928. This chapter examines the topic of questionnaire (form) design and describes the model forms contained in appendix 2.
928. This chapter examines the topic of questionnaire (form) design and describes the model forms contained in appendix 2.
929. There are many approaches to questionnaire design, which may be considered more of an art than a science. Questionnaires should be constructed on the basis of established standards for effective form layouts and for questions, classifications, and instructions that are clear and simply stated.
930. In designing a form, it is essential to consider the reactions of those who may receive it. The immediate reaction of most people depends upon whether or not forms are actually addressed to them. Many people receive a large volume of mail; if they perceive it is not their responsibility to reply, they may divert the form to someone else or allow it to languish unanswered. Subsequent reactions typical of recipients are:
“Do I understand the questions on the form?”
“Can I supply the information without consulting any records?”
“Can I obtain the required information from records readily available to me?”
“Can someone else supply the information fairly readily?”
“Will completing this form be a major task requiring much coordination on my part?”
931. It may be inferred from these questions that levels of effort on the parts of reporters will vary. If forms can be completed and returned with little effort, reporters may do so immediately; this objective underlies the approach in model exploratory form 1. However, form completion usually requires more than a minimum effort, and the next question reporters may ask is: “When will I have time to complete this form?” The amounts and types of work required of reporters will be important determinants in obtaining responses. The collection compiler should, by direct and indirect means, encourage reporters to assign a high priority to form completion.
932. Reporters may also ask:
“Do I approve of this survey?”
“Will this survey supply information of use to me or this enterprise?”
“Does the survey seek sensitive or confidential information?”
“Do I have a legal obligation to comply?”
“What will my supervisor’s attitude be?”
933. A well-designed form will contain inherent answers to many of these questions, ease the burden on reporters, and increase their willingness to cooperate. But form design is only part of the equation. The compiler must also establish and maintain effective contacts with reporters. Paragraphs 967-977 of this chapter note some opportunities for development of cooperative rapport between compilers and reporters.
Page One of the Form
934. The initial page introduces the questionnaire, Therefore, explanatory comments on general issues should precede presentation of more specific topics (questions and instructions). The introductory page also creates an overall impression in the mind of the recipient. The compiler may wish to convey the message that:
“This form is important; note the official logo.”
“We can help if you have any difficulties in filling out the form; we can be contacted by telephone.”
“The information you provide will be treated confidentially; your competitors will not see it.”
“We trust you; your careful estimates are acceptable.”
“We are not officious bureaucrats; however, completion of this form by the due date is compulsory.”
“We appreciate the time and effort you expend in completing this form. By providing information for the compilation of BOP statistics, you are contributing to the well-being of your nation.”
“We have done our best to minimize your costs and inconvenience; for example, we have enclosed a postage-paid envelope for your reply.”
936. Examination of sample page one reveals that the address label is likely to be the most eye-catching feature. This label is usually affixed to preprinted forms. In most surveys of enterprises, the label should contain the name of the enterprise, the address, the name—and possibly the title—of the person responsible for completing the form. The label should also contain a reference number, which is essential for identifying the reporting enterprise. It is important that this information be up-to-date; a reporter’s respect for the compiling organization is lost when details are incorrect. If the information is wrong, the form may even be treated as junk mail and tossed away.
937. Other items at the top of page one—such as the compiling organization’s logo (the symbol of a genuine inquiry); the form number; the compiler’s name, address, and telephone number (which are provided in case the reporter has to contact the compiler); the title of the collection; and the reference period—are all important pieces of information.
938. Items in the middle section of page one state important general information. In most of the model forms, this section begins with the clear direction, “Please read this first.” Items in this section include:
The relevant legislation should be quoted or specifically referenced to show that completing the form is mandatory and to establish the compiler’s credentials.
a statement on confidentiality
This statement should be linked to the one on legislation; the confirmation of confidentiality reassures the reporter as to treatment of sensitive information.
a broad description of the purpose of the collection
This statement should make the reporter aware of the importance of BOP data and of the data required by the questionnaire.
a note on the general layout of the questionnaire
The statement should briefly describe the location of questions and instructions and suggest effective approaches for responding to the questionnaire.
939. The last section on page one requires reporters to identify themselves and thereby take responsibility for completing the form. Information on a reporter’s identity also constitutes a starting point for the compiler to initiate follow-up or query action.
Questions and Instructions
940. The body of a form consists of questions and instructions. Determination of the information to be collected, the instructions to be given, and the presentation of questions and instructions are central concerns of the collection compiler. Often, the compiler must decide what to leave out, as well as what to include. Users of BOP data generally want more information, and reporters naturally prefer to supply less. The compiler must balance these conflicting perspectives by using astute judgement. The compiler should be aware that seeking additional—or more rather than less—information results in a greater reporting burden, slower response, more follow-up, and higher processing costs in terms of staff and computing.
941. To balance these opposing factors, the compiler might: (a) introduce tailored forms and/or special forms for various segments of the population; (b) collect more information from a lesser number of enterprises; (c) collect less information from smaller enterprises than from larger enterprises; and (d) develop supplementary questions and instructions for certain reporters with unusual or more complex transactions. The compiler should concentrate on collecting items that are important analytically and not on collecting classification categories of little importance. For example, it would not be necessary for a compiler to collect detailed data on portfolio investment liabilities if there is very little of that activity in the country concerned.
942. The compiler should avoid seeking information that can be obtained from sources other than the survey unless there is good reason to believe that alternative data would be unsatisfactory.
943. Survey questions should request information in a form that is consistent with the record-keeping practices of the data provider. If information that the data provider can supply easily is not completely consistent with BOP requirements, it is preferable to obtain this information and make adjustments than to require information that is consistent with BOP criteria but difficult or impossible for the data provider to supply.
944. Instructions should be written in language that is plain and unambiguous. BOP terms used by the compiler are not necessarily familiar to reporters.
945. It is often difficult but important to achieve the proper balance between instructions that should be included on a collection form and those that should be supplementary. Typical cases should be covered by instructions included on the form. As lengthy instructions can make forms appear cumbersome and difficult to complete, instructions for less typical cases could be provided in supplementary form and issued on an ad hoc or selective basis, such as correspondence with individual reporters. An option for positioning instructions on forms would be to place a general set of instructions at the beginning of the form and specific instructions with questions to which they relate. This approach may be effective if the specific instructions are not too long. On the other hand, separating instructions from questions may avoid a cluttered presentation. Another option may be to issue separate instruction booklets—an approach that may work well in some circumstances but generally is not favored.
946. There are many options in form presentation. Good rules of thumb are to avoid unnecessary graphics and to follow a consistent layout. The primary objectives are to ensure a straightforward presentation and to obtain accurate information.
947. To determine the most appropriate presentation, the compiler could prepare several versions of forms and test these with reporters. The version that elicits the best results could then be chosen.
948. Collection forms should initially be kept fairly simple; questions should seek basic information. More detailed questions can be added as reporters become familiar with collection methodology. In many countries, successful use of detailed and frequent questionnaires is the result of a long historical process. An approach that gradually increases the amount of data collected may be harder to adopt in an ITRS. Nevertheless, reporters in an ITRS should accept the principle that collections must remain flexible and responsive to changing circumstances.
949. In general, however, changes should be made to questionnaires only when there is a strong case for change. Any revisions should be properly tested before being implemented, and reporters should be notified about changes as far in advance as possible.
Types of Questions
950. A compiler may include several different types of questions on a BOP form.
951. In a single item question, the reporter is asked to supply a single value. For example:
952. “Yes” or “no” questions or multiple choice questions that permit reporters to mark boxes corresponding to answers make form completion and processing relatively easy. For these reasons, such questions are used extensively in the model exploratory form.
953. Self-coding questions, for which reporters enter single characters or codes from a predetermined list in the space provided, also facilitate processing.
954. In a table (matrix) approach, reporters fill in cells in tables. This approach is used often in the model forms. It may combine some features of other types of questions. For example, reporters may be asked to record values in some columns or rows and to record codes in others. If the table approach is used, it is important that rows and columns be labelled for easy reference.
955. The table approach is useful when data being collected are related to each other in some way, such as being components of a total or a reconciliation between stocks, transactions, and other changes. If these relationships are clear to reporters, it is more likely that accurate information will be provided than if a number of separate questions seeking similar information are asked.
956. With a descriptive question, reporters are asked, for example, to describe some activity or to provide the name of a partner country.
957. Some questions may act as trigger questions. That is, if a particular question is answered in a particular way, additional information is sought from the reporter through a supplementary questionnaire or a telephone call. This approach prevents forms from being overloaded with questions and instructions and constitutes an effective means of handling less frequent activity.
958. In addition, there are other devices to make the collection of information easier. A questionnaire may include “of which” questions that seek supplementary answers to previous questions. For example, a question may ask, “Of the amount reported in question 3, how much was in foreign currency?” There may be sequencing questions that have accompanying skip instructions. For example, “If you answered ‘no’ to this question, go to question 5.”
Some Miscellaneous Issues
959. The compiler should determine, in consultation with users, the level of accuracy required. Along with other things, the magnitude in which data are collected will affect the level of accuracy. Generally, data involving money values are collected in thousands or millions of units of currency. Use of higher rather than lower magnitude greatly reduces collection and processing costs. However, the compiler should check completed forms carefully to ensure that the correct magnitude has, in fact, been used.
Revisions to Previously Reported Data
960. Most forms request data for a single collection period. However, a reporter may occasionally wish to advise that previous data have been reported incorrectly. Therefore, most of the model forms contain a section seeking information on significant revisions to reported data. The word significant is used as, in normal circumstances, the compiler can tolerate a margin of error. If information supplied by the reporter is insufficient for processing revisions, the compiler should approach the reporter for additional information.
Population Frame Maintenance and Other Supplementary Questions
961. From time to time, the compiler may add additional questions to a collection form to maintain an up-to-date population frame, to carry out specific inquiries (for example, on financial leasing or securities issued abroad by residents) or to determine how particular transactions are reported. Supplementary questions should be kept fairly simple to avoid prolonging response time and processing of results.
Signing Off and Checking the Form
962. The reporter should sign the form and thereby take responsibility for completing the form correctly and for checking to ensure that information entered on the form is correct. To facilitate the checking process and to emphasize important points to be checked, model forms contain a series of boxes to be marked as checks are undertaken. This feature highlights important parts of the form and provides the reporter with an overview of the task that has been completed. Studies have shown that the inclusion of such checking procedures on forms is viewed favorably by reporters.
963. A colored form stands out from the white paper that accumulates in most offices and may serve as a reminder to the reporter to complete the form. In follow-up action, the compiling staff can refer to the green (or whatever color) form. This color reference helps the reporter to identify the form quickly and thus facilitates follow-up and query action.
964. Compilers in some countries use colored forms with white answer spaces. The white spaces draw the reporter’s attention and make completion of the form more straightforward. The disadvantage with this approach is that it significantly increases printing costs.
965. Forms should be developed to facilitate processing. Forms may include boxes for codes or form layouts may be arranged in special ways. Form design should involve consultation with those who will be responsible for processing completed forms.
Collections Via Computer Medium
966. Compilers in some countries collect data by means of computer media, including computer tape, floppy disk, and on-line (modem) reporting. Computer media are particularly useful for large volumes of data. For more simple surveys, the paper collection form is often preferred, although this preference may change in the years ahead. Many requirements—such as clear instructions and comprehensive testing—for effective design of paper forms are also relevant to collections made by computer media.
Steps in Questionnaire Design
967. The first step in questionnaire design is to define, in consultation with users, broad requirements for data. Questions that should be asked include: What data items are required? What classifications are required? What are the priority areas? How frequently and quickly are data required? (This consultation process is described in more detail in chapter 18, paragraphs 839–840.)
968. Step two consists of initial consultations with potential reporters of data. The compiler should ascertain, in the context of user requirements, the amount of information available and the nature of record-keeping practices of potential reporters. Draft forms should not be produced prior to this stage for two reasons. Confusing or incorrect assumptions could be made about availability of data and, as enterprises often place low priority on form completion, there may be an attempt to restrict these initial discussions to lower level staff if the compiler presents forms at this time. In initial discussions, it is essential to obtain a broader perspective and support from higher level staff. Also, initial discussions should include a selection of reporters from large and small enterprises.
969. The third step consists of informing and seeking the assistance of the compiling agency’s technical and support staff. In this phase, the most appropriate collection methodology options are formulated. Discussions with technical staff provide insights, highlight pitfalls, and facilitate the articulation of options.
970. Step four consists of preparing a proposal that outlines various options, projects option costs, and sets out a development timetable. The proposal should be submitted to the management of the compiling agency—and possibly to users—for a decision.
971. Step five is the actual design phase, which is the subject of paragraphs 934–966 of this chapter.
972. Step six is the phase in which the questionnaire is extensively tested; test results are used to modify the form. This process may involve distribution of draft questionnaires to potential reporters and subsequent personal interviews of reporters from large enterprises. A few reporters from smaller enterprises should be interviewed in person, and many should be interviewed by telephone. Form testing should be implemented in successive stages, and comprehensive testing should be undertaken only after draft forms have been tested on a representative sample of reporters. Short-cuts at the testing phase often lead to poor results later on. Form testing should reveal whether the presentation and content of the form are clear and whether data to complete the form are available. Form testing should also identify issues or problems causing concern for particular enterprises. Attempts should be made to have actual reporters complete forms during the test phase.
973. Compilers in some countries use observation studies during the form testing stage. In these studies, the BOP compiler delivers a draft form to potential reporters and observes their reactions to the form. The compiler notes whether parts of the form appear to bewilder reporters, whether some parts are ignored, and the order in which reporters respond to the parts of the form. These studies are most effective when reporters actually complete the form in the presence of the compiler. If this option is not available, observation studies may still be useful for determining general reactions of potential reporters. Obviously, studies of this type can be quite expensive and, for this reason, should be restricted to a small number of potential reporters.
974. Step seven involves designing and testing office and computer processing systems, which are discussed in other chapters of this Guide.
975. Step eight consists of reporting to management on results of the work in progress. Required modifications to original specifications are outlined and, if necessary, further direction is sought.
976. Step nine is implementation. No matter how much testing is undertaken, actual implementation may reveal problems that require further form modification. In the implementation phase, many or all of the significant reporters should be visited to discover how they completed the form and what problems they encountered. A selection of less significant reporters should also be visited or contacted by telephone. Such contacts contribute to the education of reporters and identify any further weaknesses in form design. Implementation may be approached in steps, and priority may be given to obtaining certain output ahead of other output. For example, if the compiler introduces a new collection form for external assets and liabilities, priority may be given to validating and publishing aggregate financial flows and income data. Validating and publishing more detailed classifications (for example, instrument of investment and partner country) and stock position data may be undertaken at a later time.
977. Step ten is the evaluation. An evaluation report should be submitted to the compiler’s management (and possibly to BOP users) to inform them of any remaining weaknesses and any further modifications required. Evaluation may be staggered to parallel implementation steps discussed in the previous paragraph and/or consist of a preliminary evaluation and a subsequent detailed evaluation. It may be best to conduct and present the detailed evaluation after procedures have become fully established. Evaluations should be carried out periodically to overcome the law of atrophy that applies to all statistical collections.
Model BOP Forms
978. A set of model BOP forms is provided in appendix 2 of this Guide. Form 1 is an exploratory form, and form 2 is an enterprise register form. Forms 3 through 5 are ITRS forms. Forms 6 through 13 are enterprise survey forms. Form 14 is for embassies and international institutions located in the compiling country. Form 15 is a household survey form for travelers. The remainder of this chapter describes the essential features of these forms.
979. For a number of reasons, compilers are not expected to adopt any of the model forms per se. The purpose of a model form is to illustrate one means of collecting data and thereby provide compilers with a starting point. Model forms are designed to demonstrate potential capture of most classifications required by the BPM and the SNA. For many countries, some of these classifications may not be important. In these cases, model forms could be simplified for actual use. Model enterprise survey forms include a variety of topics and cover large segments of the population. Under certain circumstances, compilers may wish to create several form types to take advantage of population segmentation, or compilers may wish to combine topics covered by several model forms into a single form. For particular countries, it may be necessary to add questions to the model form. Also, the form design process (outlined in paragraphs 967–977) requires that compilers hold discussions with users and enterprises prior to development of draft forms.
980. Although there may be some variation in layout from form to form, most have an introductory page that includes some or all the features mentioned in paragraphs 934–939 of this chapter. In addition, most forms have an office use only section. This section is for individuals to verify that certain processes have been completed. Form processing staff record the date the form is received (in the box marked rec.), check the form for obvious errors, and mark the box labeled edit. The supervisor may also check the initial validation and record his or her verification (in the box marked check). Such verifications encourage processing staff to take responsibility for their work, provide a record of the processing stage that forms have reached, and establish the processing trail should some error occur or procedure fail. It is necessary to identify problems so that corrective measures, such as training and procedural changes, may be undertaken.
981. When possible, similar approaches and classifications have been used throughout the model forms. As form design standards imply, questions and instructions should be consistent from form to form, unless circumstances dictate some modification. Also, the approaches and, therefore, the forms for an ITRS and for ES will be similar in many ways. For example, both types of collections may approach the same types of institutions, and the output (BOP data items and classifications) of each collection is the same. Each form usually concludes with a series of items to be read and marked by the reporter and a place for him or her to “sign” off, as described in paragraph 962 of this chapter.
982. Aspects of many model forms are discussed in chapters 3, 4, 5, 6, and 9, which describe some of the collections that could be used to meet BOP requirements. The following discussion avoids, as far as possible, duplication of discussions in previous chapters and concentrates on the strategy chosen for each form. Appropriate cross references are provided.
Model Exploratory Form 1
983. The exploratory form is designed to identify enterprises (within an enterprise group) that:
are direct investment enterprises or direct investors;
engage in international trade in goods and services and, if so, the type and size of this activity;
employ foreign workers and, if so, the approximate wages and salaries paid to them;
have external financial assets and liabilities, and the type and size of this activity.
Also, the exploratory form requests information on the enterprise group structure.
984. The purpose of the exploratory form is to identify the population to be included in BOP enterprise surveys. However, the approach could be readily modified
to identify enterprises that (a) may have accounts with nonresident banks and other external assets and liabilities and (b) could be included in an ITRS collection.
985. The exploratory form makes extensive use of questions with boxes to be marked in response. When values are requested, only approximations are required. The intent of this approach is to obtain quick responses. Most reporters should be able to complete this form with little or no reference to their records. Enterprise profiles developed from exploratory forms should be sufficient for designing a BOP enterprise register and for selecting participants for a sample (or partial coverage) survey. The results of an exploratory survey could also be tabulated and used, if necessary, to estimate data for enterprises falling below collection thresholds or to provide benchmark estimates in sample estimation procedures. The compiler could simplify the form by omitting unnecessary questions. For example, if the compiler does not require information on importing and exporting activity, question 6 could be omitted. Also, questions 9B, 11B, and 12 may not be required or may be truncated. A shorter questionnaire usually elicits more timely responses.
986. Another advantage of this form is that it initiates the reporter in completing a BOP collection form. As completing the exploratory form should be a relatively easy exercise, it may make the reporter more receptive to completing more complex BOP forms.
987. Questions 1 and 2 ask whether the enterprise to which the form is addressed is a part of a larger enterprise group that may already be on the BOP enterprise register. If the enterprise is a lower level entity, such as a subsidiary, it is merely necessary for the enterprise to name, in response to question 12, the ultimate parent enterprise in the group and to return the form. On the other hand, if the enterprise has subsidiary enterprises in the reporting economy, the questionnaire should be completed for the group of enterprises as a whole.
988. A question about the level of future activity could be added to the exploratory form. Such a question serves a useful purpose when new enterprises are being established and the level of recent activity may not be a good indicator of future activity. Illustration 19.2 contains a sample question of this type.
989. For certain areas of BOP activity (particularly international transportation and travel), more specific exploratory questionnaires could be developed.
Model BOP Enterprise Register Form 2
990. Form 2 is an office form, not a collection form, and it is included to illustrate the type of data that may be stored for units on the BOP enterprise register. Instructions for completion accompany the form. Basic data to complete model form 2 would come from a completed form 1 or a similar form. Model form 2 may be more detailed than necessary for particular countries. Whatever is not required could be omitted, or additional questions could be inserted.
Model ITRS Forms
991. Additional information on the model ITRS forms is provided in chapter 3, paragraphs 117–120.
Model Form 3—Payments
992. Form 3P is designed to capture bank client payments in foreign currency and payments to nonresidents in domestic currency. BOP transactors should complete form 3P for all such transactions above certain thresholds. (Excepted are transactions undertaken through enterprise accounts with nonresident banks or foreign currency accounts held with domestic banks. These transactions are measured via model ITRS form 5.) A corresponding form (form 3R) would be necessary for receipts. While not included in appendix 2, form 3R would be similar in structure to model form 3P.
993. Instructions for completing form 3P are presented on page two of the form, but classifications and codes that the bank client must refer to are included in the model ITRS Form 3C—Classifications. In a number of European countries, the form providing banks with instructions for making payments is comprised of an original and several carbon copies. Instructions and classifications can then be printed on the backs of original forms and carbon copies rather than on separate pages.
994. Model form 3M has been provided to illustrate how supplementary data on payments for imports of goods may be collected. A model of the corresponding form 3X for export receipts is not presented in appendix 2. Such forms may be necessary if an ITRS is used to compile the goods item. Special supplementary forms may also be required to obtain details on items such as goods for processing.
995. Forms in the “3” category are designed on the premise that reporters can record and code data on the form—a realistic expectation. As can be seen from general notes to the form, a high threshold is used to reduce the reporting burden; however, there would be a tendency for certain reporters to complete the form repeatedly.
Model Form 4—Banks
996. Two model bank collection forms are included among the model ITRS forms in appendix 2. Instructions accompanying the forms describe the overall collection process (especially the procedures for handling the ITRS form 3 series) and present instructions for completing ITRS bank forms 4A and 4B. Some of the forms referred to (such as forms 3PS and 3RS for the sampling of small transactions) are not included in the set of model forms, but the description given in form 4 should enable the compiler to produce them.
997. Form 4A collects:
reference information (part A);
transactions of a bank on its own account (parts B and C);
a summary of all the transactions passing through the bank’s accounts (part D)186;
a reconciliation of these transactions with the bank’s nostro and vostro account balances (parts E and F);
the bank’s external asset and liability positions (part G).
Form 4B requests a record of all bank-related foreign currency transactions and all transactions with nonresidents. Such a form is useful to ensure complete reporting of these transactions.
998. Collection instructions for forms 4A and 4B could be treated as a separate, permanent document. These instructions are somewhat longer than those on the other model forms. However, more detailed instructions are appropriate as banks play a major role in an ITRS collection, and staff in each bank should develop in-depth knowledge of the complete collection procedure.
Model Form 5—Enterprises
999. This form is for enterprises that have accounts with nonresident banks, foreign currency accounts with domestic banks, or other external assets and liabilities. Thresholds could be used to exclude small contributors.
1000. Form 5 collects:
reference information (part A);
details of foreign payments and receipts (part B);
a reconciliation of payments and receipts with bank accounts (Part C);
a check on the exchange rates used in conversion (part D);
information on other external assets and liabilities (part E);
supplementary data on trade transactions (part F), which would only be required when an ITRS is the source for the goods items.
Separate parts B and C would be completed each month for each bank account.
Other ITRS Collection Forms
1001. Previously described model forms represent the principal ones that compilers may have to design. However, it may be necessary to supplement these forms with more specialized forms, such as those for enterprises involved in international transportation and travel or those for measuring reinvested earnings attributable to direct investment. Space prohibits the inclusion of such specialized forms in the Guide, but the reader should find that many issues relating to the collection of such data are addressed in enterprise survey forms included in the model set. Another specialized form could be required for financial intermediaries involved in certain security transactions. Model form 13 (which is discussed subsequently) or a similar form could be used for this purpose.
Model Enterprise Survey Forms
1002. The model enterprise survey forms presented in this Guide cover a range of BOP activities. The compiler may wish to combine some of these forms or rearrange the grouping of questions to address a particular collection strategy. The forms are intended to be illustrative rather than definitive.
Model Form 6—Goods
1003. Form 6 begins with the usual introductory page, which is followed by two pages of instructions and questions in eleven parts. The form concludes with a standard question on significant revisions to past data and a series of questions pertaining to checking the form.
1004. Form 6 presents a number of collection possibilities, and compilers are not likely to use this form in its entirety. As most countries rely on other sources to collect data on goods, few countries would use a survey to collect the data requested in parts A and B. Also, it is unlikely that a compiler would collect both selective data on goods (parts H and I) and across-the-board data on goods (parts A and B) in the same survey. In addition, enterprises involved in processing, repair, and merchanting activities are usually specialized, and it may be more appropriate to use specialized forms for questions and instructions relating to these activities.
1005. In parts A and B of form 6, two approaches are used to record a commodity breakdown. The compiler should decide upon the most appropriate approach. More accurate data may be obtained for exports (and imports) if reporters are asked to describe the commodity and to provide information on quantities involved. Parts H through K are included to illustrate collection of selective data to use as the basis for timing adjustments to ITS or certain BOP projections. Further information on form 6 is provided in chapter 4, paragraphs 133-146.
Model Forms 7 and 8—Transportation Services
1006. Form 7 is designed to collect information from resident transport operators, and form 8 collects information from resident entities that provide various goods and services to nonresident transport operators or that receive selected services, such as inland freight and mail, from them.
1007. Identifying enterprises that fall within the scope of these forms is often a difficult task. While the exploratory form (form 1) collects broad information on international transportation activity, it may be desirable for the compiler to develop a more detailed exploratory form to identify enterprises with transactions covered by the BOP transportation item. Such an exploratory survey should collect information on whether enterprises are involved as (a) operators, (b) suppliers of goods and services to operators, or (c) acquirers of transportation services from nonresident operators. Such an exploratory form should also collect some general data on the size of the activities involved.
1008. One issue that must be resolved is whether to collect data on a transactor or settlement basis. In part A of form 8, for example, data are collected on a settlement basis. That is, the resident enterprise that settles the claim reports the information rather than the resident enterprise that provides the service. The case for employing one approach rather than the other may vary from item to item. Most information on the airline industry is available from the branch office of a nonresident airline operator, so the settlement basis would be more appropriate. For other industries, the choice is less clear. Therefore, the compiler should investigate the issues before finalizing a collection strategy.
1009. Further information on forms 7 and 8 is provided in chapter 5, paragraphs 212–217.
Model Form 9—International Travel
1010. As mentioned in chapter 4, paragraph 152, ES of travel businesses may measure instruments used to settle travel expenditure or services provided to nonresident travelers. The former methodology can measure both travel credits and debits while the latter measures travel credits only. Form 9 covers both approaches. Parts A, B, and C measure various instruments (such as travelers’ checks, credit and debit cards, and prepaid and advance tour purchases) used to finance travel. Parts D and E measure goods and services supplied to foreign travelers by hotels and other enterprises. In practice, the compiler is unlikely to include all these approaches on a single form. Therefore, the actual form would be relatively simple.
1011. The compiler may choose to measure transactions on the basis of those undertaking the transactions or those settling them. In form 9, the enterprise that settles the transactions has been chosen as the reporter.
1012. Chapter 4, paragraphs 152–160 provide further information on form 9.
Model Form 10—International Services
1013. Form 10 measures selected service transactions between residents and nonresidents. Two matters associated with the collection of data on services require the compiler’s attention. The first concerns determination of the borderline between the activities of a branch and its head office. A branch of an enterprise operating abroad is not regarded as providing international services to residents of the country in which the branch is located. Rather, the branch is regarded as a direct investment enterprise and its impact on the BOP is reflected as part of financial and income flows from direct investment. The second matter is to ensure that intra-group services (such as management fees paid by related companies to each other) that cross international borders are reported. Both of these issues should be focused on during form testing and implementation phases.
1014. Parts C, D, and E of form 10 show the data ideally required to measure international insurance transactions. Some amalgamation of categories may be possible in practice. In particular, transactions in insurance of goods and other casualty insurance could be combined as BOP treatments of these types of insurance are similar.
1015. Paragraphs 161–166 of chapter 4 provide more information on insurance questions contained in form 10, while paragraphs 167–168 of chapter 4 provide more information on the remainder of the form.
Model Form 11—Compensation Payable to Foreign Workers
1016. Form 11 illustrates the type of data on employee compensation and related transactions that may be collected in a survey of employers. Some items on the form require the employer to provide estimates of employee expenditure and taxes in the host economy and of repatriations abroad. This type of data may be readily available from some employers in some countries. In other situations, the compiler may prefer not to ask employers for such data.
Model Form 12—External Assets and Liabilities
1017. Model form 12, which is presented in appendix 2, is comprehensive; it provides modules that could be used to develop a collection or collections of data on external assets and liabilities.
1018. Model form 12 includes the usual introductory page. The instructions follow on consecutive pages. For actual forms, instructions could be presented on odd-numbered pages, issued separately, or handled in other ways. Paragraph 945 of this chapter should be consulted for further information on this point.
1019. In model form 12, questions are divided into seven parts. Part A (which requests data about enterprise claims on nonresidents) contains separate columns for opening and closing positions, gross and net transactions, exchange rate and other changes, and investment income. Three categories of claims are shown: claims on direct investment enterprises, claims on direct investors, and claims on other nonresidents. Each category is further subdivided by instrument of investment.187 Terminology is explained in the instructions. Part B requires the reporter to classify key items from part A by partner country. Parts C and D seek similar information for liabilities to nonresidents.
1020. Part E seeks information on financial fees and withholding taxes. Part F contains supplementary questions on the valuation of direct investment equity. Parts G and H seek information on reinvested earnings of direct investment enterprises. Paragraphs 180–186 of chapter 4 describe form 12 in more detail.
1021. Form 12 is designed on the assumption that the reporting entity is both a direct investor and a direct investment enterprise—an atypical circumstance. The form could be modified for other types of enterprises. When an enterprise is a direct investor but not a direct investment enterprise, all instructions and questions relating to direct investors abroad may be omitted. When an enterprise is a direct investment enterprise but not a direct investor, all references to direct investment enterprises located abroad may be omitted. When an enterprise is neither a direct investor nor a direct investment enterprise, all references to both direct investors and direct investment enterprises may be deleted.
1022. Further simplifications may be made. For example, a country may receive significant investment in the form of direct investment and loans from nonresidents but otherwise receive no external financing and have little investment abroad. The compiler has few resources to mount a major collection effort and, consequently, wishes to keep collection and processing costs to a minimum. Illustration 19.3 shows how parts A and C of model form 12 may be arranged in a simplified format that would suit a data collection for such a country. While information on reinvested earnings attributable to direct investors should still be collected in the detail shown in part H of model form 12, part G of this form could be reduced to a single item because reinvested earnings from abroad are likely to be very small.
1023. Another form could be developed for enterprises that are not direct investment enterprises by omitting rows 1 through 3, row 6, and item 2B in question 9.
1024. On the other hand, when more detailed information is required, a different approach could be used. The first column in part A (instrument of investment) could be replaced with three columns that would permit reporter coding of (a) instrument of investment, (b) claim or liability, and (c) country of nonresident liability or claim holder. The reporting enterprise must be instructed to complete a row for each instrument, claim/liability, and country combination. Parts B through D would be deleted. This approach is similar to that used in many of the ITRS forms.
Model Form 13—Security Transactions
1025. Chapter 6 describes data that possibly should be collected from resident intermediaries in respect of securities issued both in the domestic market and abroad. As stated in that chapter, data on transactions and associated stock positions handled by resident intermediaries cannot, in many cases, be collected from the resident principals involved. Model form 13 is designed to measure, as part of an ITRS or ES, security transactions and associated stock positions handled by resident intermediaries.
1026. Many approaches are possible, and the model form 13 presented in the Guide represents only one. The form is designed to collect details on a security-holding basis from resident intermediaries. For securities issued by residents and owned by nonresidents, the form collects data on the security reference number and owner code, as well as details on financial transactions, stock positions, income, fees, and withholding taxes. The security reference number enables the compiler to access a security database, maintained by the compiler, that contains for each security: (a) the name of issuing enterprise; (b) the type of instrument;188 (c) an institutional sector and industry code for the issuer; (d) the currency of denomination of the security; and (e) the country of issue. The owner code enables the compiler to ascertain the country of residence and possibly the institutional sector of the holder. For securities issued by nonresidents and owned by residents, similar information is collected.189 In these cases, the security reference number identifies characteristics such as type of security, country of residence, institutional sector of issuer, currency of denomination, and country of issue.
1027. Model form 13 is based on the assumption that all data can be collected from one institution. As discussed in chapter 6, in some countries, the roles of security broker and portfolio manager are combined; in others, the roles are split. Therefore, the form may have to be split so that, for example, portfolio managers provide details of positions, income payments, and certain fees and taxes, while brokers are approached for purchases and sales, issues and redemptions, and associated fees. Even if the collection is fragmented, a full accounting of each security could be assembled from the different data sources. Security reference numbers could be used as the link. Sometimes, the compiler may be able to collect only partial data—for example, stocks but not flows and income or vice versa. Using methods described in chapters 14 (income) and 16 (stocks and transactions), the compiler can derive uncollected items from data that has been collected.
1028. The collection form is also designed on the assumption that computer database facilities can be used to assemble and supply relevant information to the compiler. As some compilers have found, if the necessary facilities are available and intermediaries can supply relevant data, the compiler’s processing costs are relatively low when such an approach is used. For intermediaries to operate efficiently in the market, they usually require database systems in which security reference code information and client account details are essential elements. The compiler should ensure that appropriate legislation exists to undertake the collection, that data are kept confidential, and that staff and computing resources to handle the work are available.
1029. For compilers in some countries, model form 13 may seem ambitious, or the compiler may not have the necessary authority to collect all the data required. In these cases, the compiler could request the intermediary to prepare tabulations that the compiler might otherwise compile. At a minimum, the compiler should attempt to obtain the items in columns C through L in parts A and B of the form. These items should be classified by: (a) sector of the issuer and country of residence of the holder—for securities issued by residents and (b) country of residence of the issuer and sector of the holder—for securities issued by nonresidents and held by residents.
Other Model Forms
Model Form 14—Embassies and International Institutions
1030. Model form 14 illustrates the type of data that could be collected from foreign embassies located in the compiling country. These data could include numbers of local and nonresident staff employed, expenditure on goods and services, compensation paid to locally employed workers, and details of official transfers and loans. Should a foreign government maintain a military base in the compiling country, the form could be modified to collect such detail. Likewise, form 14 could be modified to collect data from international institutions located in the compiling economy. More information on model form 14 can be found in chapter 9, paragraphs 380–383.
Model Form 15—Travel
1031. Form 15 is designed to collect, from nonresidents, information on travel and related expenditure in the compiling country. The form is designed to identify types of goods and services acquired by travelers rather than the means by which expenditures were made. Form 15 also seeks information on income earned by these persons while they were in the compiling country; this data could be used to measure part of compensation of employee debits. The form should be completed by travelers shortly before they depart from the compiling economy. Alternatively, questions on the form could be asked in interviews with these persons.
1032. A similar form could be developed for residents returning from travel abroad. Paragraphs 315–324 of chapter 7 provide further information on conducting surveys of travelers.