tar wmcn thts DaRnersnlo al
TMP 1

Identltyn&
number of TMP

I

Adaress or aeslgnarea TMP

1 . “1 O..w.lr, -llm 0”I”I 1.11-I.a.1.. IWl.l..

*

SCHEDULE (Form 1065)

K.1

Partner’ s
own*‘ %

Share of Income. Credits,
For caiewar year 1987 or IISC~I year

Deductions,

etc.

i

a~~+0 !545m9

,11C’ DC..“ C na, “ Scncc
Partner’ s Panner

jcolnmc” l.31 mc T~CIIUY

.. .

. ._ . . . . . . . . .._.

:98’ aenQ~ng...,a”

. ..__._.__ _____.___,19---

~--iK

identifying
5 name.

number

t

address. ana ZIP code

! PaRnershlp’ s Identifying number w s 1 Pannersw’ name. adaress. ana ZIP code

A(1)
15 1~

‘ Ye5 Llo

uanner

a general

oanner’

-

D

Enler Partner’ percentage s
Protlt sharmg Loss sharmg

of “~,~&$r~” ‘ _ _ _ _ _ _ _ _ -96 flied return

(II) EM 01 “ tar % 96

If ‘ fes-

to QuestIon

A( 1) ot tnc wtntrsniol tScc 1065 lnrtruct~onr Leave

_ _ _ __ _ _ __% .-_ ._.___
. ..______

(2) Dd th18 Dartner materlallv Dartic~oatem the trade
0, DUSI~~SS actlvltv(w

I
7

pap 12 of the form
(3)

I E
Yes ZNo
‘ F 1 i G(1)
1

OwnersnIp of capital IRS Center where pannersntp
Tax Shelter Did the Reglstratlon Partner’ s ownershtp

_ __ _ _ _ _ -96 w__._.___
b .A._.-.__________ b . .._--~-_-._-_________._ Interest m the partnersh,o C]Yesa

%

blank If no Iraae 0r buslners actlvlllef I D,d thus Dartntr aclwcIv ~arlic~~ate I” Ihe rental iel, estdlc
acl~~~!v(‘ es) Of the Da’ :“erShtD’ (See

-

Numaer

Increase atier Oct. 22. 1986?

jaae 13 01 the Form 1065 Inslroclions teare E!anl II “ re” 0 tdl B Partner
5

No

redI estate actlvllles

)

z

Ye5 C

IJO

(
I

snare or :laoll1tie5

Nonrecourse

s .____.__...........

1’ attach statement. (See page 13 al tnc Form 1065 InstructIons Yes. ) (2) Did the PaRnershlp start or acwre a new actlvlty aeer Oct. 22. !986?. 0 Yesa ho If Yes. attach stat~ent. (See page 14 of the Form 1065 Instructions.)

/
Cautm: Refer to dltacned Panner s ~mVucOonS for Schedule K.1 (Form 1065)

I

I

before enrennq mformatton
(b) Amount

from thts schwufe
lmoun(

I

on y,7ur tax retum, In column(b)
enter the

(a) Dlrtnbutlve

share 8tem

(C) 1040 lllcn

on:

‘ 1
2 3
4 a b c d c

Oralnary Income

mcome (loss) from trade or busmess acflvtty(ies)

.

I

Income or loss from rental PorVol10 mcome (loss). Interest Dwdenas Royalt!es Net ShorVterm Net long.term Orher portfoIl

real estate actwty(les)

or loss from other rental acttwty(les)

Sch. 8. Part I, line 2

.
caprtal gam (loss) capital gam (loss) . mcomc (loss) (attach

.
. . schedule) . . .

. . . . .
.I

Sch. 6. Part II. lme 4 Sch. E. Part I, Ime 5

. .

. Sch. D. line 5. col. (r) or (g) Sch. D. line 12. col. (r) or (g) (Latw 01 w+ubh hna d WC rdu<n:

f 5 6 7
a 9 10

.

Cluaranteed

Payments
under sectton

_ .

.

.

Net gam (iorr)

Other (anach

schedule)

1231 (other than due to casualty or theft)

18x5, ( Scnou*K.L,FWrn 1

!+a ?mauV* lmt-

IOT

(Lala an wobub* Lno d lovr “ fura)

Charitable C011~lkrt~OnS . Expense seduction recovery property(sect~on for Deducttons related to portfolio income . .
Other (attach schedule)

. 179 ) .
. . . . . . . .
.

.

See Form 1040 Instructtons sa hmr , InMMDI* I_ own ( 5uw~L.L (‘ ,065) 1

. I

I1 .2a
b c d
l

Credit for Income tax wlthheld
Lowncome housmg credit .

. . .

@aIdled
actMy Credit(s)

rchabllltatlon

erpendtitures

(attach schedule) related to rental real estate acrtvlty(les)
schedule)

12c (attach

.

.

I3 I

Credqs) related to rental actMy(tes) (attach schedule). Other credits (attach’ schedile) .’ : ^ .

. . 12b and . . . . . . . other than 12b. 12~. and 12d
other than

related to rental . .

. . . . real estate

See Form 1040 Instructions Form 0586. lme 8

sa Puvm’ **lnnrn b l ,ra Lo6))) ( -L-l

:

:

:

:

:

:

:

:

:

:

:
Schedule K-1 (Form 1065) 1967

Property Subject lo Recapture 01 Investment Credit
lnve~lmcnl

Intere,t --_

Tax Preference Items rll~tm loyme

.

SCHEDULE

C

Profit or (Loss) From Business or Profession

(Form

1040)

a
9

10

11
12 13 14 15 16 17
Insurance

c Enter20%of

IlN

26b subvxt

to

d Subtract lme 26~ from 26b

I

Interest 1 Mongage (pad to firuncul
b Other

b Jobs credit
mstltutlons) 29 c Subtrrct l~nc 28b from 284 Other CXDc”seCI (Ilst type

18 19 20 21 22 30 31 32

La~nary ana Ofrlcccxpens4.

ctean~ftg
YW~C”

Legal ana PrOfesslWl

I f I

PensIon A”0 protlt-shrrmg pIA.% Rent on busmess prowrty Ada Amounts m columns

ana rmount) .___,_........._............__._ . . . . ..__........................ . .. ...... ..... .. .... .. ... ..... .. .. . .. . .. . . .. . . .. . .. .. ... . . .. . . .
.I 30 31

I

lor Imes 6 througn 29. These 4rc the total deductions

Net profit or (loss). Subtract lme 30 horn hnc 5. If l Profit. enter here and on Form 1040. line 13. and on khcdule SE. lme 2 (or lme 5 01 Fwm IDO 1 or Form 104 IS) II s 1-s. you MUST go on to lme 32 II you haYe I loss. ycu MUST 4”s~r If ‘ Yes.’ lc41S). thn Ouestw)“: ‘ Do you have Amounts Iof whocn you Arc not you MUST ettacn Form 6198. If ‘ NO.’ enter the IOU on Form 1040. w Form 1040 Instrwtl~ns. lme 13. and on Scheduk

at nskI” thn busmess?’ (Set InstnrNOns ) 0
SE. lme 2 (or lme 5 of Form

Yes 0

NO

1041 or Form

For Peperworb Reductzon Act Notkr.

Schedule C (Form 1045) 1987

jc7eoutcC

lFDlm 1000)

19a7

31..

;

Cost of Goods

Sold and/or

Operations

(See

Scheoule

C lnstructtonsfor

Part

111)

xc)* x159 x75 3889

X32 5257 0273 5299 Xl.4 0430 355 cd71 5885

30:2

Real Estate. Insurance.
finance.
5512 5538 5553

and Related Services

3!!0 3318

::::

; 5710
3079 3045 1736 1751 i777 a839 a854 a870 9396

Manufacturing. Including Printing and Publishing
0612 0638 0653 ,x79 0695 3810 :a36 :a51 3877 3893 .016 LO32 !OS7 :073 1099 :;:5 :313 :339 Isa-

_-

rrantpartation.

Eommuntcations. Public Jtilities. and Related Serwces 5114
5312 5336 5510 5536 5551 5619 a35 5650 3715 3731 3736 :;:: 3921 3939 39% :zG 3966 4119 4317 4333 6676 5692

Servicer (Providing Personal. Professional. and Euriness Services)
7096 7211 7237 7419 7435

1z$G i 9415
I9431

Wholesale Trade-Selling Goods to Other Businesses, Government. or lnrtitutlons. etc 4614 Our*& soM% Indudlng I rnXhlmr% .qulom.nt. rood. nu”h tic
2618 5e1l~ngforywro.m*ccou”t

4416 r432 rr57 4473

7450 7476

7617 7633 76% 7674