0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

Costa Rica’s national accounts are compiled and disseminated by the BCCR, based on Organic Law 7558 of the Central Bank dated November 3, 1995 and subsequent amendments. Although that law does not stipulate that the national accounts must be compiled by the BCCR, Article 14 (d) establishes that each month the BCCR must publish data on Costa Rica’s economic situation, with information on at least production, prices, credit, exports, imports, and gross and net international reserves. The national accounts are compiled at the BCCR by the Department of Social Accounting (DCS), a unit within its Economic Division.

In accordance with the terms and conditions of the National Statistical System (SEN) Law 7839 of November 4, 1998, establishing INEC and giving it the role of directing the SEN, the national accounts are part of the national statistics that INEC must prepare. The original version of that law provided that the national accounts would be transferred within a year from the BCCR to INEC, i.e., in 1999. Given that INEC had not managed by that time to consolidate its organizational structure and was also conducting the Population and Housing Census, this transfer was postponed until January 2003. Nonetheless, pursuant to the law, the BCCR would continue for five years, starting in 2003, to finance the national accounts function in INEC. This task would then pass to the government budget, at a current cost of US$1.6 million per year.

Although the SEN law establishes that INEC has authority to designate by regulation the institutions that will make up the SEN, these institutions have not been designated as yet. Thus, the agencies that will make up the SEN and their responsibilities in compiling statistics have yet to be defined. In view of this situation, the DCS has made a proposal regarding the statistical units that have an obvious function as compilers of economic data and should thus be incorporated in the SEN, beginning with the BCCR itself.

In practice, the BCCR has taken on the principal activities involved in gathering basic statistics given that, due to limited resources, INEC—formerly the Directorate General of Statistics and Census—has only made a limited contribution to preparing the basic statistics needed to compile the national accounts.

0.1.2 Data sharing and coordination among data producing agencies are adequate

The principal sources of basic statistics or primary data for compiling Costa Rica’s national accounts are censuses and economic surveys conducted by INEC, the manufacturing sector survey conducted by the Institute for Economic Research of the University of Costa Rica (IICE), and administrative records such as those of the Costa Rican Social Security Fund (CCSS), the Directorate General of Direct Taxes, and the Directorate General of Customs.

Although the acquisition of specific information is formally based on contracts (e.g., on the economic surveys for base year 1991) and transfers (as in the case of the 2001 household survey), in practice there are no formal permanent mechanisms established for coordinating the provision of data. There are even duplications in the preparation of some statistics as in the case of foreign trade data, which is handled by both INEC and the BCCR. Access to administrative records is not governed by formal agreements and, although in practice there have generally been no problems involving access for the compilers of the national accounts, there is a trend towards deterioration evidenced by the difficulties encountered in obtaining data from the tax records and other public sources such as the Ministry of Agriculture and Livestock.

In the case of the Directorate General of Direct Taxes, the information obtained is partial due to the confidentiality conditions imposed by the reformed Tax Procedures Code of 1995, which barred access to private company financial statement data from that source. More recently, processes to simplify tax procedures have reduced the amount of information that the Tax Administration requests of taxpayers; thus the ability to obtain information from this source has been further eroded.

As there are no permanent formal mechanisms for obtaining data from public sources, the data are supplied with a time lag. This problem limits the ability to reduce the delivery time for national accounts data. In addition, current legal provisions prohibit requiring the private sector to supply information, as this is contrary to constitutional precepts. The BCCR is currently making efforts to submit a bill that would make it mandatory for the private sector to submit information, based on jurisprudence from a resolution of the Constitutional Court ruling on the Consumer Protection Law. In addition, preliminary approval in the current legislature of the draft Law on Financial Administration and Public Budgets represents an important step toward establishing norms to support data collection for the national accounts, as the language of that draft law specifically provides that the information systems of government institutions must consider the need for national accounts data.

Finally, the Foreign Trade Promotion Enterprise (PROCOMER) has been formally commissioned to supply data on exported goods based on single form records. However, staffing problems in that agency sometimes affects the timeliness of these data. Similar problems also occur in the Directorate General of Customs.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

According to Article 40 of the BCCR’s Organic Law:

“Government offices and units and autonomous institutions shall be required to provide assistance to the departments of the Central Bank, so that they can carry out their functions efficiently. For this purpose, they must promptly supply them with the data, reports, and studies they request.

Failure to meet this obligation on the part of responsible officials in government offices and agencies and autonomous institutions shall be considered a serious dereliction of duty.

Employees of the Central Bank of Costa Rica have access to tax information for statistical purposes only. They must respect the prohibitions and limitations established in Article 117 of the Code of Tax Standards and Procedures, Law 4755 of May 3, 1991. They shall also be subject to the provisions of Article 203 of the Penal Code.”

While that law does not specify the private sector’s obligation to provide information to the BCCR for statistical purposes, requests for information from the Social Accounting Department for national accounts compilation always indicate the confidential nature of the data to be provided by the sources. Based on the opinion of the BCCR’s Legal Counsel, this fact alone assures the confidentiality of the information gathered. In addition, Article 37 (g) of Chapter IX of the Internal Labor Regulations of the BCCR indicates that “Employees must maintain the utmost secrecy regarding the operations, business, and actions of the bank and its clients,” and this is supplemented by Chapter XII on the sanctions applicable for failure to comply with this standard.

In addition, the law establishing the SEN and that would in future support the national accounts compilation function includes sanctions in this area. In effect, Article 4 of Law 7839 establishes that all agencies belonging to the SEN are required to compile and process statistical data in accordance with principles of statistical confidentiality. Moreover, Chapter IV, Article 51 of that law refers to the system of sanctions applicable to public officials who violate their duty to maintain the confidentiality of statistics.

Computer programs have devices that help to protect confidential data. In effect, the processing of information is protected by personal access codes held by authorized personnel only.

In addition, measures are adopted to protect information when it is destroyed, an action that is subject to the National Archives Law. The Internal Audit Office of the BCCR monitors compliance with these provisions. Data storage facilities have security measures for data protection.

0.1.4 Data reporting is ensured by legal mandate and/or measures implemented to encourage voluntary reporting

As indicated in the above item, existing legal provisions require the reporting of information useful for national accounts that are available in public sector administrative records, except in cases where such data are expressly limited by confidentiality regulations. With respect to the private sector, there are no legal provisions or norms regarding the compilation of data and the provision of information. Given that providing data in this case is largely voluntary, considerable efforts are made to build a good relationship with data providers in order to obtain information. In many cases, they are offered statistical information produced by the BCCR as a service in exchange for the information requested.

0.2.1 Staff, financial, and computing resources are commensurate with institutional programs

The DCS staff (26) assigned to compile national accounts is not of adequate size because, in addition to compiling activities proper to the national accounts function, staff members must carry out basic information gathering activities not undertaken by INEC or other institutions due to limited resources and changes made in the scope of the functions assigned to the various public entities.

Efforts are being made so that staff quality will be adequate and include experience with and knowledge of concepts and practices. These efforts include the contracting of international consulting services since 1997 to develop the 1993 SNA project In addition, there is ongoing and informal transfer of knowledge from more experienced staff to new staff, and officials are periodically sent to overseas national accounts courses and seminars. Finally, the BCCR has planned a training course for 2002 through the direct contracting of international consulting services.

Recently, the incentives structure and work environment suitable for retaining qualified personnel have deteriorated, primarily owing to uncertainty as to whether the national accounts function would remain in the BCCR. The departure of experienced staff has led to the hiring of staff with less experience with the function, and this has challenged the education and technical training processes mentioned above.

The national accounts program has financial resources that are appropriate and certainly proportionally higher than the resources that other statistical entities usually have. In this context, the problem lies more in the ability to absorb the financial resources productively. Even entrusting the conduct of statistical projects to third parties has not yielded good results, owing in great part to the lack of cooperation from informants. This has been the case, for example, with the economic surveys for base year 1991. This reaffirms the need to establish legal and administrative instruments that include a requirement to provide information to the SEN entities and provide sanctions on those who fail to do so.

The computer resources for automating the process of compiling and disseminating the national accounts are sufficient. The BCCR’s Computer Department consistently upgrades to the latest generation of computers. In addition, efforts are being made to take advantage of the possibilities offered by modern computer technology, particularly by developing initiatives—to materialize in 2002—that will make it possible to move from the current linked spreadsheet system to compiling processes that benefit from database platforms.

0.2.2 Measures to ensure efficient use of resources are implemented

The BCCR has implemented a standardized cost control system by homogeneous products that makes it possible to measure the relative cost of the national accounts program compared to other statistical compilation processes of the BCCR. In addition, the restructuring of the Social Accounting Department in 1996 facilitated the adaptation of the workloads of the various sections that compile the national accounts and the more efficient integration of the processes generating these statistics.

0.3.1 Processes are in place to focus on quality, to monitor the quality of the collection, processing, and dissemination of statistics, to acknowledge and deal with tradeoffs within quality, and to guide planning for existing and emerging needs

The BCCR values and broadly supports efforts designed to improve quality in statistical work as a prerequisite for promoting transparency and strengthening confidence in government policies. However, these principles may not necessarily be applied in units outside the BCCR that compile the basic statistics for the national accounts function.

The law establishing the SEN granted its guiding body, INEC, supervisory powers over the statistical processes carried out by government institutions. In this respect, the SEN has the obligation to ensure the application of a standardized system of concepts, definitions, classifications, nomenclatures, and codes that will make it possible to compare data, evaluate the quality of SEN statistics, and promote research, development, improvement, and application of the statistical methodology in the agencies that generate basic or summary data, as well as to support and provide technical assistance to the statistical services of the State and other users through mutual cooperation agreements.

Nonetheless, such goals have not been fulfilled in practice. Users—both the public and the authorities—have been continuously if informally concerned about the quality of statistics.

Mechanisms designed to consider current and emerging needs for statistical data are informal. In some cases proposals are taken up by the national accounts unit itself, for example, the tourism satellite account; in other cases they are directed to INEC. The DCS is always on the lookout for initiatives to expand and strengthen statistics. In this respect, it has submitted to INEC a plan to improve the basic statistics for the national accounts, a plan that includes both immediate actions and strategic actions involving the future change of the base year for the national accounts.

1.1.1 Statistics are compiled on an impartial basis

Only technical considerations are involved in compiling national accounts. Professionals are fully independent in choosing the sources and statistical methods and techniques they consider most appropriate for compiling the estimates. Results are not subject to any external or internal influence.

Legal provisions do not explicitly stipulate the professional independence of the staff responsible for compiling the national accounts. However, professional criteria guide the selection of management staff. Although there is no provision to this effect, in practice there is no undue outside influence on the content or dissemination of statistical data.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

All decisions regarding data sources are made in the DCS and agreed with the Economics Division. This means that compilers of statistics are free to select the basic data in terms of quality, timeliness, cost, and the response burden on informants.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The Social Accounting Department is authorized to comment on erroneous interpretations or misuse of statistics. The practice is to hold quarterly meetings with the press for such purposes. Public disclosure of economic results, particularly projections of economic activity, is a topic reserved to the Office of the President of the BCCR and is carried out through the public relations unit of the issuing institute. Subsequent clarifications for users with respect to published data are in practice made by the DCS.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

As indicated, the terms and conditions under which the national accounts are compiled are explicitly stated in Article 14 (d) of Law 7558. These provisions are available on the BCCR Internet site and in printed form.

Methodologies and quarterly estimates of the national accounts and the Indicador Mensual de Actividad Económica (IMAE) have already been made public on the Internet site; information on the annual accounts will soon be made available. In addition, all the methodologies are physically available at the BCCR Information Center.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

No governmental authority has access to data before they are published. National accounts are disseminated to all users at the same time.

1.2.3 Products of statistical agencies/units are clearly identified as such

Data released to the public are clearly identified as products of the Social Accounting Department. The BCCR indicates in its publications the source of the information, whether in the case of its own products or when it reproduces data from other statistical agencies. The source of information is usually mentioned in a footnote.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

There is no advance notice of changes in methodology, source data, and statistical techniques. These changes are disclosed to the public the first time that data incorporating the new methodology are published.

1.3.1 Guidelines for staff behavior are clear and publicized

Although strictly speaking there is no code of ethics, staff regulations clearly establish employees’ obligations. Recently, in the context of processes for renewal and institutional modernization, the BCCR prepared a basic set of values that includes ethical principles such as honesty and loyalty. This set of values is found on the Intranet site of the issuing institution.

Personnel are informed regarding the Internal Labor Regulations of the BCCR. Chapters IX, X, XI, and XII indicate the obligations, prohibitions, and sanctions that apply to BCCR staff.

2.1.1 The overall structure in terms of concepts and definitions follows international standards, guidelines, or agreed practices

National accounts are compiled according to the conceptual framework and definitions in the 1993 SNA.

2.2.1 The scope is broadly consistent with international standards, guidelines, or agreed practices

National accounts cover the economic activities of all resident institutional units, based on 1993 SNA recommendations. Consistent with the concept of residence, this includes the activities of companies located in free zones, Costa Rican embassies, and workers residing abroad for a period of less than one year.

Production boundaries are consistent with the 1993 SNA, regardless of the coverage obtained. Illustrated below is the extent to which the value of production is measured for some specific activities: the quantification of own-account production for all goods intended for final household consumption is partial, given that consumption of own-produced agricultural products is estimated and consumption of own-produced manufactured products would be covered by the 1994 survey of small manufacturing enterprises. Own-account research and development activities are not identified separately. The production of goods for own-account fixed capital formation is quantified in principle: the construction component through construction permits and machinery and equipment through the commodity flow. Petroleum exploration is quantified. The production of literary or artistic originals for amusement and the sale of illegal products to buyers disposed to purchase them are not quantified due to practical difficulties in obtaining the information.

The scope of assets generally coincides with the 1993 SNA. Given the absence of a regular army, all defense expenditures on new construction and equipment (domestic security) are considered assets that may have civilian uses. In addition, valuables and historical monuments have not been included. Works in progress in agriculture are included for those products for which that calculation is relevant, among them coffee and sugar cane. However, they are valued at cost and not at basic prices. This calculation is also done for construction activity. Computer applications are included in assets only when purchased from third parties and not when companies develop them in-house, as is also required by the 1993 SNA. Original art and patented entities are not included in production or in balances.

2.3.1 Classification/sectorization systems used are broadly consistent with international standards, guidelines, or agreed practices

The classification of economic activities corresponds to the United Nations International Standard Industrial Classification of All Economic Activities, Rev. 2 (ISIC Rev.2), to four digits for the annual accounts and one digit for quarterly indicators. The classification of products used is based on the Central Product Classification (CPC). A breakdown of household consumption is done by ad hoc national classification and not by categories in the Classification of Individual Consumption by Purpose (COICOP). A functional classification of the government’s final consumption expenditure has not been implemented.

2.4.1 Market prices are used to value flows and stocks

In Costa Rica’s national accounts, transactions are recorded, in principle, at the market prices in effect at the time the transactions occur, consistent with the valuation principles of the 1993 SNA. Output is valued at basic prices; intermediate consumption is valued at purchase prices, excluding the deductible portion of sales and excise taxes, as applicable (value added taxes); the components of final consumption expenditure and gross capital formation are valued at purchase prices, including distribution margins and the applicable taxes on products. Total exports and imports are valued at f.o.b. prices.

2.4.2 Recording is done on an accrual basis

Generally all transactions are recorded on an accrual basis with the exception of some government income and expenditure items that are recorded on a cash basis. Expenditure is then adjusted for commitments based on the execution of the budget when compiling the final figures.

2.4.3 Grossing/netting procedures are broadly consistent with international standards, guidelines, or agreed practices

The procedures for calculating gross and net figures are consistent with the 1993 SNA. Thus, for example, transactions among establishments belonging to the same enterprise are recorded on a gross basis.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

There is no complete and up-to-date business register that could be used as a basis for sample surveys covering all activities and types of units. However, for the manufacturing survey, the IICE has maintained a frame of manufacturing enterprises since 1976, updated starting in 1991 based on the CCSS registry. This frame has been used for regular updates of the manufacturing survey sample. New activities, offshore processing, and the free zone have been included.

In order to conduct economic surveys of trade and services for the 1991 base change, in 1989 the Directorate General of Statistics and Censuses (DGEC) developed enterprise sampling frameworks by using administrative records, primarily from tax and social security sources. These frameworks have not been updated since 1992, a function explicitly entrusted to INEC under Law 7839. There is also an updated sampling framework for exporters and importers and recently an area sampling framework for agriculture (2001).

Only for the manufacturing sector are enterprise statistics compiled through a regular program of surveys that solicit quarterly and annual data from establishments. The universe covers establishments employing more than 20 workers; thus coverage would amount to 85 percent of the aggregate value of all manufacturing. ISIC Rev. 2 and ISIC Rev.3 classifications are used. The response rate is satisfactory (more than 85 percent) in the case of the quantity indicators. Although financial statements are requested, the response rate to this is low. In addition, response is very low for the production account questionnaires. The sample design and the calculation procedures are representative of the survey universe. Random scientific sampling techniques are used, and suitable imputation methods are employed to take non-responses into account. Expansion factors are scientifically derived on the basis of the sample design. The calculation procedures and design of the sample are reviewed and corrected regularly (see 3.3.1 i).

Household income and expenditure surveys are not conducted annually.

Comprehensive surveys/censuses have been conducted for population and housing (1984 and 2000) and for the agricultural sector (1984). Suitable methods and procedures have been used in the population and housing surveys. However, problems with the agricultural census have made it unhelpful for national accounts purposes. In addition, household income and expenditure surveys have been conducted, most recently between November 1987 and November 1988. This survey had national coverage and was based on a probabilistic sample design. Given the undervaluation of the principal aggregates, the survey was only marginally used in the new base year 1991 in some per capita items and consumption structures.

The principal data sources for annual estimates of GDP on the production side are as follows: (i) for the agricultural sector, the compilation of the national accounts uses different sources of information provided by government offices involved in this activity, producer associations, and foreign trade statistics; (ii) private construction without public financing is based on administrative data on construction permits; (iii) for transport activity there are data from various sources on amounts provided by type of service and the number of vehicles registered. In addition, there are consolidated financial statements from the banking system and detailed financial data from public communications, electricity, and water and sewerage companies. The general government production account is based on accounting data from each of the levels of government.

The information used in the estimates of personal and business services is among the weakest given that there are only employment indicators and some indicators of service quantities and prices.

There are no surveys of establishments or companies at higher than annual frequencies, with the exception of the survey of small and medium manufacturing establishments to obtain quarterly (ITCE) and monthly (BCCR) volume indices consistent with the annual data, as mentioned in point 3.1.1.ii.

For the remaining activities, the principle data sources for quarterly estimates of GDP on the production side are:

• Agricultural. The data on quantities harvested, sold on the local market, or exported are obtained from government offices, customs records, and records from producers or traders of agricultural products and processing industries.

• Construction. There are monthly records for the private component (construction permits).

• Electricity, gas, and water. Administrative monthly records are used.

• Trade. There are monthly administrative records for 400 commercial establishments since 1998.

• Transportation, storage, and communications. Different sources provide quantity indicators for services provided. The information includes, among other data, statistics on tons of merchandise moved in ports, number of tourists entering and leaving the country, minutes of telephone communication, and passengers/kilometers transported per plane.

• Financial intermediaries. There are financial statements from banks, financial corporations, and institutions of other sectors.

The principal data sources for quarterly GDP estimates by expenditure are:

• Monthly report of expenditure on final household consumption compiled on the basis of administrative records from commercial establishments (see iv of previous point).

• For gross fixed capital formation, INEC and the Directorate General of Customs (DGA) contribute information on private construction and expenditure on machinery and imported equipment, respectively.

• The DGA provides information on exported and imported goods. The balance of payments provides data on services.

Ad hoc economic surveys are usually conducted to complete the data necessary for the change of the base year.

An expanded economic survey was actually conducted on the manufacturing sector for this purpose. It covered all companies with more than 20 employees and incorporated questions appropriate for determining intermediate consumption. This survey was conducted by the IICE and financed and commissioned by the BCCR. In addition to the manufacturing industry survey, there was additional information from administrative sources that had detailed information on transactions by enterprises subject to special tax or financial benefits and financial statements available in administrative records.

In addition, there was a survey of small manufacturing enterprises that gathered data for 1994. It was based on a probabilistic sample, for which the sampling framework included enterprises registered in the CCSS and enterprises included in other administrative records, as well as verification of that framework in the field. Extrapolation methods are used for the series estimates of this stratum.

The BCCR also financed economic surveys of trade and services activities that were conducted by what was then the Directorate General of Statistics and Censuses. The samples were selected by following a stratified random sample design in which large establishments were selected with certainty and smaller establishments were selected with probability proportional to size. The results of the trade and services surveys were far from satisfactory due to the low response rate obtained. The data from these surveys were also supplemented with the company financial statements available in administrative records.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

In some cases, the source data are not consistent with the definitions, scope, classification, and valuation of the national accounts. Thus, adjustments are made to the source data as part of the process of validating the data for compiling the accounts. For the base year, the consistency of the results from the surveys was reviewed and adjustments were made as indicated, including those due to undervaluation of the results. The corrections included adjusting the valuation to account for underreporting of income and overvaluation of production costs. The opinion of accounting experts interviewed contributed to this exercise.

3.1.3 Source data are timely

Price, international trade, balance of payments, and monetary statistics are timely. Lag time is greater for obtaining government finance statistics.

3.2.1 Data compilation employs sound statistical techniques

3.2.2 Other statistical procedures (e.g., data adjustments and transformations and statistical analysis) employ sound statistical techniques

The sampling frameworks used to select samples and calculate grossing-up factors cover registered enterprises, including self-employed workers, and thus would exclude the activities of individuals who do not report to the tax authorities or the CCSS. This omission in the survey would involve the informal sector of the economy attributable to the institutional household sector (unincorporated household enterprises). These activities could be important in trade, transport, and personal services, as well as in some groups of business services. However, the coverage of these activities in the national accounts is comprehensive given that trade and transport services produced by the informal sector were covered by the sources and estimation techniques used.

No official estimates were prepared for activities not covered in the national accounts.

3.3.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessments are monitored and made available to guide planning

In the case of the quarterly manufacturing survey, the only survey conducted, the sampling error amounts to 15 percent. There is no information on non-sampling error. The selection of samples is adjusted every five years. The non-response rate is almost 15 percent. Atypical values are checked through direct consultations with the enterprises. The survey is audited to verify the accuracy of individual responses.

In compiling the estimates of GDP by activity, each year the observed data obtained either through surveys/censuses or appropriate administrative records cover approximately 70–80 percent of output. For the quarter there are output indicators for 60 percent of production; the remainder is obtained by a quarterly allocation of an annual projection, leveled out using the Chow-Lin “no indicator” method. It should be pointed out that in the monthly indicator of economic activity (MIEA) the monthly change in economic activity is estimated with and without the “high-technology electronics industry,” as agreed with the basic data sources.

There are no data for estimating intermediate consumption in accordance with an input/output matrix on an annual or quarterly basis.

When compiling estimates of GDP by expenditure components, the estimate of annual household expenditure is extrapolated from the base year with appropriate indicators and in some cases by using fixed ratios. These results are adjusted in the process of reconciling supply and use. Quarterly household consumption is estimated by using sample administrative records that encompass 56 percent of that variable. There is no independent calculation of final expenditure by nonprofit institutions serving households.

In the estimate of gross fixed capital formation, the observed data derived from administrative records represent more than 90 percent of the annual figure and 70–80 percent of the quarterly figure. As in the case of the estimate of quarterly value added by activity, the remainder of gross fixed capital formation is obtained by distributing an annual projection over the quarters, leveled out using the “no indicator” Chow-Lin method.

There are predetermined and systematic procedures for validating important parts of the source data. Thus, for example, the validation of the tapes from customs is made through the informatics system used to process the custom data, which allows the detection, inter alia, of duplications and erroneous codes. Furthermore, the consistency of the foreign trade variables over time is carefully observed. Similarly, the consistency of wages and salaries reported by the industrial survey is compared with related CCSS sources.

The consistency of data in terms of definition, valuation, and reference periods was checked for the new base year. Basically, source data are corrected in the context of revisions.

3.3.2 Main intermediate data are validated against other information where applicable

The main intermediate data are validated using independent source data and statistical frameworks in all cases where such information is available. This is the case with the manufacturing industry and agriculture, where there are relatively independent surveys and administrative records.

3.3.3 Statistical discrepancies and other potential indicators of statistical problems are investigated and made known to users

The general scarcity of fully independent sources does not facilitate the identification of statistical discrepancies between GDP based on production activities and GDP based on expenditure components. In addition, the methods used to determine changes in inventories can contaminate this variable with differences in intermediate consumption at constant prices derived from the production accounts by assuming fixed ratios and the value of intermediate consumption obtained from the expenditure side.

3.4.1 Studies and analyses of revisions are carried out routinely to inform statistical processes

Studies and analyses of the revisions are not performed.

3.4.2 Revision studies and analyses are disseminated and used to guide users

As mentioned above, revision studies are not carried out. However, in the case of the new base year (1991) disseminated in April of 2000, the sources and methods employed were very widely disseminated in public and specialized settings, as was the revision of the published figures that incorporated the new results.

4.1.1 Processes to monitor the relevance and practical utility of existing statistics in meeting users’ needs are in place

There are no explicit reviews to assess the relevance of the program. However, the project to implement the 1993 SNA is being developed in order to expand the statistical products offered to date to users within the central framework of that system. In addition, specific requests have been received from public and private institutions to support the compilation of the satellite tourism account and the satellite health account. In these cases, the role of the program has included compiling the account with the basic information provided by interested parties in the first case and with technical support provided to carry out the second study.

4.2.1 Periodicity follows dissemination standards

The periodicity of the national accounts is consistent with the SDDS.

4.2.2 Timeliness follows dissemination standards

The quarterly accounts are released one quarter after the conclusion of the reference period, consistent with the SDDS. The annual data from the definitive national accounts are released no later than 11 months after the end of the reference year. Dissemination is subject to the timeliness and availability of the principal source data.

4.3.1 Statistics are consistent or reconcilable over a reasonable period of time

There are uninterrupted and consistent time series for a period of no less than five years. The new national accounts series covers the period 1991–2000. The previous series, with a base year of 1966, covered the period 1966–1998. The erratic movements detected once the trend behavior of the series is analyzed are examined on the basis of additional information available that could explain those movements. The DCS has not linked the new national accounts series with the 1966-based series and users were so informed.¹ Whenever new annual data are generated, the consistency of the indicated structural ratios is analyzed, as well as others that may be considered relevant for the analysis of the Costa Rican economy.

4.3.2 Statistics are internally consistent (accounting identities observed)

There are no discrepancies between the GDP estimate on the production side and on the expenditure side as the latter is derived in the context of the supply and use balances in which the estimates of final expenditure components are submitted to analysis for logical and numerical consistency. Residual discrepancies are absorbed in the estimation of changes in inventories, which in large part is obtained as the residual of total supply minus the sum of other uses.

The annual and quarterly data are consistent in conceptual terms, in that they utilize the same definitions. The discrepancy between the sum of quarterly estimates and the annual estimate is corrected using the Chow-Lin method.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources or statistical frameworks

There are differences, albeit minor, between net credit/net borrowing taken from the national accounts and the sum of current account balances and the capital account taken from the balance of payments.² The source of such differences is:

• the inclusion in 1991 and 1992 of alcohol exports not taken into account in the balance of payments;

• the downward adjustment of the level of exports reported in the balance of payments for 1996–1998, for the national accounts, when fictitious sales designed to unduly obtain export subsidies were detected;

• the omission from the national accounts of specific re-exports made in 1999. This difference will be corrected when the final figures are released for the year 2000; and

• the existence of different criteria in the national accounts and the balance of payments with respect to the treatment of some services. Although in some cases agreement has finally been reached (educational and insurance services), in other cases the differences still continue (aircraft repair) because there is no relevant information for a joint decision.

4.4.1 Revisions follow a regular, well established, and transparent schedule

Although there is no pre-determined and pre-announced schedule for the publication and revision of the annual figures, they follow a stable pattern over time. The dissemination and revision dates for the figures are coordinated with the gathering of principal source data and the programming of important official documents on economic policy.

In effect, there is an internal schedule for compiling final data and annual projections. The preliminary version of the national accounts indicators for the year is prepared in February of the following year for the bank’s Annual Report. This version basically covers the data on the first three quarters of the reference year.

A revised preliminary version is calculated in mid-year, based on data on the four quarters of the reference year, administrative records on foreign trade, and results from the manufacturing survey. This version is released to the public and used internally in the BCCR for purposes of monetary programming. The final version, which is prepared between August and November, adds the pending administrative records and is used internally by the BCCR to consolidate projections at the closure of macroeconomic accounts for the current year.

Quarterly data are subject to the schedule agreed upon with the IMF for the SDDS, which is an advance schedule.

4.4.2 Preliminary data are clearly identified

Preliminary and revised data are identified in the tables. The revised data are accessible at sufficiently detailed levels.

4.4.3 Studies and analyses of revisions are carried out periodically and made public

There is no formal mechanism to explain to users the reasons for discrepancies between preliminary and final data. However, BCCR officials are normally informed of the reasons for these discrepancies. These officials usually transmit this information to the general public. Any questions users may have are amply handled via e-mail or by telephone.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

Detailed estimates of the national accounts are disseminated with their corresponding time series. The data are catalogued to make it easier for users to identify and consult them and reflect the formats frequently used by the national compilers of these statistics and by international organizations. The DCS has not received any comments from users regarding the formats used.

5.1.2 Dissemination media and formats are adequate

The annual and quarterly estimates of the national accounts are not disseminated in a specialized annual or quarterly publication. However, these estimates are available electronically on the BCCR’s Internet site as well as at the BCCR Information Center. In addition, the DCS handles all requests for information in this area that it receives directly, using various media for this purpose.

5.1.3 Statistics are released on a pre-announced schedule

As indicated in 4.4.1, there is no pre-established schedule for the annual accounts. As of July 2001, the quarterly accounts have a pre-announced schedule.

5.1.4 Statistics are made available to all users at the same time

As stated in 1.2.2., statistics are released at the same time to all interested parties.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

Upon request, nonconfidential information is provided with the same level of detail used in the compilation, as long as this does not violate the principle of confidentiality and the BCCR’s commitments in this regard.

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from international standards are noted

The BCCR Information Center has methodological documents on the annual and quarterly national accounts. These documents do not indicate the links with the principal data systems or differences with respect to other systems.

5.2.2 Levels of detail are adapted to the needs of the intended audience and the type of compilation

Metadata with different levels of detail according to the intended audience are not provided.

5.3.1 Contact person for each subject field is publicized

There is no person assigned to provide assistance in the area of national accounts. However, all requests for assistance are handled.

5.3.2 Catalogues of publications, documents and other services, including information on any changes, are widely available

Information on publications, documents and other services is available on the BCCR Internet site and at the BCCR Information Center.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The consumer price index (CPI) is compiled and disseminated by the National Institute of Statistics and Censuses (INEC). The legal foundation of INEC is specified in Law 7839 of November 4, 1998, which creates the National Statistical System (SEN) and establishes INEC as its governing body. Article 15 of the law empowers INEC to produce the following statistics: (i) statistics based on administrative records; (ii) population and housing statistics based on censuses; (iii) statistics based on multi-purpose household surveys, agricultural surveys, household income and expenditure surveys, economic surveys, price surveys for the consumer price index and the producer price index), foreign trade surveys, etc.

For the consumer price index, practice is consistent with the law. INEC is self-sufficient in producing and compiling both the household income and expenditure survey, which provides the basis for selecting the basket and weights for the index, and the index itself.

0.1.2 Data sharing and coordination among data producing agencies are adequate

As mentioned above, INEC is completely self-sufficient in producing the CPI and no other agencies are involved in its compilation. The data are collected on a timely basis as is clear from the fact that the index is published only two business days after the end of the reference month, consistent with a provision approved by the institution’s Board of Directors. Ongoing contact is maintained with the data providers through the normal process of data collection, which is carried out through personal interviews.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

Article 51 of Law 7839 establishes that violations of the statistical confidentiality required by law of government employees or other individuals or legal entities that provide service to SEN units will be penalized under Article 203 of the Penal Code and will also constitute a serious labor violation when government officials are involved. These provisions are also strengthened in Articles 33 and 34 of the law’s regulations.

Article 4 of Law 7839 establishes that the agencies and units that make up the SEN must compile and handle statistical data according to the principles of confidentiality, transparency, specialization, and proportionality. Further, the regulations for the National Statistical System Law published on August 24, 2000 in the official gazette, La Gaceta, establish detailed rules regarding the handling of data by SEN entities, which are compatible with confidentiality. For example, Article 31 stipulates for information requested from legal entities and individuals that “the confidentiality of the information provided shall be guaranteed, wherefore data shall be disseminated and published only in the form of aggregates covering three legal entities or individuals, making it impossible to obtain individual facts or figures by a process of elimination.”

Computerized databases for compiling the CPI containing information from the individual records of respondents are protected by access codes. Only two authorized persons, who are responsible for compiling the index, handle this information. In addition, the physical installations of INEC have surveillance services and security devices that prevent access by unauthorized persons.

Only upon initial contact are respondents given a letter signed by the INEC manager explaining to them the purpose of the survey and how the data they provide will be used. In addition, they are given a brochure summarizing the methodology. The confidentiality of the data is properly protected in the physical installations of INEC, which have appropriate security services allowing entry only to authorized persons. The confidentiality of price data is not very sensitive given that by-law merchants must make the prices of their products public; nevertheless, the law dictates that INEC cannot provide the individual data.

In practice, INEC has established data aggregation procedures that prevent the disclosure of individual data, including those that could be identified by a process of elimination. Access to individual data is limited to officials directly involved in compiling the index. In addition, when files are delivered to research institutions, the data files are prepared in a way that maintains their confidentiality.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The law establishing the SEN does not include the requirement that individuals or legal entities in the private sector must provide information for statistical purposes to INEC or other statistical agencies. The original draft of the law did consider both that requirement and the establishment of penalties for violations. However, the final version of Law 7839 was approved without these stipulations. The law does specify the manner and conditions under which the information must be provided, and Article 11 of the law provides that the information must be submitted on hardcopy or electronically without charge to the statistical institutions.

INEC makes efforts to improve the public’s willingness to collaborate with the statistical research the institution conducts by focusing on improving its own image and that of its statistical products. To this end, workshops and seminars are conducted to present the products.

Finally, it should be pointed out that in practice INEC has no difficulty obtaining price data.

0.2.1 Staff, financial, and computing resources are commensurate with institutional programs

It is considered that the staff available is sufficient in terms of quality and quality to carry out the CPI program in its current context. It should be pointed out that one of the considerations in determining the geographic coverage of the index was specifically its adaptation to the available resources. The staff involved in compiling the index includes nine people, of whom five are interviewers and four are professionals. The latter group includes a computer specialist and others who have training in statistics or economics. Most of the staff responsible for compiling the index have more than three years’ experience with such work. In terms of training, some of the staff have participated in seminars overseas conducted in the context of cooperation with Spain and devoted to conceptual and methodological issues in the compilation of the CPI.

Salary incentives have improved with the establishment of INEC compared with what existed in the Directorate General of Statistics and Censuses, which it replaced.

The CPI is one of the priority programs of INEC and to date has had adequate financial resources.

Computing resources are adequate and the equipment has been upgraded recently. During the past year a new computer system was designed and introduced; this allows greater speed in data processing and expands the possibilities of analysis. The entire process of calculating the index has been automated.

0.2.2 Measures to ensure efficient use of resources are implemented

An accounting system has not been established in INEC that would make it possible to measure the costs of the CPI program. Currently its costs are known only in approximate terms. There is now a plan to incorporate a costing system designed to establish prices for statistical services.

0.3.1 Processes are in place to focus on quality, to monitor the quality of the collection, processing, and dissemination of statistics, to acknowledge and deal with tradeoffs within quality, and to guide planning for existing and emerging needs

Within INEC there is great awareness of the need to produce quality products as a means of improving institutional image and winning the public’s confidence. In practice, there is an ongoing process to monitor the quality of the information produced. The quality of the index has been improved with the new computer system, as was mentioned previously, owing to greater versatility in analyzing primary data and intermediate output.

Although the SEN Law calls for the creation of an Advisory Committee, this has not yet happened. There are no procedures for consulting with users in order to obtain their reactions to the index.

Only in the design stage of the index were there consultations with users to consider their information needs. It should be pointed out that, as a result of wage negotiations in the current year and at the request of the labor unions, a committee was created to examine the CPI methodology and to suggest technical improvements. The committee is made up of the Ministry of Economy, the Ministry of Labor and Social Security, the unions, and INEC. Experts from the universities and the BCCR are being invited to participate. In addition, this year a new initiative targeted consultation with users by organizing a workshop in February 2001 with users from the public and private sectors for the purpose of learning their views on INEC’s products and their information needs.

1.1.1 Statistics are compiled on an impartial basis

Only technical criteria are applied in compiling the CPI. There is no outside influence in any part of the process of compiling the CPI, and the compilation is done according to a pre-established methodology.

INEC is an autonomous institution of the State and as such enjoys independence in carrying out its functions. The institution’s manager is appointed by the Board of Directors, which is made up of representatives from the various sectors of society as established in the SEN Law. INEC staff is appointed through an internal and external competitive process based on professional qualifications. The power to make these appointments lies with the manager.

In principle, the administration of INEC supports the preparation of methodological documents and analyses but, for practical reasons relating to constraints on staff time, this activity is insignificant. In addition, there exists a practice of giving talks about the CPI to universities and unions.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

The compilers of the CPI are completely independent in the selection of data sources, which they make only on the basis of technical considerations.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

INEC is authorized to make public clarifications regarding erroneous interpretations or misuse of its data.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The terms and conditions under which the CPI is compiled and disseminated are determined in Law 7839 on the SEN, which is publicly available. The law was published in the official gazette of November 4, 1998 and is also available from the INEC Internet site. Printed copies of the law can be purchased for the cost of printing at the INEC Information Center.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

It has not yet been publicly disclosed that the Minister of Economy has access to the data one hour before they are published.

1.2.3 Products of statistical agencies/units are clearly identified as such

All the publications of INEC identify the source of the published data. The press release reporting the data to the media bears the printed logo of INEC. INEC requires that explicit reference be made to the data source when its data are reproduced in the publications of other agencies and in the press.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

Methodological changes are incorporated only when a change is made in the base of the index, which happens infrequently. The CPI methodology is then published in a special publication that can be purchased by the general public for a nominal price.

1.3.1 Guidelines for staff behavior are clear and publicized

The INEC service regulations establish the standards of behavior for its staff in Title V, Chapter I. These are made known to staff in orientation sessions when they join the institution and are periodically reviewed in monthly bulletins issued by the INEC human resources unit.

The service regulations are available to all INEC employees. Ethical standards are observed in practice and management carries out checks in this regard.

2.1.1 The overall structure in terms of concepts and definitions follows international standards, guidelines, or agreed practices

Costa Rica’s CPI bases its weights on the 1987/88 household income and expenditure survey. Household consumption expenditures used in the weights for the index are consistent with the conceptual framework and definitions of the national accounts. The concept of household consumption expenditure used in the household income and expenditure survey on which the weights of the current index are based corresponds to purchased consumption, i.e., it includes goods and services purchased on the market by households, consumption of goods produced for one’s own consumption, and payments of wages and salaries in kind. However, consumption of rental services for owner-occupied housing is not imputed.

In compiling the CPI, INEC uses very precise specifications for the products (varieties) for which prices are collected. These specifications are detailed in the Manual of Specifications for the Interviewer.

2.2.1 The scope is broadly consistent with international standards, guidelines, or agreed practices

Costa Rica’s CPI covers urban households in provincial capitals (San José, Alajuela, Heredia, and Cartago) in the country’s Central Region, which represented 44 percent of total national population in 1988. Excluded from the index’s reference population are households consisting of a single individual and those consisting of more than 12 people. All socioeconomic strata are represented in the CPI. In the case of households with unincorporated businesses, household consumption expenditures are separated from expenditures relating to production. The selected coverage was based on cost criteria. In any case, the coverage of the CPI with 1995 as the base year is greater than that of the previous CPI, which was limited to the metropolitan area alone.

With the exception of rent attributed to owner-occupied dwellings, there are no differences, in principle, between the coverage of the products included in the scope of the CPI and those included in household consumption expenditure in the national accounts. The weights of the CPI include all market goods and services intended for household consumption and also goods produced by households for their own consumption. The weights of the CPI do not include illegal goods and services.

Although the normal user of the CPI may not be aware of the different coverage in comparison with consumption expenditures in the national accounts, these differences are known to the compilers of the national accounts and the CPI as well as to economic analysts in the public and private sectors.

2.3.1 Classification/sectorization systems used are broadly consistent with international standards, guidelines, or agreed practices

In principle, family units (households) are defined according to the definition of the corresponding institutional units in the system of national accounts. It is relevant to point out that, although the design and execution of the 1987/88 housing survey predate the appearance of 1993 SNA, the latter did not change the earlier definition of this institutional sector. Similarly, transactions classified as household consumption expenditures correspond to those in the 1968 System of National Accounts (1968 SNA), which is marginally different in this respect from 1993 SNA. The CPI is compiled for 46 subgroups, eight groups, and the total. Classification corresponds to the categories recommended in 1968 SNA for classifying household consumption expenditure. The eight groups correspond to (i) food, beverages, and tobacco; (ii) clothing and footwear; (iii) housing; (iv) furniture and accessories; (v) medical care; (vi) transportation; (vii) entertainment and education; and (viii) other goods and services.

2.4.1 Market prices are used to value flows and stocks

The CPI weights are based on household expenditure on goods and services valued at the market prices in effect at the time of purchase. These are purchase prices, i.e., they include both taxes and distribution margins for the products in the price paid by the consumer.

Product specifications are detailed and only prices of cash transactions are collected. Offer prices and promotions are excluded.

2.4.2 Recording is done on an accrual basis

The prices used in compiling the index are the market prices in effect at the time the data are collected.

2.4.3 Grossing/netting procedures are broadly consistent with international standards, guidelines, or agreed practices

Purchases of used goods by households, particularly vehicles, are included in the index weights using net figures, in accordance with recommended practice for recording these transactions.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

The framework for selecting the sample for the income and expenditure survey was an area framework with national coverage. The framework exhaustively identified all the households in the country. The basic information for the framework came from the 1984 population and housing survey and from updates (the segments included in the sample were updated through field work).

The household income and expenditure survey is conducted at infrequent intervals. The last survey was done in 1987/88. The survey before that was in 1974. There are now medium-term plans to conduct a new survey to update the CPI weights and basket. The 1987/88 survey covered all households in the nation based on a random representative sample of 4,884 households. The survey collected information at detailed product levels (goods and services) that included about 2000 products. Expenditure reflected household consumption according to the national accounts concept.

Both the sampling design and the calculation techniques followed appropriate procedures. The sampling design represented a random phase-stratified sample. Enumeration segments were selected in the first phase and households were selected in the second phase.

Prices are collected monthly through direct interviews at establishments conducted by INEC enumerators. Surveying is done every day of the month and every effort is made to visit each establishment in the sample on the same day each month. The establishments were selected following a random sampling procedure. In total, 1,100 establishments are visited, chosen from among those where consumers most frequently make their purchases. From these, 12,500 price observations are obtained. Prices for electricity, water, and telephone are obtained directly from the companies that provide these services, and these prices are checked against those published in the official gazette. A particular case is housing rents, for which prices are collected monthly from a sample of 150 households. These households are divided into three groups of 50, each of which is visited every three months in the same order within each quarter.

The sampling frameworks for selecting establishments was based on the 1990 directory of establishments, which served as the framework for conducting the economic surveys for changing the base year in the national accounts. This directory was updated based on administrative records. Random selection methods were used to select a representative sample for each type of establishment in which consumers make their purchases, such as supermarkets, stores, markets, and farmers’ markets.

There are detailed specifications for collecting prices that ensure that observations in different periods represent the same variety.

No ad hoc surveys are conducted. New products and changes in weights are included only when there is a new household income and expenditure survey.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

Both the price data collected and the expenditure data based on the household survey are completely consistent with the definitions, scope, and classifications of the CPI. The prices collected correspond to the market prices in effect for cash transactions on the day of the interview.

3.1.3 Source data are timely

Source data are received on a timely basis. INEC does not use prices collected for other systems of price statistics in compiling the CPI.

3.2.1 Data compilation employs sound statistical techniques

As mentioned earlier, the level of detail in consumer expenditure provided by the 1987/88 National Household Income and Expenditure Survey included about 2000 products (goods and services) classified according to household consumption expenditure in the national accounts. The results of the survey were obtained at the six-digit level in the product classification.

Household consumption expenditure used for the CPI weights corresponds to the definition of household consumption expenditure in the national accounts but, as mentioned earlier, it has more limited scope in that it excludes the imputation for rental services of owner-occupied homes. Financing costs incurred for credit purchases were deducted from the value of the goods purchased, when it was possible to separate them out. Goods produced by households for their own use were valued at market prices for similar products.

The CPI basket of goods and services consists of 264 items, representing 75 percent of household consumption expenditure of the index’s reference population. These items were selected on the basis of the criterion that they should be consumed by at least 10 percent of the households and that their relative share in total consumption expenditure should be equal to or greater than 0.09 percent. The basic indices correspond to those of the 264 items included in the compilation of the index, which are calculated as short-term relatives (compared to the previous month) of average prices. These indices are aggregated at higher levels of 46 subgroups, eight groups, and the general index using the modified Laspeyres formula. The weights are already obsolete because they are based on the 1987/88 expenditure structure. The reference period of the index is January 1995 and for these purposes expenditures in the 1987/88 base period were aligned with the reference period by applying to them the specific product price changes occurring between the two periods.

The techniques used to deal with problems caused by products that temporarily disappear from the market are consistent with the options that are common practice in the compilation of price indices. Costa Rica follows the option of using the last price observation seen during the three months following the disappearance of the product. If the product disappears for more than three months, it is replaced by another with specifications that are as similar as possible to those of the product that has disappeared. The replacement is done with appropriate techniques, taking into account differences in quality so as to avoid introducing bias. Differences in quality are handled with the overlapping method, i.e., two consecutive monthly observations are awaited for the replacement product in order to apply the price change to the cumulative index of the product replaced. New products are only included in the compilation of the CPI in Costa Rica at infrequent intervals when the index base is changed and on the basis of results from a new household income and expenditure survey. The infrequency with which these surveys are compiled would require utilizing other sources (such as statistics on production and imports or conducting ad hoc surveys) to identify new products that have gained a significant share in the consumer goods and services market.

The treatment of seasonal products in the compilation of the index is not relevant in Costa Rica given that the index basket contains no seasonal products. This is probably because of the relative lack of importance that such products might have had in the base for the index weights, but also because the country’s climatic conditions and its openness to foreign markets mean that the supply of agricultural products continues throughout the year.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

As mentioned earlier, the geographic coverage of the CPI is limited to the Central Region, whereas the implicit deflator of household consumption in the national accounts naturally has national coverage. In addition, although the scope of coverage of consumer activities in both systems is similar (except for rental expenditures attributed to owner-occupied homes), the nature of the national accounts estimates is generally more comprehensive than that of the estimates obtained in the household income and expenditure surveys.

3.3.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessments are monitored and made available to guide planning

The household income and expenditure survey on which the CPI is based was conducted with a high level of methodological soundness and yielded estimates of sampling errors at aggregate levels of groups of products. The survey emphasized quality in the results with complex procedures to verify the data in the field and in-office analysis of the questionnaires with the help of computer programs designed especially to detect omissions, inconsistencies, and extreme values. The non-response rate was about 15 percent, and corresponding adjustments were made to the size of the sample.

In the CPI estimate, all household consumption expenditures used in the weights come from the results of the 1987/88 household income and expenditure survey. The expenditures of the CPI reference population constitute 68 percent of total national household consumption expenditure according to the survey. The products included in the index’s basket represent 75 percent of the consumption expenditures of the index’s reference population during the base period of the weights.

3.3.2 Intermediate results are validated against other information where applicable

Prices collected are validated at various phases of CPI data collection and compilation. There is constant supervision of data collection, and prices are analyzed by comparing them with those recorded in earlier periods by fieldwork supervisors and checked again with the establishments. Price fluctuations between the various observations collected for a same variety are also compared. Analysts conduct manual and automated examinations of the data collected at different phases in the processing of the primary data included in compiling the index. No checks of intermediate results are done by comparison with other price index systems, such as the producer price index (PPI).

3.3.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated and made available to guide users

In Costa Rica the results are not revised and the CPI is final when published.

3.4.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

This is not applicable in Costa Rica given that the CPI published figures are not revised.

4.1.1 Processes to monitor the relevance and practical utility of existing statistics in meeting users’ needs are in place

As mentioned earlier, there is no process of regular consultation with CPI users. Based on a recent initiative by the unions, a committee has been created with representatives from the government and the unions to analyze the methodology of the index and to suggest improvements. In the past, users were only consulted at the stage when a change was being introduced in the index base. In addition, INEC’s participation in meetings and seminars on prices organized by international organizations is irregular.

4.2.1 Periodicity follows dissemination standards

The periodicity of the CPI of Costa Rica is monthly and consistent with the IMF’s Special Data Dissemination Standard (SDDS).

4.2.2 Timeliness follows dissemination standards

Costa Rica’s CPI is published on the second business day following the close of the reference month, so that it exceeds SDDS timeliness requirements, which stipulate a 30-day period.

4.3.1 Statistics are consistent or reconcilable over a reasonable period of time

The CPI series are consistent for sufficiently extensive periods of time. Methodological changes are introduced only when changes are made in the base year, which happens infrequently. Erratic price movements are analyzed, and the products that have the greatest effect on the monthly change in the index are published monthly.

4.3.2 Statistics are internally consistent

This is not applicable to Costa Rica given that the index is not calculated separately by items and by geographic areas.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

Changes in CPI prices are not regularly compared to those obtained in other price index systems for comparable items.

4.4.1 Revisions follow a regular, well established, and transparent schedule

As mentioned earlier, no revisions are made to the CPI figures, which are final when published. In addition, the index is published on predetermined dates that follow a stable pattern. Publication of an advance release calendar for the index was recently started (on the BCCR Internet site). INEC must follow this standard in order to meet the requirements for subscription to the IMF’s SDDS.

Publication of the index is not subject to the preparation of any official economic policy document.

4.4.2 Preliminary data are clearly identified

This is not applicable as the data are final when published.

4.4.3 Studies and analyses of revisions are carried out routinely and made public

This does not apply in the case of Costa Rica as no revisions are made in the CPI.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of texts, tables, and charts)

Long and detailed series are available at the INEC Information Center. The regular monthly publication provides detailed information for the current month only. In addition, the monthly series for the last five years at the group level are published on the INEC Internet site. A brief analysis of the results is presented in the monthly bulletin.

5.1.2 Dissemination media and formats are adequate

The CPI data are presented in the form of indices and changes in a dedicated monthly publication. The data are also published through a press release and through the INEC Internet site.

5.1.3 Statistics are released on a pre-announced schedule

The results of the CPI are published on a very timely basis and according to a pre-announced release schedule.

5.1.4 Statistics are made available to all users at the same time

In principle, the estimates are made available to all users at the same time. Technically, the Minister of Economy receives the information within the hour prior to release for purposes of preparing his public response.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

Upon request, users can obtain information from the indices at the level of each item included in the CPI basket.

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from international standards are annotated

There is a CPI methodological publication with base year 1995, which is very limited in scope. In addition, methodological summaries have been prepared that will be included on the institution’s and the IMF’s Internet sites when Costa Rica subscribes to the SDDS.

5.2.2 Levels of detail are adapted to the needs of the intended audience and type of compilation

A methodological description that is more detailed than the published version is available for internal INEC use.

5.3.1 Contact person for each subject field is publicized

The INEC Information Center assists users in obtaining the information they need and directs them to specialists in the corresponding area when necessary.

5.3.2 Catalogues of publications, documents and other services, including information on any charges, are widely available

INEC recently finished preparation of a catalogue of publications and has included it on its Internet site.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The PPI is prepared and disseminated by the BCCR based on the Central Bank’s Organic Law 7558 of November 3, 1995 and subsequent amendments. Although that law does not indicate that this price index must be compiled by the BCCR, Article 14 (d) does establish that the BCCR must publish information on Costa Rica’s economic situation every month, providing information on at least production, prices, currency, exports, imports, and gross and net international reserves. The unit within the BCCR that compiles the PPI is the Department of Social Accounting (DCS), which reports to the Economics Division.

Under the terms and conditions of the law on the National Statistics System (SEN), Law 7839 of November 4, 1998, which establishes INEC and gives it the role of directing the SEN, producer price indices for goods and services are part of the national statistics that INEC is to compile. However, the function is currently being performed by the DCS, which is also responsible for compiling the national accounts. In accordance with the law referred to above, the national accounts will be transferred to INEC in January 2003. The preparation of the PPI has not yet been transferred to INEC, in anticipation of making an orderly transfer of functions that does not hinder the normal production of these statistics.

0.1.2 Data sharing and coordination among data producing agencies are adequate

The data sources for constructing and compiling the PPI are primarily private sector establishments. Although there are no formal agreements on providing the data, in practice there is a good degree of collaboration. The BCCR makes an effort to foster adequate submission of data by respondents.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

Pursuant to Article 40 of the BCCR’s Organic Law, “Government offices and units and autonomous institutions shall be required to provide assistance to the departments of the Central Bank, so that they can carry out their functions efficiently. For this purpose, they must promptly supply them with the data, reports, and studies they request. Failure to meet this obligation on the part of responsible officials in government offices and agencies and autonomous institutions shall be considered a serious dereliction of duty.

“Employees of the Central Bank of Costa Rica have access to tax information for statistical purposes only. They must respect the prohibitions and limitations established in Article 117 of the Code of Tax Standards and Procedures, Law 4755 of May 3, 1991. They shall also be subject to the provisions of Article 203 of the Penal Code.”

While that law does not specify the private sector’s obligation to provide information to the BCCR for statistical purposes, requests from the Social Accounting Department for data for the price indices always indicate the confidential nature of the data to be provided by the sources. Based on the opinion of the BCCR’s Legal Counsel, this fact alone necessitates the confidentiality of the information gathered. In addition, Article 37 (g) of Chapter IX of the Internal Labor Regulations of the BCCR indicates that “Employees must maintain the utmost secrecy regarding the operations, business, and actions of the Bank and its clients,” and this is supplemented by Chapter XII on the sanctions applicable for failure to comply with this standard.

In addition, the law establishing the SEN and that would, in future, support this price index compilation function includes sanctions in this area. In effect, Article 4 of Law 7839 establishes that all agencies belonging to the SEN are required to compile and process statistical data in accordance with the principles of statistical confidentiality.

Computer programs have mechanisms that help to protect confidential data. In effect, the processing of information is protected by personal access codes accessible to authorized personnel only.

The requests for information that the DCS issues regarding the information on the price indices always indicate the statistical purposes of the data being requested, as well as the confidential nature of the data the respondents provide, which is expressed in an official letter from the BCCR. Based on the view of the BCCR Legal Counsel, this fact alone requires that the information collected for the price indices be treated as confidential.

In addition, measures are adopted to protect information when it is destroyed, an action that is subject to the National Archives Law. The BCCR’s auditing unit monitors compliance with these provisions. Data storage facilities have security measures to protect the data.

No special aggregation standards have been developed to ensure the protection of individual data. However, in practice, due to the confidentiality requirement, the aggregation principles followed are consistent with this requirement. Only employees responsible for data processing have access to the individual data. When files are provided for research purposes, precautions are taken to avoid disclosing individual data, eliminating opportunities to identify the respondent, or aggregating the data in a way that is compatible with the principle of confidentiality.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

There are no legal provisions or standards regarding the compilation of data and the provision of information. Given that providing data in this case is largely voluntary, considerable efforts are made to build a good relationship with data providers in order to obtain information. In many cases, the BCCR offers them statistical information produced by the BCCR so as to provide a service in exchange for the information requested.

0.2.1 Staff, financial, and computing resources are commensurate with institutional programs

The number of staff (4) assigned to index compilation is adequate, given that the data are collected electronically (by fax and e-mail) and processing is largely automated.

Staff is highly trained and includes two graduate statisticians with significant experience in the design and compilation of price indices. These employees have participated in various overseas seminars on modern techniques for compiling price indices. In addition, there are two employees with training and experience in gathering basic information.

There are no problems retaining staff. The average longevity of the professional staff involved in this work is more than 10 years.

There are no financial resource problems in compiling the PPI. The program has adequate financial resources, which have been used to update the base of the index two times in the last 10 years.

The program also has the computing resources needed to provide timely and quality results.

0.2.2 Measures to ensure efficient use of resources are implemented

The BCCR has implemented a standardized cost control system by homogeneous products. In recent years efficiency in the production of the price indices has actually increased through the combination of a reduction in the staff involved in this work and greater automation of the processes.

0.3.1 Processes are in place to focus on quality, to monitor the quality of the collection, processing, and dissemination of statistics, to acknowledge and deal with tradeoffs within quality, and to guide planning for existing and emerging needs

The BCCR appreciates and broadly supports efforts designed to improve the quality of statistical work as a prerequisite for promoting transparency and strengthening confidence in public policies. For this reason, the DCS has established systems to control the quality of and validate the data derived from its sources.

The quality standard of maintaining an appropriate number of products and informants in the index baskets equal to the average observed in the preceding year is followed.

Considerations regarding new and emerging needs for data are resolved internally. The mechanisms are determined basically according to national accounts needs, and thus requirements for new data are resolved internally at the Economic Division level.

1.1.1 Statistics are compiled on an impartial basis

Only technical considerations are applied in compiling the price indices. The professionals are completely independent in selecting the sources and statistical methods they consider most appropriate for making the estimates. The results are not subject to any external or internal influence.

The legal provisions do not explicitly stipulate the professional independence of the staff responsible for compiling the price indices. However, professional criteria guide the selection of management personnel by means of a publicly open and competitive process. Although there are no regulations to this effect, in practice there is no undue outside influence on the content or dissemination of statistical information.

In practice, comprehensive documents have been prepared on the methodology of the index at the time of changes in the base of the index. Similarly, activities have been organized to present the methodology and results of the index in different forums (press, internal workshops). Finally, the various users have the option of asking the BCCR for talks to learn about or clarify the conceptual and methodological aspects of the index.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

All decisions regarding data sources are made in the DCS. This means that the compilers of the statistics are free to select the basic data in terms of quality, timeliness, cost, and response burden of the informants.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The DCS is authorized to make comments on errors in interpretation or misuse of the statistics. The practice is to hold quarterly meetings with the press for this purpose.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The terms and conditions under which the PPI is compiled are explicitly defined in Article 14 (d) of Law 7558. These provisions are available from the BCCR Internet site and in printed form.

The methodology is available on the Internet site and physically at the Central Bank’s Information Center.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

No governmental authority has access to the data before they are published. The PPI is disseminated simultaneously to all users.

1.2.3 Products of statistical agencies/units are clearly identified as such

Data released to the public are clearly identified as products of the DCS. The BCCR identifies the source of the information in its publications. The source is usually mentioned in a footnote.

1.2.4 Advance notice is given of major changes in methodology, data sources, and statistical techniques

There is no advance notice of changes in methodology. These changes are disclosed to the public when the data incorporating the new methodology are published for the first time.

1.3.1 Guidelines for staff behavior are clear and publicized

Although there is no code of ethics strictly speaking, personnel regulations clearly establish workers’ obligations. Recently, in the context of renewal and institutional modernization processes, the BCCR developed a basic set of values that include ethical principles such as honesty and loyalty. This set of values is also available on the BCCR Internet site.

As indicated, Chapters IX, X, XI, and XII of the Internal Labor Regulations of the BCCR consider the obligations, prohibitions, and sanctions to which its employees are subject. Staff is informed of these rules.

2.1.1 The overall structure in terms of concepts and definitions follows international standards, guidelines, or agreed practices

The design of the PPI of Costa Rica followed the model established by the United States Bureau of Labor Statistics. The valuation principles and time of recording for weights and prices correspond to the System of National Accounts 1993 (1993 SNA). The scope of the index is the production of manufactured goods intended for the domestic market. Production in free zones, bonded warehouses, or factories operated by companies subject to special customs regimes and producing for the foreign market is not included.

The weights of the published PPI were calculated on the basis of net sales of manufacturing establishments on the domestic market, which is consistent with internationally accepted practice. A price index has also been calculated for exports of manufactured and agricultural goods for 1999, 2000, and 2001. It is expected that the index will be published this year. An aggregate index would also be compiled (products for the domestic market and exports) at the four-digit level of ISIC Rev 3.

The production of goods to be used by another establishment belonging to the same company was considered in cases where there is an exchange “price.” Barter, changes in inventories, or works in progress were not valued (in many cases the companies do not calculate these data and, even when they do maintain records of this information, it is not considered feasible for the companies to submit the data).

Indices are not calculated for stage of processing of the goods or according to the use to which the products will be put.

The definition or specification of the individual goods that make up the PPI basket follows the recommendations issued by the United States Labor Department, described in the methodological publications of the Bureau of Labor Statistics. In this regard, consideration is given both to the characteristics of the products and to the conditions of sale and type of buyer when there are price differences for these reasons (in the case of transactions in the domestic market, it is up to the discretion of the companies to provide this information).

2.2.1 The scope is broadly consistent with international standards, guidelines, or agreed practices

The scope of the published index includes market manufacturing enterprises that direct their production (or part of it) to the domestic market. Thus, free zones, customs depots, or factories operated by companies under customs control that direct their production to the foreign market are not included.

Based on data from the year 2001, the baskets of goods produced and exported for both the manufacturing and the agricultural sectors will be incorporated, with the exception of companies operating under the free zone and inward processing regimes. The latter will also be incorporated in the PPI starting in 2002.

As mentioned earlier, the coverage of the index published to date differs from that of the national accounts, in that it only considers goods manufactured and intended for the domestic market and thus constitutes a subset of manufacturing production measured in the national accounts. The users know this difference in coverage to the extent that the PPI methodology mentions its limited scope.

As indicated earlier, all goods produced by companies in the manufacturing sectors and sold on the domestic market during the reference period were taken into account in the construction of the PPI weighting structure. The production of goods for use within the same company was taken into account in cases where there is an exchange “price.” Sales of illegal market goods sold to voluntary buyers were not valued.

2.3.1 Classification/sectorization systems used are broadly consistent with international standards, guidelines, or agreed practices

The classification of units and transactions is consistent with 1993 SNA. Establishments are classified according to ISIC Rev.3 and also ISIC rev. 2 to guarantee their utilization in the national accounts, which, for the present, continue to use the latter classification. The classification of products corresponds to the characteristic branch of production.

2.4.1 Market prices are used to value flows and stocks

The data on sales used for index weights are valued at basic prices.

Exported production (in the case of the as-yet-unpublished exports index) is valued at f.o.b. prices. As mentioned earlier, indices are not calculated for intermediate consumption or for production according to stage of processing. Thus, the valuation of goods for intermediate consumption is not relevant in the case of Costa Rica’s PPI.

The product specifications include the characteristics of the transactions in cases where such characteristics lead to different prices. In this regard, consideration was given to aspects such as the payment period and type of buyer, including in the latter case the different prices for preferred clients and transactions between different units of the same company. In practice, the coverage and accuracy of the information regarding the characteristics of transactions on the domestic market are subject to the willingness of companies to provide information voluntarily. It has not always been possible to obtain this type of information considering its sensitivity to the reporting companies. In such cases, the most commonly reported transaction was selected.

2.4.2 Recording is done on an accrual basis

The sales data used for the index weights correspond to values at basic prices of transactions at the time they occur.

2.4.3 Grossing/netting procedures are broadly consistent with international standards, guidelines, or agreed practices

This standard is not relevant in the case of the PPI of Costa Rica given that indices are not compiled by stage of processing of the product or by category of economic use, as indicated earlier.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

To select the sample of enterprises, the BCCR constructed a complete sampling framework for 1996 and updated it to 1997 based on records from companies registered with Social Security in Costa Rica (the coverage of such records is deemed to be good at the level of medium and large companies in the manufacturing sector).

To replace products and respondents that cease to participate in productive activity, the raw records (not refined by classification) from Social Security for the year in progress are used as a provisional sampling framework.

For exported production, there is an exhaustive framework of all export products (customs records) that is updated monthly and with a four-month lag.

There is no periodic program of surveys for complete updating of weights and baskets. Comprehensive updates of both are performed when the base for the index is changed. However, there are partial updates for company products and companies in specific branches subject to analysis. The weights for exported production will be adjusted when they diverge by 10 percent from the actual data (as reflected in the updated population totals available from the customs registry and classified by ISIC).

The survey used as the basis for the PPI weights covers more than 80 percent of the manufacturing industry (the calculation of this figure does not take into account the handicraft sector, for which there are no updated and reliable measures). The survey universe was made up of all manufacturing companies employing 10 or more people that sold their production in the domestic market. In constructing the PPI weights for the domestic market, small companies (less than 10 employees) were not taken into account because (i) they represent 75 percent of the enterprises and only 13 percent of national production, (ii) they are very erratic in terms of their activity and location, and (iii) a high percentage of them cease to operate in the short term. All of these factors introduce a significant amount of statistical noise and makes the fieldwork difficult. In the case of exported goods, the weights are based on population and the extraction of samples was done on the basis of the entire framework.

Both the sampling design and the estimates follow an appropriate procedure. The sampling design to obtain the PPI weights for the domestic market is a two-way stratified sampling (the sampling framework was stratified according to ISIC division to two digits and within each classification strata were constructed according to size). The size of the sample within each ISIC was calculated using the Neyman formula. This sampling size was distributed by optimal allocation and finally samples were selected within each stratum using the selection methods with probabilities proportional to size. The stratum of 100 and more was assigned a unit probability. The size strata were 10–20, 21–40, 41–60, 61–100, and more than 100. The fraction of the total sampling was 22.5 percent. All branches of significant activity are represented in the sample. In the case of exports, the coverage approximates the census data (about 87 percent).

For the 1997 survey, it was not necessary to impute data to cover nonresponse as the design incorporated a 5 percent increase in the sampling size to anticipate the failure to respond, which was 2.1 percent. The approach was to adjust the expansion factors accordingly.

The data obtained from the companies corresponded to items produced, sales, and percentage exported. With this information, a master sample was constructed that consisted of a set of generic products (with a nonspecific definition). To determine the index basket, all generic products in the master sample were included. These were broken down into specifications following the criterion of obtaining at least three specifications by generic product for three or four informants. The specification was also selected on the basis of a random sampling technique with probability proportional to size.

The price survey sample contains 216 establishments, from which prices were obtained for 1,044 items. The sampling design for the basket and index weights is revised only when there is a change in the index base.

The products are specified with sufficient detail to identify a product that is unique in its price. There are no groupings by product “category” at a higher level of detail than in the ISIC.

No periodic surveys are done for unincorporated household enterprises.

When necessary, ad hoc surveys are conducted to complement the program of periodic surveys (e.g., to identify new products and provide source data for weights and prices).

When necessary, partial surveys are conducted in specific branches to identify changes in relative importance, substitutes, and new products or informants.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

The prices collected reflect the definitions, scope, classification, valuation, and time of recording of the PPI. These are sales prices on the domestic market, at basic prices, for transactions in the reference month.

3.1.3 Source data are timely

Source data are timely and are obtained throughout the reference months from enterprises (in the case of local transactions) and from export invoices (in the case of exported goods). Other price statistics systems are not used as sources of information.

3.2.1 Data compilation employs sound statistical techniques

The manufacturing survey used as the basis for the PPI weights yielded representative results with an accuracy of 90 percent and a confidence level of 95 percent for manufacturing activities at the two-digit ISIC level. The results obtained are considered valid for most classes (four ISIC digits).

The survey provided detail on the products at the generic description level (a total of 116 generic items were obtained). The sample included all these items, which were opened in a second phase of consultation with the companies to determine different specifications. The products are specified with sufficient detail to identify a product that is unique in its price. There are no groupings by product “category” at a higher level of detail the ISIC branch of activity.

For each generic item there is a variable number of specifications, which correspond to similar products (representative of a generic product) produced by different companies. The indices for each specification are aggregated as a weighted average of each specification’s share in the total value of the generic item.

The PPI weights are sales in the domestic market, which is consistent with internationally accepted practice for compiling indices.

The PPI is calculated for the entire manufacturing industry and at the level of 64 classes, 38 groups, and 19 divisions of ISIC Rev.3. The basic component of the indices is the ratio of the prices for each specification compared to the previous month. These ratios are aggregated at the branch of activity level as a weighted average based on their share of the total branch in question using the modified Laspeyres formula, which is also applied at higher levels of aggregation of the indices.

The period of the PPI weights is the year 1997 and, as mentioned earlier, the reference price is the price in the month of December 1999. The weights were not aligned with the reference period.

When the indicator weights are changed, the new index is “linked” to the old, “moving” the new level backwards with the variations in the old index. In the only change of weights made to the PPI, the link was only at the level of the general index (more disaggregated levels were not linked due to the human resource costs, given that the previous PPI was calculated using ISIC 2 and the new one using ISIC 3).

When product prices are temporarily missing, they are imputed as follows: (i) by applying the price change for other similar products, (ii) by using the price change for other goods produced by the enterprise, or (iii) by imputing the change in the branch to which the missing price belongs. Seasonal products are completely eliminated from the sample during the time they are not marketed. This situation occurs only in the case of agricultural products for export, as the manufacturing sector PPI contains no seasonal products.

Products that disappear from the market permanently are eliminated from the calculation of the index (removed from the base period and the current period). In some cases there is substitution, when the market is directed to buying a substitute product. In this case the internal weights of the companies or the specific product within the companies vary.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The coverage of the PPI is quite complete in that it covers more than 80 percent of manufacturing production, as stated earlier. The national accounts provide a deflator of the production of this industry with complete coverage. Estimates of the national accounts for the production of small business were based on an ad hoc survey for 1994 and extrapolations for subsequent years.

3.3.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessment are monitored and made available to guide planning

There is information on the non-response rates for all sampling stages for the two occasions when the structure of weights and the baskets of products and informants for the PPI were constructed (1991 and 1999). There is no information on sample errors, and the complexity of the design makes it difficult to calculate them. With respect to the percentage of source data that are imputed, records are maintained that can be used to calculate them.

The PPI is representative of more than 80 percent of manufacturing production. The coverage of the sample of establishments is 22.5 percent of the target population of the survey to determine the weights. The prices are collected for a sample of specific items that cover between 15 and 20 percent of the index’s universe.

The source data for the weights were analyzed at the enterprise level by comparing the ratios with companies in similar circumstances. In addition, the overall structure of relative weights was validated by comparison with that calculated for the national accounts and, in absolute terms, the comparison was made at the general level and at aggregate levels similar to the ISIC divisions (pairs were not compared due to the different classifications used in the national surveys and in the PPI, i.e., ISIC 2 and ISIC3, respectively).

The prices collected are submitted to monthly critical analyses, as well as an analysis of the consistency with other variables over longer periods.

3.3.2 Intermediate results are validated against other information where applicable

The results are periodically validated at the level of very important products and at the level of similar product groupings.

3.3.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated and made available to guide users.

Discrepancies in the calculation of the PPI are not relevant in that the PPI is calculated only by branches of activity.

3.4.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

The PPI data are always calculated in final form each month. No delay in reporting by key establishments has been noted.

4.1.1 Processes to monitor the relevance and practical utility of existing statistics in meeting users’ needs are in place

There is no formal process for consulting with users outside the BCCR. The index is compiled primarily to meet national accounts needs and secondarily to supplement the information on price trends.

4.2.1 Periodicity follows dissemination standards.

The periodicity of the index is monthly, consistent with the requirements of the IMF’s SDDS.

Timeliness follows dissemination standards.

The index is published five business days after the end of the month, exceeding the requirements established in the IMF’s Special Data Dissemination Standard (SDDS).

4.3.1 Statistics are consistent or reconcilable over a reasonable period of time

The series of the indices, even at detailed levels, maintain consistencies for periods of at least five years. The 1992 base series included data for the period 1991–1999. The new base is expected to be maintained for at least five years. As mentioned, retrospective series were not created for the subgroups of the index when the change was made in the base year. In order to provide a service to users, the old and new series are linked at the general index level.

4.3.2 Statistics are internally consistent

This is not relevant in that the index is compiled only by branches of activity.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

Comparative analyses of the components of various indices are performed, particularly of the CPI, in order to be certain of the reliability of the results of the indices. These procedures also use information regarding the demand for manufactured products, making it possible to detect possible changes in the trade margins, which could explain possible divergences in the trends observed in these two indices.

4.4.1 Revisions follow a regular, well established, and transparent schedule

There is no calendar of revisions of provisional estimates as the data that are disseminated are always final. With respect to revisions in weights and methodology, there is no predetermined schedule.

The publication dates of the index are not subject to the dissemination dates of any official document, but rather they represent the minimum amount of time possible for collecting and compiling the indicator.

4.4.2 Preliminary data are clearly identified

This does not apply as there are no preliminary PPI estimates.

4.4.3 Studies and analyses of revisions are carried out routinely and made public

This does not apply for the same reasons as in the previous point.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of texts, tables, and charts)

The original series and the monthly and cumulative rates of change in the general index since 1991 are published on the BCCR Internet site, as is the monthly index for each of the 19 divisions of ISIC, revision 3. Printed monthly publications contain the same information as is released on the BCCR Internet site. None of these cases includes analyses of the results. On the other hand, if users require a level of detail greater than that published, they can request it from the BCCR, which will provide it bearing in mind the principle of confidentiality that governs the statistics of the BCCR.

5.1.2 Dissemination media and formats are adequate

The PPI estimates are disseminated through electronic media, by using a voice mail system, and through data sheets available at BCCR headquarters. There are no press releases. As mentioned above, additional or more detailed information is available upon request by interested parties.

5.1.3 Statistics are released on a pre-announced schedule

An advance release calendar for PPI data was incorporated in the BCCR Internet site while the mission was in Costa Rica; this is consistent with SDDS requirements.

5.1.4 Statistics are made available to all users at the same time

Data are released simultaneously to all users.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

As mentioned earlier, this is a usual practice of the BCCR.

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from international standards are annotated

A comprehensive methodological document is produced at the time of a change of the index base. It is available to the public at the BCCR Information Center (for the cost of photocopies) and on the BCCR Internet site.

5.2.2 Levels of detail are adapted to the needs of the intended audience and type of compilation

Metadata are available with the same level of detail for all users, in the document mentioned in the preceding point.

5.3.1 Contact person for each subject field is publicized

The Basic Statistics Section of the BCCR, which is responsible for compiling the index, has the function of responding to users’ questions about the indicator. The BCCR Information Center also supports this work. Currently no employee responsible for providing this service has been identified. Nonetheless, starting with the country’s subscription to the SDDS, this information will be available on the BCCR Internet site.

5.3.2 Catalogues or publications, documents and other services, including information on any charges, are widely available

The BCCR has no catalogue of publications.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The Central Bank of Costa Rica (BCCR) compiles, processes, and disseminates balance of payments statistics. The Organic Law of the Central Bank of Costa Rica (BCCR), Law 7558 of November 3, 1995, does not specifically assign these functions to the BCCR. However, Article 14 (d) of the law indicates that each month the BCCR shall publish a statistical summary of the country’s economic situation, to include at least information on production, prices, currency, credit, exports, imports, and gross and net international reserves. In addition, the Organic Law empowers the BCCR to establish and publish the methodology it will use to prepare this statistical summary, as well as the changes it makes to the methodology. The economic situation report includes data on the external sector. The Balance of Payments Area (ABP) of the BCCR’s Monetary Department is the unit that compiles balance of payments statistics.

The Organic Law of the BCCR establishes that the country’s economic institutions must collaborate with the BCCR so that it can achieve its goals in the best way possible. Article 40 of the law establishes that the offices and agencies of the State and autonomous institutions shall be required to provide their assistance to the departments of the BCCR. However, nonfinancial private sector entities are not required by law to provide information to the BCCR. Currently underway is an initiative to establish the requirement that private sector entities provide statistical information to the BCCR.

Law 7839 of 1998 creating the National Institute of Statistics and Censuses (INEC) gives the institution the role of compiling foreign trade statistics and national accounts, a role currently carried out by the BCCR; but the understanding is that there is no intention to transfer the preparation of statistics to INEC in the immediate future. The law, for which no regulations have been drawn up as yet, creates the National Statistical System (SEN) to streamline and coordinate statistical activity. However, INEC, the governing body of the SEN, has still not designated which entities will be part of that system.

0.1.2 Data sharing and coordination among data producing agencies are adequate

The BCCR has established operational mechanisms to obtain information on components of the balance of payments from public and private sector entities. The Directorate General of Customs (DGA) provides monthly information on imports and exports; the Foreign Trade Promotion Agency (PROCOMER) provides monthly information on export licenses; the company Componentes Intel de Costa Rica S.A. provides monthly information on its foreign transactions; the Costa Rican Tourism Institute (ICT) provides information on travel; and the Treasury of the Ministry of Finance (MOF) provides information on public external debt. In addition, the ABP obtains information directly on services and direct investment of private companies through surveys and has established appropriate mechanisms to obtain data from other BCCR departments or areas like the Operations Recording and Control Department and the Monetary Programming Area.

The ABP holds periodic meetings with the Foreign Investment Committee, made up of the BCCR, PROCOMER, the ICT, and the Coalition on Development Initiatives (CINDE, a private institution), for the purpose of coordinating information requirements and avoiding duplication of efforts. There are also regular meetings with the Ministry of Finance (MOF) to coordinate the needs for information on foreign public debt and transfers and with the ICT to coordinate information under the travel heading. ABP staff consults in person and by telephone with the principal companies providing basic statistics for compiling the balance of payments, in order to discuss problems and mechanisms for reporting data.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

Article 14 of the Organic Law of the BCCR establishes that the BCCR is required to protect the confidentiality of the individual information provided to it by individuals and legal entities. In addition, Law 7839 creating the SEN clearly establishes that the information obtained for statistical purposes shall be strictly confidential, except for information coming from public institutions and nonstate public institutions, which shall be freely available to citizens. The data collected to prepare statistics can be used only for these purposes. In addition, the October 1987 Internal Labor Regulations of the BCCR establish that employees are required to maintain the utmost secrecy regarding the operations, business, and activities of the BCCR and its clients.

The cover letter that accompanies surveys and other requests for statistical information informs companies or institutions that the information requested of them will be treated in a strictly confidential manner and in aggregated form for the purpose of compiling balance of payments statistics.

The Internal Labor Regulations of the BCCR and the Statistics Law provide procedures to prevent the disclosure of information on individual respondents. The regulations characterize as gross misconduct (i) the commission of a serious indiscretion regarding BCCR operations and the interests of its clients or (ii) the discovery, with or without authorization, of the content of work-related or nonwork-related books or documents for the purpose of making improper use of these data or transmitting information to third parties. The Statistics Law establishes that public employees or other individuals or legal entities that provide services to agencies of the SEN, which violate statistical confidentiality, shall be penalized under Article 203 of the Penal Code. The Statistics Law also establishes that data on individuals or legal entities must not be provided or published individually, but rather as part of aggregate figures, except with prior authorization from the party that provided the data. For these purposes, aggregate figures shall be deemed to mean those that refer to three or more individuals or legal entities.

In the ABP, access to individual data is restricted to staff that requires such information to carry out its professional activities. Some measures have been taken to protect this information, e.g., information on companies is coded, and only responsible staff has access to the list of codes and company names.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The Organic Law of the BCCR does not specifically assign the BCCR the role of compiling balance of payments statistics, and the BCCR is not empowered by law to require the reporting of information by private sector companies and individuals for purposes of compiling balance of payments statistics. Given that the BCCR does not have legal authority to request information from the private sector, it is likewise unable to impose sanctions on those who do not report. However, ABP staff has established procedures to encourage voluntary reporting, e.g., in the letter that accompanies surveys the BCCR explains to respondents the importance of the information being requested of them for purposes of compiling the balance of payments and asks for their cooperation in providing the information sought in the surveys.

The ABP gathers direct information from public and private companies for purposes of compiling the balance of payments. In order to ensure adequate reporting, ABP staff visits large companies and provides assistance in filling out the form to those companies that request it. The survey forms include the telephone number and e-mail address of the person who can answer questions. To promote collaboration on the part of respondents, the BCCR sends brochures with statistical information to companies that provide information to the BCCR for purposes of compiling the balance of payments.

The BCCR’s Operations Recording and Control Department, which prepares the external debt statistics used to compile the balance of payments, has implemented some mechanisms to improve the compilation of data on public external debt. Nevertheless, although the BCCR has legal power to request information from the public sector, public institutions are not required to report statistics on public debt in a regular and systematic way.

0.2.1 Staff, financial, and computing resources are commensurate with institutional programs

Financial, computing, and staff resources are generally sufficient to compile the balance of payments statistics. However, the demand for new activities, such as implementing quarterly surveys to compile the balance of payments, has created the demand for additional resources.

The ABP has 10 professional staff members, including a chief, who are responsible for collecting, processing, and disseminating balance of payments statistics.

The staff has adequate computing resources to compile and disseminate balance of payments statistics. Each staff member has an individual computer, with Internet access and appropriate software. Linked Excel spreadsheets are used to compile the balance of payments and international investment position. Import and export data are obtained on line from the BCCR’s Department of Social Accounting. In addition, computing resources include a database for the quarterly and annual surveys, access to the integrated Debt Management and Financial Analysis System (DMFAS), access to the database used to generate monetary statistics reports from the BCCR and the rest of the National Financial System, and access to the Financial Accounting Information System (SIFC) required to generate the reports on international monetary reserves. A large percentage of private sector responses is sent by e-mail.

The staff has solid knowledge of and familiarity with the concepts and methodology recommended in the fifth edition of the Balance of Payments Manual (BPM5). Most of the members of the section have many years of experience and/or have received in-country or overseas training in the methodology of compiling the balance of payments. New staff members receive theoretical and practical training in data collection methods from employees with extensive experience, and they are encouraged to participate in courses on balance of payments methodology in the BPM5. BCCR management promotes staff training through its participation in balance of payments courses and seminars organized by the IMF, the Center for Latin American Monetary Studies (CEMLA), the Central American Monetary Council, and other institutions. In the year 2000, the BCCR organized a balance of payments course in San José in collaboration with CEMLA; various employees from the BCCR and from other Latin American countries attended it.

0.2.2 Measures to ensure efficient use of resources are implemented

In order to ensure the efficient use of resources, an operational program has been introduced at the management level. The Annual Operating Plan, sets goals and objectives at the section level; consolidation occurs at the departmental level. Each unit prepares the budget with common departmental goals and specific goals for the section. The goals include outcomes and there is quarterly follow-up of the Annual Operating Plan, which compares plans with actual achievements. In addition, the BCCR seeks to take advantage of the latest advances in computing in order to save resources in compiling and disseminating statistics.

On various occasions the BCCR has requested consulting services from CEMLA, the IMF, and the Central American Monetary Council to train its personnel and evaluate the methodology being applied. In recent years, the IMF has provided technical assistance for the adoption of the balance of payments methodology according to BPM5.

0.3.1 Processes are in place to focus on quality, to monitor the quality of the collection, processing, and dissemination of statistics, to acknowledge and deal with tradeoffs within quality, and to guide planning for existing and emerging needs

There are no formal quality assessment processes. However, the management of the BCCR is very interested in establishing procedures to promote and assess the quality of the information that the BCCR produces. To this end, BCCR employees are receiving training in the total quality management program, ISO 9001. The head of the ABP has received training in this program, and two employees are attending the ISO 2001 course. The APB employees who have received ISO 2001 training should implement the total quality control processes in the ABP, including preparing manuals on compiling and processing balance of payments statistics. In addition, procedures have been established to conduct ongoing quality assessment of the surveys and verification of data collection and processing.

ABP employees participate in seminars and conferences on the balance of payments and seek to implement international standards in the country, for which they sometimes request assistance from international organizations.

Tradeoffs among quality dimensions are acknowledged (e.g., between timeliness and accuracy and reliability). However, these inverse relations are not communicated to users, although they generally relate to the use of preliminary data. For example, the preliminary export data are based on export license records and are subsequently replaced with final data received from customs.

There are no surveys of users to assess quality. However, there are mechanisms to obtain comments regarding data quality from the members of the BCCR Board of Directors, which includes representatives from the public and private sectors. Although there are no formal mechanisms to obtain information on new statistical needs, requests regarding new information requirements are handled.

1.1.1 Statistics are compiled on an impartial basis

The BCCR is empowered to establish the methodology that it will use to prepare the statistical summary of the economic situation. The Organic Law also establishes that the BCCR must publish this methodology. APB staff members have a significant degree of autonomy in their work and try to implement international standards in compiling and processing balance of payments statistics. The assignment of tasks in the area of data compilation is based solely on technical and professional criteria.

The BCCR organizes conferences and seminars and promotes staff training through attendance at courses organized by international institutions in the country or overseas such as the IMF, CEMLA, and the Central American Monetary Council. In addition, it promotes post-graduate courses for employees (specialization, masters, and doctorate). The BCCR is authorized to pay the costs for its staff to attend training courses in the country or overseas. Staff hiring is the responsibility of the BCCR Human Resources Department. However, such hiring is done at the recommendation of the area involved.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

APB staff choose the source data and statistical methodologies used in compiling the balance of payments, taking into account statistical considerations only, evaluating the timeliness and quality of alternative sources. The selection and design of surveys are based solely on statistical considerations and take into account international methodological standards. The introduction of new surveys is based exclusively on the importance of certain activities, as in the case of new activities in the services sector.

Important methodological changes, such as the adoption of the methodology recommended in the BPM5, are disseminated. Balance of payments reports and analyses are done with objectivity and independence; the BCCR is the only agency that publishes and analyses these statistics.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The BCCR regularly releases explanatory reports on balance of payments developments in the review Weekly Economics, in the Annual Report of the BCCR, and in the annual report Balance of Payments of Costa Rica. The Economic Indicators section published on the BCCR Internet site includes quarterly balance of payments data but not explanatory notes. The Executive President of the BCCR regularly holds press conferences to discuss new figures and analyze general developments in the economy, including balance of payments data.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

Laws and regulations governing the collection, processing, and dissemination of statistics are released on the corresponding Internet site. The Organic Law of the BCCR is disseminated on the BCCR Internet site, which also includes the name of the person who can be contacted to obtain additional information on the statistics its publishes. This person refers requests for information to the relevant area. The Statistics Law is published on the INEC Internet site.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

The BCCR Board of Directors, made up of the President of the BCCR, the Minister of Finance, and private sector members, has access to some statistics before they are published, but the public is not informed of this practice.

1.2.3 Products of statistical agencies/units are clearly identified as such

Publications of the BCCR clearly indicate that the balance of payment statistics have been compiled by that institution and identify the data sources.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

There are no advance notices of substantial changes in methodology, source data, or statistical techniques. However, important changes in methodology, such as converting the balance of payments information from BPM4 to BPM5, are published simultaneously with the new statistics. In future, the BCCR plans to provide advance information of the adoption of new methodologies and the incorporation in the data of results from new surveys and special studies.

1.3.1 Guidelines for staff behavior are clear and publicized

Ethical standards for staff behavior are contained in the Internal Labor Regulations of the BCCR, which establish rights, prohibitions, and penalties for failure to meet these standards. The regulations are disseminated on the BCCR Intranet site and made known to all new employees when they assume their positions. However, they are not communicated to the public.

2.1.1 The overall structure in terms of concepts and definitions follows international standards, guidelines, or agreed practices

The structure and classification of the balance of payments statistics are consistent with the methodological standards established in the BPM5. The BCCR began compiling and publishing balance of payments statistics according to the BPM5 in 1997. The publication Balance of Payments of Costa Rica 1992–1997 includes a methodological note explaining the major variations and changes made with the introduction of the BPM5. The international investment position was partially compiled in 2000.

2.2.1 The scope is broadly consistent with international standards, guidelines, or agreed practices

The balance of payments statistics include the transactions of institutional units resident in Costa Rica with the rest of the world. The criterion of residence is determined on the basis of principles established in the BPM5. The statistics include transactions with nonresidents by companies that process goods for manufacture, operating under two regimes: inward processing (“maquiladoras”) and free zones. There are no estimates for contraband, undervaluation of imports, or overvaluation of exports. The financial account excludes private bank external holdings of equity securities and long-term foreign liabilities³. In addition, residents’ financial assets and liabilities transactions with offshore banks associated with the private banking system are excluded, except for those deposits that might be included in the surveys or information compiled by the Bank for International Settlements (BIS), which are used as sources for obtaining such information. The coverage of some services is limited, particularly in the quarterly calculations. Discrepancies with respect to the principles outlined in BPM5 are not identified

2.3.1 Classification/sectorization systems used are broadly consistent with international standards, guidelines, or agreed practices

The balance of payments statistics are organized according to the structure and classification recommended in the BPM5. The unit of account for the balance of payments is the U.S. dollar. Following BPM5 guidelines, the balance of payments is compiled according to the principal standardized components classified in the main groups of accounts: the current account, which includes goods and services, income and current transfers; the capital account; and the financial account The analytical balance of payments is published, including reserves as a below-the-line component.

In the balance of payments, exports and imports are recorded at f.o.b. values. Customs records imports at f.o.b. values and c.i.f. values. Under the goods heading, there is separate identification of general merchandise, goods for processing, repairs on goods, goods procured in ports by carriers and nonmonetary gold. Goods for processing include goods that are exported or imported to be processed under the free zone and inward processing (“maquiladora”) regimes. The foreign trade figures processed by the Accounting Department of the BCCR exclude repairs on goods and goods procured in ports.

The capital account records capital transfers and the financial account records the assets and liabilities of Costa Rica compared to nonresidents. The Economic Indicators Section published on the BCCR Internet site includes quarterly balance of payments data in which the financial account is classified by functional categories under direct investment, portfolio investment, and other investment; reserve assets are presented as a below-the-line item. The components of the financial account are also classified by financial instruments and by institutional sectors. In the Annual Report and the Economic Review of the BCCR, the principal classification of the financial accounts is by institutional sectors: the public sector (broken down into disbursements, repayments, and other) and the private sector.

In the financial account, the data on direct investment in the country are obtained from the survey of companies. The information on portfolio investment is obtained from the surveys of public and private companies and from administrative records of the MOF and the Stock Exchange. The data on foreign transactions of commercial banks used to be obtained from the survey of banks but are now obtained directly from the report on the banks’ foreign assets and liabilities. This report, prepared by the Monetary Programming Area, excludes long-term foreign liabilities and the banks’ foreign holdings of equity securities. The data on international reserves are obtained directly from the balance sheet accounts of the BCCR, denominated in U.S. dollars.

2.4.1 Market prices are used to value flows and stocks

In the balance of payments, transactions are valued at market prices. However, in the international investment position, stock data are recorded at accounting prices and not at market prices, as recommended by BPM5.

The unit of account of the balance of payments is the U.S. dollar. Some components of the balance of payments, such as imports and exports and public external debt, are obtained from records posted in U.S. dollars. The BCCR requests information in the surveys in U.S. dollars or colones. Direct foreign investment companies or those subject to special regimes report the information in U.S. dollars. The information provided to the BCCR in colones is converted to U.S. dollars using the average exchange rate for the period.

Flows of the banks’ foreign financial assets and liabilities are obtained from the difference in stocks. The stock data, expressed in domestic currency, are converted to U.S. dollars using the buying rate prevailing on the date to which the balance sheet relates, which is the exchange rate used by the banks to convert stock data in U.S. dollars to domestic currency. The flows obtained are not adjusted to exclude valuation changes.

International reserve transactions are obtained as the difference in dollar-denominated balances of the BCCR. The flows obtained are adjusted to eliminate valuation changes.

2.4.2 Recording is done on an accrual basis

In general and consistent with the BPM5, the basis for recording transactions in goods is the date when ownership is transferred, the date when the service is rendered, or the date on which direct investment is made. Interest on public external debt is recorded on the due date; no adjustments are made to convert interest to accrued values.

2.4.3 Grossing/netting procedures are broadly consistent with international standards, guidelines, or agreed practices

Consistent with international standards, most components of the current account show gross values and most entries in the financial account are recorded in net values, i.e., each component is shown as a credit or as a debit. In addition, in the financial account assets and liabilities of the individual components are shown separately.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

It is considered that the source data used in compiling the balance of payments are adequate. The principal sources of information used in compiling the balance of payments are: (i) import and export statistics based on customs records and compiled by the DGA, including goods not subject to special regimes (customs territory) and those subject to the special inward processing regime (“maquiladoras”); (ii) PROCOMER, which provides statistics on foreign trade and overseas transactions of companies subject to the free zone regime (this entity also provides statistics on intended exports that are used to compile preliminary statistics on exports); (iii) INTEL, a company subject to the free zone regime, which reports directly to the BCCR for preparation of the national accounts and the balance of payments (the company does not fill out the survey); (iv) surveys of direct foreign investment companies and national companies that carry out foreign transactions, conducted by the BCCR to obtain information on transactions and stock data from direct investment, portfolio investment, and other investment, as well as other items in the balance of payments; (v) surveys conducted by the BCCR to obtain information on services and transfers; (vi) the Ministry of Commerce (COMEX) survey on intended foreign investments; (vii) the ICT, which prepares estimates on travel items; (viii) the Assets and Liabilities Management Division of the Operations Recording and Control Department of the BCCR and the Treasury Office of the MOF, which provide information on public external debt; (ix) information from the Monetary Programming Area on banks’ foreign assets and liabilities accounts; (x) foreign assets and liabilities accounts from the BCCR’s balance sheet, which provides information on international reserves and other overseas accounts of the BCCR; (xi) information from the Ministry of Finance on the government’s foreign transactions, e.g., transfers received from abroad, embassy expenses, etc.; (xii) the Stock Exchange, which records information on residents’ holdings of securities issued overseas and securities issued by the government overseas; and (xiii) information from the nonfinancial public enterprises’ external transactions on, e.g., transfers received from abroad, external expenses, banking deposits abroad (all these public enterprises have reported the data to the BCCR on a quarterly basis since 2001).

The ABP maintains a registry that includes all import and export companies, including companies engaged in processing goods imported for later export, which operate under the following special processing regimes: inward processing (“maquiladoras”) and free zones. In addition, there is a list of special respondents that are surveyed; this list includes companies that provide services, government entities, embassies, and international cooperation agencies.

The registry of companies is updated periodically with records of new companies carrying out transactions overseas provided by the Development Initiatives Coalition, COMEX, and PROCOMER. The annual sample is one of convenience, which includes all processing companies and the most representative companies carrying out transactions overseas. The sample is revised annually to eliminate companies that close and mergers and to incorporate new companies. Starting in 2000, quarterly surveys have been introduced based on a reduced sample of the principal companies, approximately 100 companies representative of the principal sectors. The INTEL company, the largest company in the processing industry, provides some monthly and quarterly data to the BCCR.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

In the last two years, substantial progress has been made in improving the source data in order to adjust them to the definitions required by BPM5 and to improve the coverage and classification of balance of payments statistics.

In 1999, the annual survey of businesses and institutions was modified and implemented to facilitate the compilation of components of the balance of payments and the international investment position in accordance with BPM5. In addition, the quarterly survey was drafted in 1999 and its implementation began in 2000. This survey includes less detail than the annual survey. Recently, developments include the Instructions for Compiling the Annual Survey Form for the Balance of Payments, which is quite detailed and includes concepts and definitions consistent with the BPM5. In addition, more concise Instructions for Completing the Quarterly Survey Form for the Balance of Payments have been prepared. The survey on intended direct investment prepared by COMEX is used to validate the information on direct investment obtained from the surveys.

The records on public external debt include bonds issued by the government abroad. The ABP adjusts this figure based on information on holdings of these securities by residents. This information is obtained from the custodial records maintained by the Stock Exchange.

The ABP compares the statistics on public external debt compiled by the Treasury Office (which includes the former DCP) with the corresponding statistics compiled by the BCCR’s Operations Recording and Control Department, in order to reconcile the figures and ensure that the coverage, the classification by sector and by instrument, and the records of transactions are consistent with international standards. In the DMFAS records, interest is recorded on the date of payment and late payments, if any, are recorded separately. In the balance of payments, interest is recorded on the due date.

The BCCR’s Department of Social Accounting compiles the statistics on foreign trade. The information on imports and exports is obtained from different sources. The DGA is the source of information for general exports and imports (customs territory) and those for inward processing (“maquiladoras”). PROCOMER is the principal source of statistical information on imports and exports in the free zones. The data on imports are reported in f.o.b. and c.i.f. values. The National Accounting Department compiles the import statistics on a c.i.f. basis, but the ABP compiles and records them in the balance of payments on an f.o.b. basis. The ABP adjusts the import data to include repairs on goods and goods procured in ports by carriers, which are not recorded in the customs statistics.

3.1.3 Source data are timely

In recent years, the timeliness of the data sources has improved. However, there are problems with the timeliness of responses to the surveys and with the figures on foreign trade, particularly exports.

The surveys indicate the due date for reporting, but a significant percentage of companies report the information with a considerable lag. ABP staff consistently follows up on those companies that have not reported data by the deadline through telephone calls or personal visits in the case of companies that have problems filling in the information.

The timeliness of the figures on foreign trade has improved considerably, but there are problems with the timeliness of the final figures on imports and exports, which are obtained after a lag of several months. The preliminary figures on imports exclude some contracts, particularly those of companies subject to special regimes, which report with a lag. The preliminary figures on exports are prepared with a lag of more than two months and are subject to constant revision. The provisional figures on exports are based on intended exports; these data are available 15 days after the end of the reference month. Therefore, trade data are subject to constant revisions.

3.2.1 Data compilation employs sound statistical techniques

Procedural manuals have been prepared that document the management, processing, and quality control of the annual and quarterly balance of payments surveys. These manuals are for internal use and are also used to train new staff. The surveys are designed to be easy to complete and process electronically, and a series of measures has been established to control quality and conduct appropriate and timely verification of the information provided in the surveys before and after their electronic processing. Procedures have been developed to verify coding and tabulation errors and additional tests when the information has been processed.

In addition, procedures have been developed to reconcile the statistics on public external debt compiled by the BCCR with those of the Treasury Office of the MOF. These are carried out electronically in special formats. The ABP has direct access to DMFAS, a system used by the BCCR to compile external debt statistics.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

Starting in 1999, import data have been gathered on a c.i.f. and f.o.b. basis. In addition, customs provides data on insurance. Previously, freight and insurance were estimated from the c.i.f. value to obtain the f.o.b. value.

Estimates for travel are compiled by the ICT in coordination with the ABP and the Institutional Sector Accounts Area of the BCCR. Information from the General Directorate of Migrants is used to determine the number of travelers visiting the country and traveling abroad. Average per capita expenditure and average stay of travelers are estimated on the basis of the surveys conducted by the ICT during the high and low seasons. Average expenditure is estimated for travelers coming from the United States, Canada, Europe, and the rest of the world.

3.3.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessments are monitored and made available to guide planning

The annual survey is a convenience sample including the most representative companies subject to the general regime and all companies operating under special regimes. Because it is a convenience sample, the sampling value cannot be determined. However, the representativeness of the results of the survey is evaluated by comparison with the principal components of the balance of payments. The quarterly surveys are also convenience samples that include the principal 100 companies that carry out operations with other countries. The results of the quarterly surveys are expanded to the universe using representativeness factors from earlier years.

The information reported by the companies is validated and respondents confirm transactions for high amounts.

3.3.2 Main intermediate results are validated against other information where applicable

Data on direct investment obtained from the surveys are compared with data on planned investment compiled by the Ministry of Commerce and PROCOMER.

Information published in the financial press is used to verify large value transactions for direct investment and other transactions, such as issues of securities abroad.

3.3.3 Statistical discrepancies and other potential indicators of statistical problems are investigated and made known to users

Statistical discrepancies are investigated but not made known to users.

3.4.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

No formal studies are conducted on revisions of the statistics.

4.1.1 Processes to monitor the relevance and practical utility of existing statistics in meeting users’ needs are in place

There is no formal process of consultation regarding users’ needs. In practice, many requests for information are received; these are handled primarily through e-mail and serve as an indicator of the users’ information requirements.

ABP staff members regularly participate in statistical meetings and seminars organized by the IMF, CEMLA, and the Central American Monetary Council. In 2000, a staff member from the area participated in a seminar on debt organized by the IMF.

4.2.1 Periodicity follows dissemination standards

The periodicity of the data is consistent with the Special Data Dissemination Standard (SDDS) and the data are published on the BCCR Internet site. The statistics on foreign trade are published monthly. The balance of payments data are released quarterly. International reserves are disseminated weekly and calculated daily. The data template on international reserves and foreign currency liquidity is prepared and distributed monthly.

4.2.2 Timeliness follows dissemination standards

The timeliness of the foreign trade, balance of payments, and international reserves statistics is consistent with the SDDS. The foreign trade statistics are disseminated 15 business days after the reference month. The quarterly balance of payments data are disseminated within 10 weeks of the reference quarter. The preliminary annual balance of payments data are disseminated within 10 weeks of the close of the reference period. The final balance of payments data are disseminated approximately eight months after the reference period. Weekly international reserves are published two days after the reference week. The data template on international reserves and foreign currency liquidity is released 15 business days after the reference month.

4.3.1 Statistics are consistent or reconcilable over a reasonable period of time

The historical balance of payments series from 1991 to 1997 were revised in 1997, upon conversion from BPM4 to the methodology recommended in BPM5. The revised series was published in the report, Costa Rica Balance of Payments 1992–1997, which included a methodological note explaining the principal changes in the classification and concepts as a result of adopting BPM5.

The historical balance of payments series for 1950–1990 was reclassified in accordance with the structure of BPM5, in summary fashion, and was published in 2000, in 50 Years of BCCR Statistics.

Any unanticipated change in economic trends is explained in the analytical text that is a part of the balance of payments publication.

4.3.2 Statistics are internally consistent (accounting identities observed)

The annual balance of payments statistics are consistent with the quarterly statistics. When the annual information is obtained, any discrepancy between the annual figures and the sum of the quarterly balances is adjusted, in the corresponding quarter if there is information on the date when the discrepancy occurred, or by distributing the difference among the quarters in accordance with other criteria.

The ABP continuously reviews methodology, procedures, and products in order to ensure the internal consistency of the statistics.

4.3.3 Statistics are consistent or reconcilable with those obtained from other sources and/or statistical frameworks⁴

The balance of payments statistics can be reconciled with the national accounts, monetary, and government finance statistics. Differences are identified and explained. The ABP and the National Accounts Area work together to reconcile the figures. There are some differences in the statistics prior to 1999, but as of 2000 the balance of payments statistics must be compatible with the national accounts statistics. Discrepancies for earlier years have been identified and explained. The greatest discrepancy in exports in 1991–99 is because an adjustment has been made in the national accounts to eliminate fictitious exports associated with improper use of the Tax Credit Certificate and this adjustment has not been made in the balance of payments.

The data on banking sector transactions in the balance of payments statistics are obtained from monetary statistics compiled by the Monetary Programming Area. The balance of payments statistics only include short-term foreign assets and liabilities of the banks; neither bank’s foreign holdings of equity securities nor bank’s long-term foreign liabilities are included.

Public external debt transactions are obtained from the debt records compiled by the BCCR, which are reconciled with the corresponding information compiled by the Treasury Office of the MOF. Public external debt transactions are reconciled with the corresponding stock data.

The data on financial account flows are reconciled with changes in the international investment position.

4.4.1 Revisions follow a regular, well-established and transparent schedule

There is no formal policy on revisions, but they are made on predetermined dates associated with the dates of BCCR publications. The preliminary annual data are compiled in February and published in the Half-Yearly Review. The Annual Report is published in March and the annual balance of payments data are revised on that date. The balance of payments figures are again revised in May and November when the monetary program is developed. Nearly final figures are compiled in August based on the information from the annual surveys.

Users are not informed of the revision policy, the schedule of revisions of the preliminary data, or the period to which they correspond. There are plans to include in the balance of payments publication the schedule and documentation for the revisions performed.

4.4.2 Preliminary data are clearly identified

In BCCR tables and documents the users are informed in a footnote that the data initially published are preliminary.

4.4.3 Studies and analyses of revisions are carried out periodically and made public

The factors that have contributed to the difference in the preliminary estimates are evaluated and explained to the BCCR authorities and to users who so request. No formal procedure has been established for disseminating studies and analyses of revisions.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The quarterly balance of payments statistics disseminated on the BCCR Internet site follow the structure and classification recommended in the BPM5.

The Half-Yearly Review includes half-yearly balance of payments data for the last two years and an analysis of recent trends in the external sector. This analysis describes the principal developments in the international environment and in the balance of payments of Costa Rica and an analysis of the factors that determined the current account, capital account, financial account, and net international reserve outcomes. A table of goods traded by six-month periods is included, which separately identifies regular exports and imports and exports and imports of goods for processing.

The Annual Report includes annual balance of payments data for the last two years and describes recent developments. Both publications classify the financial account by public and private sector and include indicators of external solvency.

The annual review Balance of Payments of Costa Rica publishes historical series covering six or seven years, following the detailed classification of the BPM5. Additional tables and graphs of the principal components of the balance of payments are included and there is a detailed analysis of developments in the balance of payments and the factors that determined the results of the principal components.

The publication Principal Statistics on International Transactions of Costa Rica disseminates time series of the balance of payments and supplementary tables of its principal components. An analysis of the balance of payments is not included.

5.1.2 Dissemination media and formats are adequate

The formats for reporting the balance of payments statistics are adequate. The statistics have been published for the first time on the BCCR Internet site. The Executive President regularly holds press conferences to discuss macroeconomic statistics.

Subsequent publications (see point above) provide more complete information and comments are included on the developments in the balance of payments. The 1950–1990 time series was published in 50 Years of Statistics in Costa Rica.

5.1.3 Statistics are released on a pre-announced schedule

The balance of payments statistics are released according to a pre-announced schedule, which is indicated on the BCCR Internet site.

5.1.4 Statistics are made available to all users at the same time

Statistics are released simultaneously to all interested parties on the BCCR Internet site on the date indicated in the dissemination schedule prepared according to the SDDS.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

The BCCR publishes quite detailed and complete information on the balance of payments and provides unpublished (nonconfidential) subaggregates to users who request them.

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from international standards are noted

In the report, Balance of Payments of Costa Rica 1992–1997, a methodological note was published explaining the concepts, definitions, classifications, and methodology of BPM5, which was adopted in 1998. However, the sources and statistical techniques used to compile the principal components of the balance of payments have not been documented. A report has been prepared for internal use on the response rates to the principal quarterly surveys used in compiling balance of payments statistics.

In preparation for subscription to the SDDS, the statistical methodologies used in compiling the balance of payments and foreign trade have been prepared but have not yet been disseminated to the public.

5.2.2 Levels of detail are adapted to the needs of the intended audience and the type of compilation

At the request of the users, and with approval from the BCCR authorities, specialized statistics from unpublished (but nonconfidential) data can be compiled and provided.

The pertinent balance of payments series are not published in seasonally adjusted values.

5.3.1 Contact person for each subject field is publicized

The BCCR provides assistance to users who request it. The ABP receives a large number of queries, generally by e-mail, regarding the balance of payments statistics and other statistics on the foreign sector and processes them as quickly as possible. The BCCR has designated a person to be the BCCR contact with the public. This person, whose name is given on the BCCR Internet site, channels requests for information to the corresponding areas. Assistance is provided in Spanish and English.

5.3.2 Catalogues of publications, documents, and other services, including information on any changes, are widely available

The catalogue of BCCR publications and documents is updated annually and released on the BCCR Internet site. Journals, bulletins, statements, and brochures are available to the public free of charge at the BCCR Documentation Center.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The government finance statistics of the central government, general government, and other levels of public sector aggregation are compiled by the Ministry of Finance (MOF) in accordance with the terms and conditions established in (i) Law 1279 of May 2,1951, (ii) Law 6821 of November 26,1982, and (iii) Decree 14375-H regulating Law 6821 dated June 6, 1983.⁵ Law 1279 governs the financial administration of the Republic and thus establishes the provisions applicable to the process, execution, and evaluation of the budget; Article 1 of Law 6821 creates the Budgetary Authority, the major objective of which is to formulate budgetary policy and ensure its fulfillment; finally, Article 8 of Decree 14375 establishes that the Technical Secretariat of the Budgetary Authority (STAP) is the competent entity responsible for compiling fiscal statistics.

Three divisions of the MOF are involved in the compilation and dissemination of government finance statistics: (i) the STAP, which compiles and disseminates government finance statistics on the operations of central government, general government, and other levels of the public sector, through the Fiscal Analysis and Monitoring Unit (UASF); (ii) the National Accounting Department, which compiles the cash flow of the central government; and (iii) the National Treasury, which compiles and disseminates statistics on central government debt, through the Registry and Control Section (SRC).

Under Article 14 of the BCCR’s Organic Law, which establishes that the BCCR is to publish a statistical summary of the country’s economic situation every month, this institution also compiles and disseminates some government finance statistics. The Government Finance and Fiscal Analysis Area compiles tables on the operations of central government and the public sector as a whole, using information from the MOF as the primary source. In addition, it uses its own data to produce a table of the domestic debt holdings of central government. Finally, the BCCR compiles and disseminates various tables on public external debt

It is important to note that Law 7839 of November 4, 1998 creates the SEN for the purpose of streamlining and coordinating statistical activity. The SEN will be made up of the centralized and decentralized public sector institutions and agencies that carry out statistical activities. The law governs the planning, production, development, and dissemination of statistical activity carried out by the entities comprising the SEN. Although the agencies that will make up the SEN have not yet been designated, it is assumed that the MOF and the BCCR will be two such entities and that thus the provisions of this law will apply to them. INEC is also created as the governing technical entity for national statistics and the coordinator of the SEN. It should be pointed out that to date no corresponding regulations have been issued to put the law into operation. Consequently, its impact on processing and dissemination is still limited.

0.1.2 Data sharing and coordination among data producing agencies are adequate

The STAP compiles the government finance statistics using information on budgetary execution as its primary data source. Consequently, the data published by the MOF following the provisions on the rendering of financial accounts contained in the document Submission of Draft Ordinary and Extraordinary Budget Law of the Republic for Fiscal Year 20XX, which are submitted to the Legislature each year, can be reconciled with government finance statistics.

As provided by Article 184 of the Constitution and Article 32 of the Organic Law of the CGR, the CGR is required to submit to the Legislature on May 1 of each year a report on developments in government finance. This report contains a statistical supplement with information on public sector revenue, expenditure, financing, and debt as a whole as well as their components (central government, general government, and public enterprises). For the most part, the information relates to the budget of the respective year and its execution. Despite the fact that the data sources of the CGR and the STAP are to a large extent the same, it is difficult to reconcile the statistics of both institutions primarily due to methodological differences (e.g., supplemental period of six months for budgetary execution) and differences in the presentation formats.

There are no formal mechanisms for collaboration between the STAP and the entities responsible for compiling data on financing for other sectors.⁶ However, there is a close working relationship between the MOF and the BCCR with respect to central government financing and debt figures, as the BCCR is the government’s financial agent and manages the systems that record domestic and external debt transactions. Furthermore, the MOF has direct access to those systems for informational purposes. Despite the close collaboration at the operational level, it should be mentioned that the government financing statistics are not reconciled with the monetary statistics. In more general terms, the government finance statistics are not reconciled with the statistics for other sectors (national accounts and balance of payments), except in the case of external debt.

There are no formal mechanisms to ensure good communications between the compilers of government finance statistics and the most important users of these statistics. However, the STAP maintains regular communications with the international and national organizations that use its statistics.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

Article 4, paragraph a) of the law creating the SEN stipulates that data obtained by law are strictly confidential. In addition, Article 98 of the autonomous regulations of services of the MOF provides for the confidentiality of the institution’s servers.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

Article 12 of the law creating the Budgetary Authority stipulates that public sector institutions must submit the economic and financial information that the Budgetary Authority deems necessary to carry out its duties. It should be noted that this provision also applies to institutions that are outside the sphere of the Budgetary Authority (local governments, universities, and banks). In addition, current regulations establish that public sector institutions and companies must designate two employees to be responsible for maintaining working and data relationships with the STAP.

Each year the STAP issues guidelines and procedures on budget formulation and execution. These guidelines and procedures establish the periods for submitting the respective information, and are issued by decree. Thus, for example, Decree 28640-H establishes that public entities must submit their budgets for the upcoming fiscal year by no later than September 30 of each year. It also establishes that public entities must submit the information on their budget execution each quarter on the 22^nd day of the month following the close of the reference period and must submit the budget settlement for the preceding year on February 16. In addition, it establishes that central government budgetary entities (and the entities included in the reduced nonfinancial public sector sample) must report monthly cash flow data, broken down for income, expenditure, and financing, 15 days after the end of the month.

0.2.1 Staff, financial, and computing resources are commensurate with institutional programs

The UASF has a coordinator and six analysts, while each of the three sectoral units of the Budgetary Authority that directly monitor the institutions within their sphere has an average of eight professionals. Each professional has a computer for carrying out his or her work, but the capacity of these computers is not always adequate. The staff assigned to the task of compiling and disseminating statistics on central government and general government operations seems sufficient. However, the computing resources supporting that task are inadequate and do not allow for greater automation of the process of compiling government finance statistics. This is the situation despite significant recent improvements in the availability of computers and computer systems based on external financial support. The SRC has five professionals, each with his/her own computer and computing support adequate to their functions.

There are no provisions for internal and external training in practical issues for the staff responsible for government finance statistics. The MOF has a training department, but the courses it gives and access to them are limited. Most of the training received by MOF employees involved in compiling government finance statistics takes the form of courses or advisory services provided through foreign financing, due to the limited availability of financial resources for training.

0.2.2 Measures to ensure efficient use of resources are implemented

The costs associated with the compilation of the government finance statistics are not calculated, and it is not known with any certainty whether available resources are used as efficiently as possible.

0.3.1 Processes are in place to focus on quality, to monitor the quality of the collection, processing, and dissemination of statistics, to acknowledge and deal with tradeoffs within quality, and to guide planning for existing and emerging needs

No formal procedures have been established to assess the quality of the statistics and information disseminated. However, there are continuous efforts to ensure the internal consistency of the data and to improve their general quality. For example, (i) the recently developed automated consolidation system represents a significant advance and (ii) the recently introduced securities management system (SATV) represents an important improvement in the recording and management of domestic debt. In addition, increased dissemination of fiscal statistics, primarily through the MOF website, encourages the compilers to improve the quality of their products.

In this regard, it should be emphasized that Costa Rica hopes to subscribe to the SDDS in the near future. For this reason, efforts are being made to satisfy SDDS requirements with respect to the periodicity, timeliness, and disaggregation level of fiscal indicators.⁷

1.1.1 Statistics are compiled on an impartial basis

The compilation and dissemination of government finance statistics are not supported by specific laws or formal safeguards that grant professional independence to the areas responsible for the government finance statistics. However, there are informal safeguards ensuring that independence because the MOF has a culture of political noninterference in technical tasks.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

Statistical techniques are free of all political influence over the selection of data and the most appropriate techniques for compiling the government finance statistics; changes in statistical procedures reflect technical and not political criteria.

The 1986 GFSM methodology has not been legally adopted but constitutes the basis for compiling government finance statistics in practice.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The compilers of government finance statistics are empowered to provide specialized advice on technical aspects of government finance statistics (but not on policy aspects). In addition, errors in interpreting government finance statistics may be discussed publicly.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The laws and regulations governing the collection and dissemination of government finance statistics are available to the public in electronic form and free of charge at the Virtual Knowledge Center of the MOF. In addition, there are no restrictions on public access to government finance statistics. Further, Article 30 of the Constitution stipulates that: “Free access to administrative departments for purposes of [obtaining] information on matters of public interest is guaranteed.”

1.2.2 Internal government access to statistics prior to their release is publicly identified

No government employee outside the MOF has access to statistics on central government and general government operations prior to their release. These statistics are disseminated simultaneously to the public at large through the MOF website. In addition, the central government’s monthly cash flow is disclosed in a press conference. The statistics on central government debt are sent by e-mail to some government entities before they are released to the public (as a courtesy), and this is not indicated in the respective publication.

The procedures that authorize the dissemination of government finance statistics are considered internal rules and are not made public. The dissemination of government finance statistics is authorized by the Vice-Minister for Disbursements, in the case of data on central government and general government operations, and by the National Treasurer, in the case of data on central government debt.

1.2.3 Products of statistical agencies/units are clearly identified as such

NOTE: In the case of government finance statistics this applies only if a statistical entity/unit is responsible for collecting government finance statistics.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

Government finance statistics users are given no advance notice in government finance statistics publications and/or special publications of methodological changes or other changes that could significantly affect the data published. Changes in methodology are disclosed to the public at the same time as the data affected are published. Any change that involves a break in the chronological series is indicated in the corresponding publication.

1.3.1 Guidelines for staff behavior are clear and publicized

The collection and dissemination of government finance statistics are not governed by a published code of conduct that ensures that the data are collected and the results presented on the basis of strictly professional considerations. However, there are internal MOF regulations, dated July 1996, that include a section on the ethical principles of public servants, among them the duty to be objective, i.e., the duty to issue objective judgments free of third-party influences.

Some efforts are being made to promote a spirit of professionalism, e.g., by organizing discussions on government finance statistics methodology. There are no systematic efforts to take advantage of methodological and administrative advances of other outside entities.

2.1.1 The overall structure in terms of concepts and definitions follows international standards, guidelines, or agreed practices

The collection and dissemination of statistics on central government and general government operations are based on IMF recommendations in that they follow the 1986 GFSM guidelines. There are plans to adopt a process to transition from the government finance statistics framework based on 1986 GFSM to the revised government finance statistics system based on 2001 GFSM, when it takes effect. In this regard, it should be pointed out that a factor that will affect the transition is the timeliness with which the Spanish version of that manual is published.

The collection and dissemination of government finance statistics on central government debt are not based on 1986 GFSM recommendations but rather on national concepts and definitions. However, the national concepts and definitions are not significantly different from the international recommendations.

2.2.1 The scope is broadly consistent with international standards, guidelines, or agreed practices

Government finance statistics are compiled and disseminated at the budgetary central government, general government, nonfinancial public sector, and total public sector levels. In addition, government finance statistics for the reduced nonfinancial public sector (a sample of nonfinancial public sector entities) are compiled and disseminated. It should be pointed out that government finance statistics are also collected at the level of consolidated central government, but only for publication in the IMF’s Government Finance Statistics Yearbook (GFSY), i.e., government finance statistics are not regularly compiled with the central government coverage recommended by the 1986 GFSM. In this regard, it should be mentioned that regularly compiled statistics on general government include all entities of the consolidated central government. Finally, the government debt figures are compiled and disseminated at the budgetary central government level.

There is information on financial flows and holdings of balances among the units of government, which makes it possible to consolidate the public sector at various levels of aggregation. However, the consolidation system used only makes it possible to provide consolidated figures for the corresponding level of aggregation, but not for its component subsectors (although each subsector is internally consolidated). For example, there are consolidated figures for the general government, but not for the budgetary central government, the attached entities, public utilities, and local governments, the figures for which are also shown in the general government table but not consolidated among themselves.

Government finance statistics are final when first published. Preliminary government finance statistics based on partial or preliminary information are not published.

All of the central government’s most important activities are the responsibility of units in that sector. Consequently, supplementary data on quasi-fiscal activities are not provided. In addition, as indicated earlier, government finance statistics are compiled for broader definitions of government.

The STAP publishes four tables on central government operations on the MOF website:

• (i) summarized monthly data on budgetary central government operations,

• (ii) summarized cumulative monthly data on budgetary central government operations,

• (iii) detailed annual data on budgetary central government operations, and

• (iv) annual data on recognized expenditures of the budgetary central government.

The first two tables are prepared by the National Accounting Department using a strict cash flow criterion and contain information on revenues, expenditure, and financing, with limited breakdown of these aggregates. The third table is prepared by the UASF, with the same institutional coverage as the other two tables, but with a greater level of detail on revenues, expenditure, and financing. In addition, this third table uses recognized expenditure as the criterion for expenditure and includes some transactions that do not involve movements of cash (public employee deductions for income tax, pension fund contributions, and other deductions from the gross salaries of public employees). Finally, the fourth table, also prepared by the UASF, contains information on recognized expenditure and on debt amortization.

The STAP also publishes on the MOF website four tables prepared by the UASF with annual data on general government, the nonfinancial public sector, total public sector, and reduced nonfinancial public sector. The level of detail on revenue, expenditure, and financing in all these tables is the same as in the third table on central government operations. The breakdown of revenue in these tables is similar to that recommended by 1986 GFSM in table A (Government revenues and grants), although the presentation of nontax revenues is different. The breakdown of expenditure is similar to that recommended by 1986 GFSM in table C (Economic classification of expenditure and net lending minus repayments to government). The breakdown of financing combines elements from tables D (Financing by type of debt holder) and E (Financing by type of debt instrument) of 1986 GFSM, although with a different presentation and without providing all the recommended detail.

Finally, the STAP publishes on the MOF Internet a table prepared by the UASF site with annual data on total public sector operational expenditures, following the United Nations Classification of Functions of Government (COFOG). A table on budgetary central government operating expenditure that follows its own classification criterion is produced as well, but not published.

In addition, the STAP has monthly data for the reduced nonfinancial public sector; these data are not published but are submitted to the Western Hemisphere Department (WHD) without consolidation.

The National Treasury publishes on the MOF website eight tables on central government debt based on nominal values. These tables present the information on total debt broken down into domestic and external debt, domestic debt broken down by type of security, maturity, and term, external debt broken down by type of creditor and maturity, and contingent external debt by type of creditor (there is no contingent domestic debt). The breakdown of central government debt combines elements of tables F (Debt by type of debt holder) and G (Debt by type of debt instrument) of 1986 GFSM, although with a different presentation and without providing all the recommended detail.

The BCCR publishes on its website four tables on budgetary central government operations: (i) monthly cash flow prepared by the National Accounting Department; (ii) a table with monthly data on revenue, expenditure, and financing, based on cash flow but with adjustments similar to those made by the STAP in its table with annual data for the budgetary central government; (iii) a table on the growth and structure of total revenue; and (iv) a table on the growth and structure of total expenditure. In addition, the BCCR publishes on its website a table on the revenue, expenditure, and financing of the nonfinancial public sector, a table on the financial performance of the entire public sector, and a table on domestic debt by holder based on transaction value. In addition, the BCCR publishes similar but more concise information in its Weekly Review and Annual Report. Finally, the BCCR publishes a weekly External Debt Bulletin, with information on total public external debt, with various breakdowns.

2.3.1 Classification/sectorization systems used are broadly consistent with international standards, guidelines, or agreed practices

Institutional sectors are defined according to 1986 GFSM, except in the case of central government, for which the STAP uses the concept of budgetary central government and not consolidated central government. Revenue, expenditure, and financing are generally classified according to 1986 GFSM methodology. However, as indicated earlier, the presentation of these aggregates differs in some respects from what the manual recommends. The sectors or subsectors for which government finance statistics are prepared are defined according to 1986 GFSM, and separate statistics are available for each of them. Separate government finance statistics are provided for central government, local governments, and consolidated general government (with the limitation of the consolidation system indicated above with respect to the subsectors that make up a specific level of aggregation).

Finally, financing and domestic and external debt are classified according to where they are placed on the primary market and not according to the criterion of residence. No surveys are conducted to determine the residence of debt holders.

2.4.1 Market prices are used to value flows and stocks

Foreign currency transactions are converted by applying the average daily exchange rate in the case of monthly budgetary central government statistics and the annual average exchange rate in the case of annual general government statistics. In the case of central government debt statistics, the foreign currency balances are converted by applying the selling rate of exchange as of the close of the period.

2.4.2 Accounts are prepared in cash values

Revenue, expenditure, and financing are recorded on a cash basis for the monthly data on central government operations. For the annual central government and general government data, revenue and financing are recorded on a cash basis, but expenditure is recorded on a recognized basis. It should be pointed out that the lag between recognition of an expenditure and the corresponding cash flow has been significantly reduced with the introduction of the electronic payments systems (SINPE). Finally, it should be noted that a transition to accrual-based accounting or budgeting is not currently under consideration.

2.4.3 Grossing/netting procedures are broadly consistent with international standards, guidelines, or agreed practices

As indicated earlier, the monthly data on central government exclude some transactions that do not involve cash movements. These relate to transactions that are recorded in net terms, such as wage payments to civil servants (minus wage deductions for various items) and procurement of goods and services (including legal withholdings). 1986 GFSM recommends that revenue and expenditure be recorded in gross terms.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

The data sources for preparing government finance statistics for the budgetary central government come primarily from the administrative systems for monitoring budgetary performance and encompass all entities covered by the budget law. The principal source of information for monthly government finance statistics (cash flow) are (i) daily/biweekly reports on income and outlays in the general fund through the SINPE and the Tax Reporting and Payment System (ILI) and (ii) the special fund account statements issued by the BCCR. The principal sources of information for annual government finance statistics are (i) revenue settlement, (ii) outlay settlement, (iii) settlement of commitments (outlays charged to the prior year’s budget), (iv) information from the National Treasury for interest, disbursements, and repayments of external debt, and (v) information from the BCCR for interest, disbursements, and repayment of domestic debt and other domestic financing.

The principal data sources for general government, which includes attached entities, public utilities, and local governments, as well as for the nonfinancial public sector and the entire public sector, are (i) the settlement of the quarterly budget (ii) the settlement of biannual commitments, (iii) the monthly financial investment report, and (iv) the Annual Report of the CGR (for local governments). It should be noted that for all institutions outside the sphere of the Budgetary Authority (local governments and universities) no information is available on financing. Additionally, for the reduced nonfinancial public sector, the information is available but is received as a whole on a monthly basis. This information is used to prepare annual but not monthly government finance statistics.

Regarding the data sources for budgetary central government domestic and external debt, the BCCR manages the recording systems for its own activities, the Securities Management System (SATV) for domestic debt, and the Financial Analysis and Debt Management System (DMFAS) for external debt. It should be noted that these systems, particularly the SATV, are used not only for recording purposes but for management purposes as well.

The STAP has the information needed to carry out the consolidation of the data on the central government, general government, and other public sector levels of aggregation. This consolidation is carried out only for annual data, with the limitation indicated earlier of the unavailability of consolidated data for the subsectors that make up the aggregate government level.

The data published by the STAP are final and based on complete coverage and thus do not include estimated data.

As indicated earlier, the data on local governments come from the Annual Report published by the CGR, which is submitted to the Legislature on May 1 of each year. There is detailed information on revenues and outlays but no information on financing. Surveys are not used to compile the local government data in the government finance statistics, as the CGR receives reports from all local governments.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

The budgetary ledger accounts are consistent with government finance statistics categories. Central government statistics are prepared by identifying the items corresponding to the accounting of budgetary execution. For the remaining public sector entities, there are no ledger accounts but rather only various classifiers developed by the CGR for income and outlays (the CGR plans to introduce uniform ledger accounts in 2003).

Budgetary management systems do not include provisions that would allow government finance statistics items to be automatically derived from budget items.

The time of recording government finance statistics in the case of revenue and financing is on a cash basis, whereas in the case of expenditures it is on a recognized basis, i.e., when payment orders are issued. Currently, as a result of the automation of the payments system, the differences between cash basis and recognized basis for expenditures are insignificant.

As indicated earlier, foreign currency transactions are converted at the daily average exchange rate in the case of central government statistics and at the annual average rate in the case of general government statistics. In the case of central government debt statistics, foreign currency balances are converted at the selling rate at the close of the period.

3.1.3 Source data are timely

Accounting or budgetary management systems provide complete data on the budgetary sector with a lag of no more than 20 days after the close of the reference period, except for the breakdown of domestic financing, which is not yet available.

There are annual data on extra-budgetary entities (including social security funds) that are sufficiently up to date to be used in compiling government finance statistics. There are also monthly data on the extra-budgetary entities that are part of the reduced nonfinancial public sector, but these are not used to compile government finance statistics. Finally, data are provided from local governments for inclusion in the annual government finance statistics collection cycles.

3.2.1 Data compilation employs sound statistical techniques

The central government and general government statistics are final when first published; thus, no estimates are done. Local government statistics are based on complete information; thus, no surveys are conducted.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The adjustments made to eliminate the discrepancy between the deficit/surplus and budgetary central government financing (and other public sector levels of aggregation) do not involve any strict statistical method but rather are simply introduced in the cash balance (budgetary central government) or in the heading “residual financing” (other levels of aggregation). In the case of entities outside the budgetary sphere, where no information on financing is available, an amount equal to the deficit/surplus (with opposite sign) is included in the residual heading.

3.3.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessments are monitored and made available to guide planning

There are no problems with omissions in the coverage or availability of information. As indicated earlier, government finance statistics are compiled on the basis of complete information. The dates established for submitting the data allow sufficient time for the entities to submit reliable data, but there is no formal procedure for assessing the accuracy of the data.

Government finance statistics data are final when first published; no preliminary data are published. Corrections are made when errors are detected and these are generally related to the classification of certain transactions.

3.3.2 Main intermediate results are validated against other data where applicable

The information sources of the STAP and the CGR are the same (budget execution). Thus, their figures can be reconciled in principle, but this is not done in practice. The BCCR uses information from the STAP for its tables on central government and general government operations, and the information can be reconciled. The debt figures come from the systems managed by the BCCR and are regularly reconciled between that institution and the National Treasury.

3.3.3 Statistical discrepancies and other potential indicators of statistical problems are investigated and made available to guide users

As indicated earlier, there are discrepancies between the budgetary central government deficit/surplus and financing (and other public sector levels of aggregation). These discrepancies have been jointly analyzed in the past by the MOF and the BCCR at the budgetary central government and reduced nonfinancial public sector levels, with the achievement of a satisfactory reconciliation of the discrepancies (the residual discrepancy is insignificant). At present, this joint analytical work has been discontinued, although the BCCR continues doing it for its own purposes.

Currently, the UASF adjusts the discrepancy at the budgetary central government level in the cash balance. This discrepancy was investigated in the past by another unit of the MOF (whose functions have been discontinued). The UASF does investigate the discrepancy at the reduced nonfinancial public sector level; the remaining discrepancy at this level and other public sector levels of aggregation is adjusted simply by introducing a residual item in financing. The public is not given any information on the discrepancy.

3.4.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

No explicit studies or analyses of revisions are carried out, although there are mechanisms within the collecting offices that make it possible to detect errors, particularly in terms of internal consistency. The errors detected are usually corrected prior to publication of the data.

4.1.1 Processes to monitor the relevance and practical utility of existing statistics in meeting users’ needs are in place

The collection of government finance statistics follows the same timeframe used in preparing and monitoring the budget, and government finance statistics are used in the process of formulating and evaluating fiscal policy. An important element in the formulation of the budget that is submitted each September is the government finance statistics for the previous period. Information is released on budgetary execution on an annual basis only, and government finance statistics for general government (and other public sector levels of aggregation) are prepared annually and released with a lag of six months. Users are especially interested in the functional classification table of public sector expenditure for assessing the extent to which the government is achieving the objectives of its social and fiscal policies. There are no formal procedures for consulting users.

4.2.1 Periodicity follows dissemination standards

Data are published on central government operations on a monthly and annual basis, on general government operations on an annual basis, and on central government debt on a monthly basis. Thus, the requirements of the SDDS, which are monthly dissemination for central government operations, annual dissemination for general government operations and quarterly dissemination for central government debt, are observed.

4.2.2 Timeliness follows dissemination standards

The timeliness of the monthly data on central government operations is no more than 20 days after the reference period, except for the breakdown between bank and nonbank domestic financing, which is currently not available, but is expected to be available soon with a six-week lag (one month according to the SDDS).⁸

The timeliness of the quarterly data on central government debt is three weeks after the reference period (two quarters according to the SDDS).

The timeliness of the annual data on general government operations is six months after the reference period, although it should be mentioned that the data for 2000 were published on July 12 (six months according to the SDDS).

4.3.1 Statistics are consistent or reconcilable over a reasonable period of time

Discrepancies with respect to expected trends in the revenue and expenditure time series are tied to underlying fiscal policy events or economic developments. To date, no case has occurred in which it has been necessary to adjust the historical series due to discontinuities resulting from methodological changes or changes in statistical systems.

4.3.2 Statistics are internally consistent (e.g., accounting identities observed)

The overall deficit/surplus is equal to total financing with the opposite sign. This consistency is achieved by including a “residual” line in financing equal to the difference between the overall deficit/surplus and financing; no note is published to this effect. The overall deficit/surplus is equal to total revenues minus total expenditure and net lending minis repayments.

There are difficulties in closing the below-the-line central government operations table, as well as in reconciling figures for financing with the monetary statistics. These difficulties are due to a variety of factors such as differences in coverage, valuation, classification, and time of recording. The BCCR makes an effort to achieve closing and to reconcile the financing figures annually; the remaining discrepancy is insignificant.

The sum of the components of total revenue, total expenditure, and total financing is equal to the corresponding aggregates. No summary table of government operations is prepared. No procedures are provided to compare the monthly data with the annual data; the annual data are equal to the cumulative sum of the months.

The sum of the government finance statistics aggregates for the different levels that make up the consolidated government or other subsectors of government finance statistics is equal to the aggregates of the consolidated subsectors minus the consolidation adjustments.

4.3.3 Statistics are consistent or reconcilable with those obtained from other data sources and/or statistical frameworks

Government finance statistics are compatible with the related national accounts aggregates, as both use the same sources of information. However, no reconciliation of the principal aggregates is done. With respect to the compatibility of government finance statistics with the monetary accounts and balance of payments data, formerly joint efforts were made by the MOF and the BCCR to reconcile these figures but such efforts were discontinued some time ago.⁹

Finally, with respect to the central government debt data, the figures compiled by the MOF and the BCCR are compatible and reconcilable. Furthermore, they are regularly reconciled.

4.4.1 Revisions follow a regular, well established, and transparent schedule

The data are final when first disseminated. If an error is later detected, it is corrected but the correction is not communicated to the users, it is simply carried to the data. Consequently, revisions do not follow a predictable cycle nor are they carried out according to a revisions schedule.

4.4.2 Preliminary data are clearly identified

As indicated in the previous section, the data are final when released for the first time. Consequently, no preliminary data are published.

4.4.3 Studies and analyses of revisions are carried out periodically and made public

As indicated in the previous section, the data are final when released for the first time. Consequently no studies and analyses are conducted of the revisions, no time series are provided on revisions of the data, and no analysis is carried out to compare the preliminary data with final data for the principal government finance statistics aggregates from earlier periods.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

Government finance statistics are compiled and disseminated generally following 1986 GFSM recommendations, which facilitates correct interpretation of data on the part of users. However, the presentation of the tables is not the best, as only a single extensive table is presented with all the information on revenue, expenditure, and financing for each level of aggregation in the public sector. It would be useful to separate this table into various tables (e.g., table of revenues, table of expenditures, financing table) and introduce a table that contains the principal government finance statistics aggregates. It would also be useful to publish time series so that comparisons can be made with prior years. Additionally, the information submitted does not include explanatory notes.

5.1.2 Dissemination media and formats are adequate

The government finance statistics dissemination policy is reflected in the advance release schedule that the MOF publishes on its website. The information is published almost entirely by electronic means (http://www.hacienda.go.cr.), although there is an annual publication on government finance statistics called Consolidation of Public Sector Revenue, Expenditure, and Financing, which is available to the public but has limited distribution. The BCCR also disseminates fiscal statistics on its website and in publications (Weekly Review and Annual Report). Finally, the CGR also discloses fiscal statistics on its Internet site and in its Annual Report.

5.1.3 Statistics are released on a pre-announced schedule

An advance dissemination schedule that provides approximate dates four months in advance and that is updated monthly is disseminated on the MOF website. In addition, this information is provided through a press bulletin. At present no information is given regarding the exact date of publication.

5.1.4 Statistics are made available to all users at the same time

The data on central government and general government operations are disseminated to all users at the same time, with no user getting preferential treatment. The debt statistics are sent (as a courtesy) by e-mail to some public agencies prior to being released to the public.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

Government finance statistics publications do not include information on the availability of unpublished data or the conditions under which they can be obtained. Users who require information in addition to published data may request it from several institutions; such information is generally provided as long as it is available or its preparation does not involve a significant drain on the institution’s staff. Currently, different international organizations periodically receive unpublished information.

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available and differences from international standards are annotated

The concepts, sources, and methods used in compiling government finance statistics are not published. Questions about methodology can be directed to the person designated as a contact. Bridging tables indicating the links between source data and government finance statistics are not provided.

5.2.2 Different levels of detail are provided depending on intended audience and the type of collection

Technical descriptions of the concepts, sources, and methods used in compiling government finance statistics are not published. Questions regarding methodology may be directed to the person designated as the contact.

5.3.1 Contact person for each subject field is publicized

The name of the person to whom questions about government finance statistics can be directed is disclosed on the MOF website. Backup and advisory services are provided on an ad hoc basis. When government finance statistics are used improperly, the corresponding authority (Vice-Minister for Disbursements) clarifies the matter.

5.3.2 Catalogues of publications, documents, and other services, including information on any charges, are widely available

There are no catalogues of publications or documents. The MOF does not charge for its publications or services.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The Costa Rican banking system is governed by the Organic Law of the National Banking System (Law 1644 of September 26, 1953 as amended), the Organic Law of the BCCR (Law 7558 of November 3, 1995), and other applicable laws, as well as by the respective regulations.

One of the functions of the BCCR is the collection, compilation, and dissemination of monetary and financial statistics. The legal basis for the compilation of such statistics is fundamentally based on the Organic Law of the BCCR. According to Article 14, the BCCR must publish (i) within the first 15 business days of each month, a statement of the financial condition; (ii) monthly, a statistical summary of the country’s economic data, including information on money and credit; and (iii) a biannual report on economic developments. Article 16 establishes that the BCCR shall publish its Annual Report, presenting the economic and financial situation of the preceding year, within the first three months of each year.

Chapter IV of the BCCR’s Organic Law establishes the organization and functions of SUGEF as the most decentralized body of the BCCR; thus, SUGEF is considered an entity attached to the BCCR. According to Article 171 of the law Governing the Securities Market (Law 7732 of January 27, 1998 as amended), the National Council for Supervision of the Financial System must approve the annual operating plan, the budget, its amendments and the budgetary execution of SUGEF within the overall limit set by the Board of Directors of the BCCR and submit them to the Office of the General Comptroller of the Republic for final approval. According to Article 174 of that same law, 80 percent of the budget of SUGEF must be financed with resources from the BCCR and 20 percent from the compulsory contributions of the supervised entities. The principal functions of SUGEF include (i) supervising and auditing the activities of the entities engaged in “financial intermediation,” for the purpose of monitoring the stability, soundness, and efficient operation of the national financial system; (ii) issuing general accounting standards for recording the operations of audited entities; and (iii) issuing accounts manuals that take into account the information needs of the BCCR.

0.1.2 Data sharing and coordination among data producing agencies are adequate

The BCCR has adopted provisions to ensure the flow of data between the Accounting Department (DC) and the Policy and Financial Programming Section (SPPF) of the Monetary Department (DM) through automated systems that allow authorized DM staff to access the

Integrated Financial Accounting System (SIFC). The SIFC has historical and current (including daily) information on all the BCCR operations, facilitating the compilation of monetary statistics for the BCCR.

With respect to other depository corporations (ODCs), since January 2001 there has been a formal agreement to ensure the flow of information from SUGEF to the BCCR. In addition, SUGEF has authorized the ODCs to submit directly to the BCCR the additional information it needs to compile monetary statistics. In order to strengthen the cooperation between SUGEF and the BCCR, in January 2001 the BCCR Board of Directors reiterated to SUGEF the need for timely submission of financial/accounting data for the ODCs to the BCCR.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

The confidentiality of the data that ODCs provide to SUGEF is established in Article 132 of the BCCR’s Organic Law. The law prohibits any individual or legal entity that provides services to SUGEF from disclosing information relating to the documents, reports, or operations of the supervised corporations. However, the following data are exempted from this prohibition: (i) data that SUGEF must provide to the public in certain cases and according to the procedures expressly provided in this law; (ii) data requested by order of the competent judicial authority; and (iii) data requested by the BCCR Board of Directors with the agreement of at least five of its members and in order to carry out the BCCR’s legal functions. In such cases, the members of the Board of Directors and other BCCR officials will be subject to the prohibition indicated above. Consequently, the BCCR maintains total confidentiality regarding the individual information it receives from SUGEF and directly from the ODCs. In effect, the monetary statistics that the BCCR publishes do not include data on specific financial corporations or transactions that could put at risk the confidentiality of the information. It should be mentioned that SUGEF publishes on its Internet site the financial statements for each ODC under its supervision, with some degree of detail.

There are no computer programs especially designed to prevent the dissemination of confidential data The confidentiality of the data depends to a large extent on the professional ethics of BCCR employees. Article 41 of the BCCR’s Organic Law establishes that the regulations of the BCCR shall spell out the duties and powers of its employees.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The legal provisions that ensure compulsory reporting of data from the ODCs to SUGEF are established in the BCCR’s Organic Law and in the Organic Law of the National Banking System. Financial corporations subject to SUGEF supervision are required to submit information to SUGEF based on the provisions of Article 119 of the BCCR’s Organic Law. Article 17 of the Organic Law of the National Banking System establishes the requirement that the ODCs submit to SUGEF all the balance sheets, statements, and statistical tables that SUGEF requests, at the time and in the manner that it determines. In addition, the law establishes that SUGEF shall have free access to all books, documents, and files of the ODCs that SUGEF considers necessary to carry out its surveillance and supervisory functions as best as possible.

SUGEF is empowered to supervise those entities that carry out “financial intermediation.” Article 116 of the Organic Law of the BCCR defines “financial intermediation” as the customary raising of financial resources in order to allocate them, on behalf of and at the risk of the intermediary, to any form of credit or investment in securities. Article 117 of the same law provides for the possibility of exempting from financial supervision savings and loan cooperatives, cooperative associations, and any other similar entity on the basis of the volume of their assets, number of members, or when they carry out operations with a closed group of members. With respect to these criteria, SUGEF exempted from financial supervision all cooperative associations and most of the savings and loan cooperatives. Financial corporations not supervised by SUGEF are not required to send information to SUGEF, except in the case of savings and loan cooperatives, which must submit summary audited reports annually in order to verify the parameters used to exempt them from supervision (based on amount of assets).

0.2.1 Staff, financial, and computing resources are adequate for the performance of institutional functions

Monetary statistics are prepared by the SPPF, which has 18 professional staff, nine of whom spend a large part of their time on the preparation of monthly and weekly monetary statistics. Each employee has a personal computer and access to the Internet.

The employees responsible for compiling monetary statistics must have at least a bachelor’s degree in economics. In addition, it is estimated that an average of one year of experience is required to acquire good knowledge of concepts and established practices. Although the employees specialize in specific areas, on certain occasions they also conduct special studies on monetary policy. Efforts are currently being made to rotate duties, so that a good number of employees will know about the various activities associated with the compilation of monetary statistics.

Although there are no established in-house programs to provide practical training on methodology, the BCCR supports its employees’ participation in training courses on the methodology recommended in the MFSM. The IMF offers these courses in cooperation with CEMLA. In the last three years, the SPPF has sent two of its employees to each of these courses and expects to continue its participation in the future.

The number of staff available for the preparation of monetary statistics is considered adequate for carrying out the current work program. However, in view of the BCCR’s great interest in adopting the MFSM methodology, particularly in expanding the coverage of monetary statistics and applying the sectorization and classification of accounts proposed in the MFSM, future needs must be evaluated in terms of the number of employees as well as their technical capability.

The computer equipment and programs used for the compilation of BCCR data were upgraded in 1999; making it possible to process BCCR data in an efficient and timely manner. Currently, three of the 18 personal computers of the SPPF have a module installed that allows for direct access to consult the accounting database. This database feeds the monetary database (MONET) from which the BCCR’s monetary accounts are derived.

With respect to ODCs, the BCCR is working on implementing an automated data entry system (entry is currently done manually). This effort was suspended in 1999 due to the actions taken by the ODCs to address potential problems related to the Y2K phenomenon. The work currently focuses on adapting the previously designed technological platform to recent technological advances and on redefining the data reporting formats to adopt the economic sectorization and classification of accounts proposed in the MFSM.

0.2.2 Measures to ensure efficient use of resources are implemented

In general, all programs in the BCCR are subject to an individualized budget with development goals and quarterly assessments. The goals are formulated at the departmental level within the context of the BCCR operating plan. The assessments are conducted in terms of results, based on a monthly check of the cost incurred for each of the development goals.

0.3.1 Processes are in place to focus on quality, to monitor the quality of the collection, processing, and dissemination of statistics, to acknowledge and deal with tradeoffs within quality, and to guide planning for existing and emerging needs

The BCCR recognizes that the quality of the statistics is essential for accurate analysis of the monetary situation and for formulating and implementing monetary policy. In addition, quality fosters users’ confidence in the data. The quality of the monetary statistics is constantly assessed by BCCR employees who compile them and by the analysts who prepare the periodic reports of the SPPF. In addition, there is implicit evaluation when reconciling statistics on certain variables compiled by other units of the DM.

The BCCR is launching an integral quality management program that provides for training courses for BCCR employees in order to improve the quality of its products. This program was organized by the BCCR Human Resources Department at the request of higher BCCR officials and is offered through outside consultants. The course lasts two months and consists of morning classes two days per week. In March 2001, the section chiefs of the Economics Division (DE) participated in this training program, during which the tasks of the DE were analyzed with respect to preparation of the monetary program and recommendations were formulated. There are plans to extend activities of this type to the rest of the staff.

There is no committee within or outside the BCCR that examines the accuracy of data or their adherence to international statistical standards. However, the Internal Audit Office of the BCCR, under Article 35 of its Organic Law, is responsible for ensuring the performance, sufficiency, and validity of the internal control system established by BCCR management. In particular, Article 38 establishes that the Internal Auditor must endorse the financial statements of the BCCR and ensure that all publications stipulated by law are produced. To this end, the Internal Audit Office evaluates the different procedures that exist in the BCCR. During 2001, the Internal Audit Office will evaluate various processes of the DM, but to date this does not include compilation of the monetary statistics. However, this activity could be subject to assessment at any time. It should be noted that the processes of the DM will be clearly defined when the recommendations from the “Total Quality Management” project are implemented.

The BCCR does not conduct periodic surveys of users, nor does it have systematic mechanisms to gather comments from users and determine their needs for monetary statistics. However, users can send queries and comments to the BCCR, through the BCCR Internet site, to the BCCR Information Center (CDI) or directly to the DM.

1.1.1 Statistics are compiled on an impartial basis

The conditions and provisions governing the BCCR are sufficient to ensure its independence and the autonomy of the staff devoted to compiling monetary statistics. The provisions emphasize the professional autonomy of BCCR staff; thus, the appointment of employees is autonomous. The Organic Law of the BCCR prohibits undue influence being exercised over the content or dissemination of statistical information. In addition, the BCCR’s Internal Labor regulations establish sanctions for employees who misuse the information to which they have access.

The BCCR promotes professionalism by encouraging staff to conduct research projects and methodological studies. In addition, the BCCR supports staff participation in international seminars and courses on monetary statistics. However, at present the preparation of weekly and monthly monetary statistics requires significant resources to verify the quality of the primary data and the subsequent compilation and analysis of the data. It is expected that when the entry, validation, and verification of data for the ODCs is automated it will be possible to provide more opportunities for the professional development of the staff devoted to these statistical tasks.

1.1.2. Choices of sources and statistical techniques are informed solely by statistical considerations

The choice of sources, methods, and definitions for compiling monetary statistics is based on statistical considerations. The sectoral balances of the ODCs are obtained from information provided by the ODCs to SUGEF for purposes of banking supervision. These balances are supplemented with detailed information sent directly to the BCCR by the ODCs. The current methodology is not entirely consistent with the MFSM, but a work plan is being implemented to adopt the sectorization and classification principles proposed in the MFSM. The BCCR seeks to encourage its staff to explain and document in manuals for internal use the methodology used in preparing monetary statistics.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

Generally there is no erroneous interpretation or misuse of monetary statistics by users. When there are significant errors in interpretation in the media, senior BCCR officials usually make clarifications.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The public has access to the legislation, regulations, and other standards that govern the collection and dissemination of monetary statistics. The Organic Law of the National Banking System, the Organic Law of the BCCR, and other standards can be found on the BCCR Internet site. The Internal Labor Regulations of the BCCR and other regulations are found on the BCCR Intranet site (for internal use) and are available to the public at the BCCR’s Information Office.

1.2.2 Internal government access to statistics prior to their release is publicly identified

No government agency has access to monetary statistics prior to their release. Once compilation of the statistical output is complete, and before it is disseminated on the Internet, it is made available to the DE and senior BCCR officials, including the Board of Directors (made up of the BCCR President, the Minister of Finance, and private sector individuals of impeccable integrity, considerable expertise, and broad experience in economics, finance, and banking). At present the public is not informed of this practice. However, a note on this subject will be included on the Costa Rican page of the IMF’s Dissemination Standards Bulletin Board when Costa Rica subscribes to the SDDS.

1.2.3 Products of statistical agencies/units are clearly identified as such

The BCCR Internet site clearly indicates that the BCCR is responsible for its statistical products, although the BCCR department that prepares a specific product is not identified. However, the printed publications of the BCCR (available to the public at the CDI) include a note that identifies the SPPF as responsible for the monetary statistics published.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

Any significant change in methodology, data sources, and statistical techniques is disclosed to the public in the Technical Note that appears in the relevant printed publications and on the BCCR Internet site, although not always in a timely manner. For example, although the joint BCCR/MOF auction took place in April of 1996, the pertinent clarifications were reported in the Technical Note two months later.

The Technical Note is sometimes accompanied by preliminary estimates that make it possible to maintain the comparability of the monetary accounts. The formulation of these estimates is time consuming and thus there is some lag between the publication of the original series and explanations of the methodological changes together with the respective estimates. Consideration is being given to expanding the use of the Technical Note to inform the public of significant changes in the monetary series as quickly as possible.

1.3.1 Guidelines for staff behavior are clear and publicized

Article 41 of the Organic Law of the BCCR establishes that the regulations of the BCCR shall contain the powers and obligations of BCCR employees. In this regard, it should be mentioned that BCCR employees are subject to the Internal Labor Regulations of the BCCR dated October 1987. Article 37 of these regulations establishes that BCCR employees must maintain the utmost secrecy regarding the operations, business, and activities of the BCCR and its clients. In addition, Article 38 absolutely prohibits employees from using their positions to obtain personal advantages of any type from colleagues or from third parties. Articles 42 to 48 establish disciplinary sanctions for violations of the rules. These standards are made known to all new BCCR employees and are available on the BCCR Intranet.

2.1.1 The overall structure in terms of concepts and definitions follows international standards, guidelines or agreed practices

The analytical framework that the BCCR uses to compile and present the central bank survey (CBS) and the other depository corporations survey (ODCS) reflects concepts and definitions that are generally consistent with the MFSM. That framework focuses on a measure of broad money that comprises the liabilities of the depository corporations (the financial assets of other sectors of the economy), and on the claims of the depository corporations on other sectors of the economy. The MFSM does not prescribe a single definition of broad money since national definitions of this aggregate vary considerably from country to country. Each country uses its own definition of broad money to determine the institutional units that make up the depository corporations survey (DCS).

The BCCR publishes the monetary survey known as Sistema Bancario Nacional (National Banking System), which consolidates the sectoral balances of the BCCR and the commercial banks; among the latter is included the Banco Popular y de Desarrollo Comunal. Publication of the DCS, known as Sistema Bancario Amplio (Broad Banking System) started in January 2001 with monthly data since December 1997. The DCS consolidates the CBS and the ODCS, which are derived from the sectoral balances of the BCCR and the ODCs, respectively. The ODCs cover all resident financial corporations and quasi-financial corporations the activities of which consist primarily of financial intermediation and which issue liabilities included in the measure of broad money.

The main monetary and credit aggregates that are identified in this analytical framework are monetary base (currency issued, deposits and short-term investments of commercial banks with the BCCR); M1 (notes and coins, private and public sector current account deposits, excluding central government, in commercial banks); quasi-money in domestic currency (financial instruments in domestic currency); M2 (M1 and quasi-money in domestic currency); total quasi-money (quasi-money in domestic currency and financial instruments in foreign currency); M3 (M1 and total quasi-money); net domestic credit (broken down into central government (net), other public sector, nonfinancial private sector, and nonbank financial private sector); and net external position (gross credit to nonresidents minus gross liabilities of nonresidents).

2.2.1 The scope is broadly consistent with international standards, guidelines, or agreed practices

The Costa Rican financial system consists of a wide variety of financial intermediaries that operate under different regulatory, legal, and institutional schemes. The financial sector includes (i) the BCCR; (ii) the ODCs (commercial banks, private financial corporations, savings and loan cooperatives, savings and loan mutual associations, and the Savings and Loan Fund of ANDE (a teachers’ cooperative)); and (iii) other financial institutions (Housing Mortgage Bank (BANHVI), the National Insurance Institute, the Cooperative Development Institute, the National Housing and Urban Development Institute, and the National Teachers’ Pension Association).

The commercial banks (three state banks, 17 private banks, and one public law bank) predominate in the commercial banking system. In September 2000, 12 of the 17 private commercial banks had parallel banks abroad to handle offshore operations. The offshore banks are parallel entities of local banks but they are located overseas (in Panama, the Bahamas, the Cayman Islands, and Montserrat). These banks were formed in the 1980s as a way to avoid legal reserve requirements and some taxes. It is estimated that the assets of offshore banks in 2000 were, on average, 1.4 times higher than the assets of local banks.

Monetary statistics cover the operations of the BCCR and the ODCs. Institutional coverage includes the activities carried out by the headquarters and all their domestic branches. In the case of the ODCs, the branches send information to the central office, which reports the consolidated balance to SUGEF. It should be mentioned that offshore operations are not recorded in the monetary accounts nor were they considered within the framework of this assessment because of the absence of detailed information on these operations. The DCS (Sistema Bancario Amplio) consolidates the Sistema Bancario Nacional with private financial corporations, savings and loan cooperatives, savings and loan mutual associations, and the Fund of the National Association of Educators (ANDE). A monthly database is available on the Internet starting in December 1997. The development of the DCS (Sistema Bancario Amplio) significantly improved the coverage of the monetary accounts because the new institutions incorporated in the survey issue financial liabilities (savings and term deposits) that are considered part of the money supply. All the savings and loan cooperatives supervised by SUGEF are included in the Sistema Bancario Amplio. It should be pointed out that the monetary accounts output reports (Sistema Bancario Nacional and Sistema Bancario Amplio) have not yet been fully adapted to the sectorization recommended in the MFSM.

Within the context of the implementation of the MFSM, the BCCR expects in future to produce a financial survey that consolidates the DCS (Sistema Bancario Amplio) with operations of pension and insurance funds, investment funds, stock brokers, and exchange houses. Currently, some of these nondepository corporations (of the nonbank financial public sector) submit hard-copy monthly or quarterly financial statements to the BCCR. However, this information is not processed because its level of detail does not satisfy the requirements of the DM of the BCCR.

2.3.1 Classification/sectorization systems used are broadly consistent with international standards, guidelines, or agreed practices

The sectorization of the economy currently used in compiling the monetary statistics of the BCCR and the ODCs is based on a national system that is not entirely consistent with international recommendations. Current sectorization first classifies the economy into residents and nonresidents and then the domestic economy as follows: (i) central government (institutional units directly dependent on the budget); (ii) other public sector (ancillary units, local governments, Costa Rican Social Security Fund, and nonfinancial public enterprises); (iii) bank financial public enterprises (BCCR and commercial banks, including three whose capital belongs to the government and one under public law); (iv) nonbank financial public enterprises (National Insurance Institute, Housing Mortgage Bank, National Housing and Urban Development Institute, Municipal Development and Advisory Institute, National Cooperative Development Institute, ANDE Fund, etc.); (v) bank financial private enterprises (commercial banks whose equity is under private control, whether domestic or foreign); (vi) nonbank financial private sector (cooperatives, financial corporations, and mutual associations) and (vii) nonfinancial private sector (institutional units not included in the other economic sectors). For analytical purposes, the accounts of the central government and the rest of the public sector are consolidated into the nonfinancial public sector.

The BCCR is implementing a new system of monetary accounts, based on the sectorization of institutional units and classification of financial instruments recommended in the MFSM, which in future will facilitate compilation of the financial survey. Under this project, it is envisaged to compile a “single sectoral balance sheet” for the BCCR and for the ODCs. All the institutional units operating in the economy have already been identified and sectorized following 1993 SNA and MFSM recommendations. The preparation of these balance sheets will require the reclassification of certain accounts, the elimination of others, and the creation of historical series (five years for annual data and two years for monthly data) for the new monetary accounts. About 70 percent of the project has been completed to date for the BCCR’s accounts; what remains is cleaning up the accounts and constructing the historical series. The SPPF has begun the revision of the accounts of the ODCs and the generation of the ODCS based on the information currently provided by the ODCs as supplemented with information that the principal public entities (and the Budgetary Authority) provide to the Government Finance Section of the DM on their operations with the financial sector. This approach will facilitate the achievement of an approximate sectorization until it becomes possible to derive the supplemental information directly from the ODCs when the data entry system becomes operational.

Notwithstanding this progress, there is lack of clarity about the nature of the transactions between the domestic banks and their offshore branches, which are currently unregulated and excluded from monetary statistics (see Section 2.2.1). Although, in principle, monetary statistics should include only the transactions of resident banks, the lack of transparency and complete data on the operations of offshore banks makes it difficult to determine the proper treatment and sectorization, in the monetary statistics, of the transactions between resident banks and their offshore branches.

The system used to classify financial instruments in the sectoral balance sheets and in the monetary surveys (Nacional and Amplio) is generally consistent with international recommendations, since it is based on the liquidity of the instrument and the legal characteristics defining the basic relationship between creditors and debtors. In future, the BCCR plans to conduct a detailed study of the classification of financial instruments in the different monetary and credit aggregates. Meanwhile, it has been determined that (i) technical reserves of insurance plans are not identified in the basic sources; (ii) there are no swap operations; and (iii) repurchase agreements (repos) are considered guaranteed loans and not sales of securities for cash. There is no transfer of ownership and only the composition of assets and liabilities changes.

2.4.1 Market prices are used to value flows and stocks

Valuation practices for financial assets and liabilities are based on accounting rules established by SUGEF in the Chart of Accounts for Financial Institutions. The BCCR will conduct a detailed study of current accounting and valuation rules and the MFSM recommendations in order to identity areas that require change.

In the meantime, it has been determined that, consistent with the MFSM, monetary gold is valued at market price and loans are valued at book value without adjustment for anticipated losses due to the likelihood of default (estimates are recorded as other assets on an individualized basis and provisions are classified as other liabilities). Financial transactions are recorded at market prices in the purchase and sale of assets and liabilities at the time new securities are issued or when ownership is transferred, except in some BCCR accounting items (e.g., the accounting system records stabilization bonds placed at a discount based on their face value, but monetary statistics reflect the transaction value of the operation).

It has been determined that the monetary surveys (Nacional and Amplio) do not include information items that supplement the information on stock market shares and capital holdings recorded as assets in the balance sheet, as recommended by the MFSM (although these amounts are insignificant). In addition, the valuation adjustments (including gains and losses on instruments held) are not presented (in the sectoral balance sheets and surveys) as valuation changes but are included in the income statement.

The BCCR follows the residence criterion of the balance of payments in sectorizing the foreign sector accounts. According to the MFSM and 1993 SNA, foreign assets and foreign-currency denominated obligations are valued at the reference buying rate established by the BCCR at the close of the month. The preliminary information is valued at the reference buying rate for the day. The reference exchange rate for day “x” is the exchange rate for the Organized Market for Electronic Currency Trading (MONED), which is determined on day x-2 based on the average of quotes for the U.S. dollar on that day. Commercial banks also use this type of reference exchange rate to value the foreign currency accounts in their balance sheets at the close of each month.

2.4.2 Recording is done on an accrual basis

Recording is done in an accrual basis in accordance with 1993 SNA. Transactions are recorded at the time ownership is transferred. For example, the auction occurs on Monday but is recorded on Wednesday when the investor actually makes payment and the BCCR credits the account.

Interest on financial assets and liabilities accrues continuously throughout the accounting period. Delays in interest payments (interest in arrears) are considered a charge to be paid and are not classified by instrument. In the case of the BCCR, there is no interest in arrears. In the BCCR asset accounts, the government is usually the principal debtor, although currently there are no delays. The ODCs follow SUGEF’s Chart of Accounts for Financial Institutions in recording late interest payments. Delays in the payment of fees or similar payments corresponding to financial instruments not included in the balance sheet are reported in the balance sheet as accounts receivable/accounts payable (miscellaneous accounts).

2.4.3 Grossing/netting procedures are broadly consistent with international standards, guidelines, or agreed practice

Monetary statistics are collected and compiled in gross values following the general principle of the MFSM and 1993 SNA. In the sectoral balance sheets of the BCCR and the ODCs, data on financial assets and liabilities are aggregated in their principal categories on a gross basis. In the case of liabilities, the principal creditors are identified as government, rest of the nonfinancial public sector, private sector, and foreign sector.

In the central bank survey (CBS) and other depository corporations survey (ODCS), credit to the central government and international reserves are presented as net values due to their analytical usefulness, i.e., by netting deposits from credit to the government, and short-term foreign liabilities from foreign assets (gross figures), respectively. The ODCS is the consolidation of the sectoral balance sheets of the ODCs vis-à-vis the other domestic sectors of the economy (excluding the ODCs subsector) and the foreign sector. The data are stocks and are consolidated, i.e., interbank balances are eliminated from the data, subtracting all outstanding claims and liabilities between the depository corporations.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

The basic information sources that the SPPF uses to compile the CBS are the BCCR balance sheet and the comparative income statement prepared by the Accounting Department of the Financial Services Division. The balance sheet consists of a general listing of all the accounts of the BCCR and the comparative income statement presents the operational income and expenditures of the BCCR. The daily report of the BCCR generated from the SIFC contains the principal assets and liabilities accounts on a daily basis and is used to feed the MONET database. The information for the BCCR is supplemented with data on the stabilization bonds that are maintained in a securities management system that is not currently linked to the SIFC. In addition, the accounting system does not classify open market operations by sector; as a result, the BCCR obtains the sectoral information from the National Stock Exchange, from the commercial banks, and from government agencies.

The source of information for preparing the sectoral balance sheet of the ODCs are the call report forms that SUGEF sends each month to the BCCR by e-mail. The information is sent in an Excel file that contains the financial/accounting listing of the financial statements for each of the ODCs (commercial banks, private financial corporations, savings and loan cooperatives, savings and loan mutual associations, and the ANDE Fund). In addition to the accounting records, the BCCR receives directly from the ODCs a set of appendices (in hard copy) with monthly and weekly information. In the latter, the principal accounts of the balance sheet are disaggregated by institutional sectors and, in the case of credit to the private sector, broken down by branch of economic activity, new placements, and repayments. The appendices also include a set of forms to collect weekly information from the ODCs, which is used to derive the monetary aggregates. These data are presented in the Weekly Report of the BCCR (for internal use), and are used to prepare the preliminary monetary accounts that are published on the Internet and feed the database for other reports.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

The BCCR is making efforts to improve the primary sources of information and adapt them to current international recommendations in this field. The structure and degree of detail of the BCCR balance sheet make it possible to adopt the sectorization standards of the MFSM; but those of the ODCs require additional information.

The call report forms of the ODCs are based on the Chart of Accounts for Financial Institutions that was developed and implemented by SUGEF in 1993 and adopted by the ODCs starting in July 1995. That chart of accounts is suitable for banking supervision purposes but lacks the degree of detail necessary to clearly differentiate the principal components of the domestic sectors of the economy in order to compile monetary statistics according to international standards. The supplementary reports on credit and deposits make it possible to sectorize the principal accounts in the balance sheet (assets and liabilities) and to identify the use of credit to the private sector, but do not provide a sufficient level of detail, thereby affecting adversely the quality of the monetary statistics.

It should be noted that in 2000 the Sistema Bancario Nacional reported 19 percent of its assets and 13 percent of its liabilities as “not classified” and the Sistema Bancario Amplio reported 15 percent of its assets and 12 percent of its liabilities as “not classified.” In the context of the implementation of the MFSM, the approach has been to use as a preliminary approximation the information that government agencies provide to the Government Finance Department and subsequently, when the data entry system becomes operational, to request the ODCs to provide all the information necessary.

3.1.3 Source data are timely

The preparation of the sectoral balance sheets for the BCCR and the ODCs is generally timely and facilitates the subsequent compilation of monetary statistics. The DM has direct access to the SIFC, which contains historical and current (including daily) information from the BCCR balance sheet and the comparative income statement. The ODCs must send their balance sheets to SUGEF within 10 business days of the close of the reference month. In turn, SUGEF transmits this information to the BCCR within five business days. The information (in the form of an electronic file) generally reaches the BCCR in the third week of each month, although there are sometimes lags primarily due to delays in the submission of information to SUGEF by the ODCs. In addition, the ODCs must send the BCCR the balance sheet, a breakdown of debit and credit accounts, and the composition of investments within 10 business days after the close of the month; credit by branch of activity 20 calendar days after the end of the reference month; and weekly information three business days after the end of each week.

3.2.1 Data compilation employs sound statistical techniques

In 1999, the computer equipment and programs for collecting BCCR data were upgraded. To this end, a new database (MONET) was introduced, making it possible to transfer data directly through an interface between the main BCCR computer and the microprocessors and to generate the BCCR monetary accounts in an efficient and timely manner. In addition, the MONET system makes it possible to execute more efficiently (i) automated loading of the ODC information coming from SUGEF; (ii) manual processing of the hard-copy information provided by the ODCs; and (iii) generation of monetary accounts reports that can be transferred to electronic spreadsheets (previously only printed reports were obtained).

The structure of the ODC call report forms makes the task of compiling the monetary statistics more difficult. In recent years, the BCCR has been trying to improve the systems for collecting and processing monetary statistics. However, there are still significant improvements to be implemented, particularly with respect to the automation of procedures, as the supplementary data report forms of the ODCs are not designed for electronic processing. The potential for committing processing errors will be reduced to a minimum once the data entry system is established.

The SPPF has a Manual for Preparing Monthly and Weekly Reports, which explains in detail the sources, methods, and classification of accounts followed in the compilation of monetary statistics. This manual is revised periodically, most recently in March 2001.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The monetary accounts cover all financial intermediaries currently supervised by SUGEF and are based on their accounting records, with some exceptions. For example, when a financial institution is subject to intervention by SUGEF, the latest information reported prior to intervention is maintained, as recommended in the MFSM. Currently there are three ODCs subject to intervention (Banco Federado, Banco Cooperativo Costarricense, and Banco Solidarista). According to the rules stipulated in Article 140 of the Organic Law of the BCCR, the intervention process may not exceed one year, at the end of which an opinion is issued. For purposes of the monetary statistics, and in accordance with the MFSM, the latest information reported in the balance sheet of the financial corporation subject to intervention is included in the preparation of the Sistema Bancario Nacional and the Sistema Bancario Amplio until the liquidation process is completed. Once this process has been concluded, the operations of the corresponding intermediary are eliminated from the monetary accounts.

3.3.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessments are monitored and made available to guide planning

The process of assessing and validating the statistics of the BCCR is automated. The SPPF consults with the DC and the sections of the BCCR responsible for processing other macroeconomic statistics to validate the income statement accounts of the BCCR balance sheet and any accounting item that may be questionable.

The methods for controlling the quality of the data from the ODCs are based partially on manual verification and validation procedures, which require significant human resources and do not completely eliminate the possibility of errors. The introduction of the data entry system that is being considered for the future will strengthen the quality control systems for monetary statistics. The SPPF analyzes the consistency of the information that the ODCs send to SUGEF and the BCCR, using pre-defined reports that make it possible to review whether the sectorization followed by the ODCs agrees with the balance sheets reported to SUGEF. In addition, reports are used to evaluate, month by month, the movements of all the balance sheet accounts. If significant changes are noted, there is direct consultation with the ODC and, depending on the nature of the operation and the justifications provided, the corrections are made directly in the monetary accounts (e.g., changes in the sectorization of items) or, where warranted, SUGEF is informed so that it can study the matter.

To move forward with the modernization of the production of the monetary accounts, the BCCR has taken up again the project (data entry system) that was suspended in 1999. This program would permit the electronic entry of the information that the ODCs send on hard copy (Monthly appendices to the balance sheet and weekly information), as well as validation of this information against the accounts provided to SUGEF. As of 1999, the technological platform had been constructed, the data entry forms had been designed, and the first training sessions had been provided to the ODCs. This project has now been resumed; however, changes introduced since then in the technological platform currently used by SUGEF require that both platforms be made compatible.

3.3.2 Main intermediate data are validated against other data where applicable

The SPPF crosschecks the ODC data sent by SUGEF with the appendices sent by the ODCs directly to the BCCR. No secondary sources are used but various crosschecks are performed using information that the DM receives from other sectors. For example, the bank assets reported by public enterprises to the BCCR (based on a sample of 12 to 15 nonfinancial public enterprises) are compared with the liabilities that the ODCs show with the public sector (using the BCCR database). In general, these crosschecks show small differences between what the public enterprises and the ODCs report.

3.3.3 Statistical discrepancies and other potential indicators of statistical problems are investigated and made known to users

The data from the ODCs are checked to prevent any statistical discrepancy due to errors in classification or omissions. When an unexplainable variation is noted, the DM contacts the respective ODC to clarify the reasons for the fluctuation. In the event that the ODCs reclassify their balance sheet accounts, an explanation is provided to the user in the Annual Report of the BCCR, in the biannual economic reviews, or directly to the user who submits a query.

In general, substantial fluctuations are not recorded in the monetary or credit aggregates, unless a change has been introduced in the methodology. When this happens, the BCCR discloses the nature of the methodological changes in the Technical Note that appears on the Internet or that is printed along with the statistics in the Black Book.

3.4.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

The data of the BCCR are final when first published and those of the ODCs, although preliminary, are updated with final data between six to eight weeks after the close of the reference month.

4.1.1 Processes to monitor the relevance and practical utility of existing statistics in meeting users’ needs are in place

The BCCR does not regularly participate in user forums nor does it organize seminars with users of monetary statistics. On occasion, at the request of academic users, the BCCR has given presentations on the construction of the monetary accounts and the monetary program. In recent years, the BCCR has also provided briefings to journalists on subjects of interest such as the BCCR deficit, developments in the financial system, government debt, direct investment, and Costa Rica’s upcoming subscription to the SDDS.¹⁰ In August 2001, the briefing on the monetary sector will deal with the definitions of the principal accounts and monetary indicators used by the BCCR. Consideration will be given to the possibility of discussing the subject of implementing the MFSM In the future.

The BCCR is aware of innovations that affect financial markets and financial institutions and tries to reflect these changes in the monetary statistics. This is not an easy task because in recent years large conglomerates have been formed that carry out specialized functions, and the financial market is issuing a series of new instruments that give investors the opportunity to invest in short-term instruments that guarantee a specific return. In addition, there is widespread use of contingent assets and liabilities accounts by the banks (i.e., transactions recorded off the balance sheet).

At the request of the General Superintendency of Securities (SUGEVAL), the ODCs have reclassified the contingent accounts and have moved some of them to balance sheet accounts. Suspense accounts are not included in the monetary aggregates because legally they do not belong to the ODCs.

4.2.1 Periodicity follows dissemination standards

The BCCR publishes on its Internet site monthly data on the analytical accounts of the BCCR (Central Bank Survey) and the analytical accounts of the banking sector (Sistema Bancario Nacional and Sistema Bancario Amplio). The periodicity of these statistics is consistent with SDDS requirements.

4.2.2 Timeliness follows dissemination standards

The timeliness of the monetary statistics is consistent with the SDDS. The BCCR publishes final monthly data on the analytical accounts of the BCCR one week after the close of the reference month, monthly preliminary data on the analytical accounts of the banking sector two weeks after the close of the reference month, and final data with a lag of six to eight weeks.

4.3.1 Statistics are consistent or reconcilable over a reasonable period of time

Monetary statistics are consistent over time. The Internet site of the BCCR and the printed publication of the Black Book have a Technical Note with information on methodological changes. The Technical Note is revised when there are operations that change the behavior of a variable making it necessary to introduce methodological changes, sometimes accompanied by preliminary estimates to maintain the comparability of the monetary accounts.

4.3.2 Statistics are internally consistent

In general, the accounting of assets and liabilities between the BCCR and the ODCs is consistent. On occasion, recording differences may arise due to operations in transit, but these are always reconcilable. In the last three years, discrepancies have been reduced due to the improvements introduced in the Costa Rican payments system that involve the use of online connections, electronic payments, third-party electronic transfers, and the definition of a banking schedule. These factors, among others, make the process quicker and accounts are credited or debited with less lag time.

4.3.3 Statistics are consistent or reconcilable with those obtained from other data sources or statistical frameworks

The monetary statistics can be reconciled with those of the balance of payments and government finance.¹¹ The accounts included in the definition of international reserves of the BCCR are consistent as recorded in the monetary accounts and in the balance of payments. International reserves include monetary gold, reserve position with the IMF, SDR holdings, foreign banknotes and coins, foreign bank deposits, and investments in foreign securities. Foreign liabilities include the use of IMF credit, other short-term foreign liabilities, and liabilities to international monetary organizations. The SPPF and the Balance of Payments Section (SBP) derive data on foreign assets from the BCCR accounts on a daily but independent basis. The SPPF provides the banks’ international reserve figures to the SBP. However, both sections calculate separately the BCCR’s international reserves daily and crosscheck the results to ensure their accuracy.

Government finance statistics on central government deposits and deposits of the principal public institutions with the ODCs are reconcilable with the corresponding monetary statistics. The SPPF reconciles the below the line accounts with the information provided by the Budgetary Authority to the Government Finance Section of the BCCR.

4.4.1 Revisions follow a regular, well established, and transparent schedule

Monetary statistics are revised regularly according to an internal BCCR schedule that establishes the publication dates. The first revision of the preliminary statistics is released within the BCCR with the production of the Weekly Report (for internal use only), which contains 12 tables and charts on the monetary accounts, in addition to preliminary daily and weekly data for the principal mobilization and credit instruments of the BCCR and the commercial banks. This Weekly Report is sent every Wednesday with information corresponding to the immediately preceding Friday (BCCR) and the Friday before the previous one (commercial banks). The Weekly Report makes it possible to monitor the principal monetary variables and is presented to the directors and chiefs of the DE. The Weekly Report is also sent to management and to the Board of Directors of the BCCR and is used as a data source for the Black Book that is published on the Internet. The figures are updated on the second and fourth Wednesday of each month with revised but not always final data.

The data for the BCCR are final when first published (except for the BCCR income accounts). The data for the ODCs and the monetary surveys (Nacional and Amplio) are preliminary when first published and are available two weeks after the close of the reference month. The final figures are available between six and eight weeks after the close of the reference month. Data for the Sistema Bancario Amplio have a one-month lag compared to the Sistema Bancario Nacional.

4.4.2 Preliminary data are clearly identified

The preliminary data are identified as such in the publications of the BCCR. For example, the Black Book clearly identifies whether the data are preliminary in a note that says “Preliminary as of (date).”

4.4.3 Studies and analyses of revisions are carried out periodically and made public

Studies on the consistency between preliminary and final data are conducted for internal purposes but are not published.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparison (layout and clarity of text, tables, and charts)

The presentation of monetary statistics facilitates their interpretation and lends itself to the monitoring of monetary and financial developments.

5.1.2 Dissemination media and formats are adequate

The media and formats used in the dissemination of data are adequate, particularly for users who have access to the Internet, as both recent data and the historical series can be consulted on the BCCR Internet site. The BCCR publishes a Monthly Report with information (tables and charts) on the principal monetary aggregates such as currency issue, monetary base, total liquidity, and total domestic credit. It should be pointed out that publication of the printed versions of the Annual Report and the Weekly Reviews was suspended more than two years ago. The historical publications are found at the CDI.

5.1.3 Statistics are released on a pre-announced schedule

As of July 17, 2001, the BCCR publishes on its Internet site an advance release calendar for monetary statistics that meets SDDS specifications.

5.1.4 Statistics are made available to all users at the same time

The monetary statistics are released simultaneously to all interested parties on the BCCR Internet site.

5.1.5 Non-published (but nonconfidential) subaggregates are made available upon request

As much as possible, the BCCR provides disaggregated data at the user’s request, provided that these do not include confidential data.

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from international standards are noted

The BCCR offers on its Internet website a brief description of the principal monetary variables. However, it does not describe the scope of the data nor does it present information on the unit of the BCCR responsible for producing the data. With the implementation of the program to adopt the MFSM, the pertinent documentation will be prepared and disseminated on the Internet.

5.2.2 Levels of detail are adapted to the needs of the intended audience and the type of compilation

The Economic Indicators Section of the BCCR Internet site contains time series for the principal variables at the level of the BCCR, the Sistema Bancario Nacional and the Sistema Bancario Amplio (monetary base, monetary aggregates and their components, net domestic credit, and the net foreign position) on a monthly basis since 1987. The First and Second Biannual Economic Reviews from the second half of 1996 onwards are included on the site and contain data on total liquidity, the money supply, quasi-money, and the net domestic credit of the national banking system by sectors. The database for the Black Book begins in 1986 with monthly data on the principal liability accounts at the level of the BCCR, the Sistema Bancario Nacional, and the Sistema Bancario Amplio (the latter starting in December 1997) expressed in millions of colones and with monthly cumulative changes with reference to December and on a year-to-year basis, expressed as percentages. The Annual Report, available on the Internet since 1995, includes analyses and commentary on economic developments.

On the occasion of the 50^th anniversary of its founding, the BCCR prepared a compact disk with information on the principal macroeconomic statistics since 1950. This information is available to the public for consultation at the CDI.

5.3.1 Contact person for each subject field is publicized

The internal users of the BCCR receive support and assistance from the compilers of the monetary statistics. External users send their queries through the BCCR Internet site. The DE receives queries on various subjects and directs them to the respective sections of the BCCR. In addition, the DE handles telephone queries on concepts, definitions, and requests for information, as well as questions that users present to the BCCR. When Costa Rica subscribes to the SDDS, the public will be able to send their questions directly to the contact person identified in the metadata.

5.3.2 Catalogues of publications, documents, and other services, including information on any changes, are widely available

The principal media for dissemination of monetary statistics is the Internet. However, the CDI has all the information that appears on the Internet (including more extensive historical series) and provides services to users via fax, Internet, information sent by e-mail, and photocopy.