Abstract

Central America has made significant efforts in recent years to improve data quality and provision, with a view to ensuring sound economic policymaking and to fostering investors’ confidence (see Box 7.1).1 The countries that are already issuers of sovereign bonds in international capital markets are well aware of the need to provide timely and adequate statistical information to market participants (see Table 7.A1).2 Thus, Panama and Costa Rica, followed by El Salvador, are at the forefront of data compilation and use of cutting-edge methodologies.

Central America has made significant efforts in recent years to improve data quality and provision, with a view to ensuring sound economic policymaking and to fostering investors’ confidence (see Box 7.1).1 The countries that are already issuers of sovereign bonds in international capital markets are well aware of the need to provide timely and adequate statistical information to market participants (see Table 7.A1).2 Thus, Panama and Costa Rica, followed by El Salvador, are at the forefront of data compilation and use of cutting-edge methodologies.

El Salvador subscribed to the Special Data Dissemination Standard (SDDS)3 in 1998 and Costa Rica subscribed in 2001; Panama began participating in the General Data Dissemination System (GDDS) in 2000,4 and Guatemala began in 2004. Honduras and Nicaragua are adapting their statistical frameworks to international standards, including by preparing to participate in the GDDS and regional initiatives to facilitate cross-country comparisons and data exchange. Nicaragua and Honduras have also made some progress in improving the quality of statistics, especially with respect to the national accounts and the monetary statistics. The drive for more harmonized statistics is motivated in particular by the trend toward increased economic integration and policy coordination in the region.

Despite the progress so far, statistical deficiencies remain, with uneven data quality across sectors and countries. Outdated methodologies, poor source data, and inconsistency across sectors affect countries to different degrees, hampering policy formulation and monitoring. A key advantage throughout the region is that the fiscal statistics generally cover the entire nonfinancial public sector, although precise identification of the overall balance by level of government is problematic in some countries.

The quality of legal frameworks for compiling and disseminating statistics varies. In Honduras, a new law that came into effect in 2000 contains comprehensive provisions that strongly support the compilation of statistics, whereas in El Salvador the national statistical law is outdated and suffers from significant gaps. The legal framework in Costa Rica does not grant compilers legal power to collect data from the nonfinancial private sector and to impose sanctions for noncompliance. Such limits hamper, for example, the collection and analysis of balance of payments data, especially on capital flows. Penalties for noncompliance are adequate in only a few countries, such as Honduras, where the real value of penalties is protected by indexation to the minimum wage. However, compilers seldom invoke laws and resort to penalties when requesting data, preferring to rely on persuasion, with different degrees of success. Therefore, an improvement in macroeconomic statistics requires both a more rigorous application of existing laws and—in most countries—changes in legal frameworks.

In most countries in the region there is scope to improve coordination among statistical agencies and for a targeted increase in resources to improve statistical output. In particular, greater coordination among the central bank, the finance ministry, and the national statistical agency would help governments better assign responsibilities for data sets, strengthen the collection of basic data, and improve the timeliness of certain data, including on the general government. There are significant disparities in resources available to support statistical production: central banks are generally better endowed than national statistical agencies and finance ministries.

Main Data Issues in Central America

National Accounts

Outdated methodologies and an obsolete base year affect the quality of the national accounts in El Salvador, Guatemala, and Honduras. Concepts and definitions in El Salvador follow the 1968 System of National Accounts (SNA) with 1990 as base year. In Honduras compilation is conducted under the 1953 SNA with 1978 as the base year. Guatemala is currently implementing the 1993 SNA and is changing the base year of the national accounts to 2001. All countries still face problems with limited source data, although efforts are under way to broaden coverage. For instance, Panama is completing a revision of its national accounts, adopting the 1993 SNA and changing the base year from 1982 to 1996. Coverage has been extended to activities that did not exist in 1982, and the accounts now include resident banks and enterprises operating in the Colón Free Zone (ZLC). In the same vein, Nicaragua recently conducted a census of enterprises in urban areas plus surveys of small enterprises and the informal sector, in the context of the implementation of the 1993 SNA (see Table 7.A1).

Prices

In general, the consumer price index (CPI) is based on internationally accepted practices and standards. Moreover, most countries have updated weights according to recent household expenditure surveys. In addition, Costa Rica, El Salvador, Honduras, and Panama publish a producer price index (PPI), and also publish related date (except for Honduras). However, coverage of the PPI is limited in most cases to the manufacturing sector and excludes key sectors such as maquila, mining, and electricity.

Government Finance Statistics (GFS)

To facilitate financial programming, countries compile fiscal data on a cash basis following the Manual on Government Finance Statistics 1986 (GFSM 1986). Some countries in the region report GFS (on a cash basis) in the format of the Government Finance Statistics Manual 2001 (GFSM 2001) for publication in the Government Finance Statistics Yearbook (GFSY). Costa Rica, El Salvador and Panama are planning to review their accounting systems to incorporate the analytical framework of the GFSM 2001, which includes recording transactions on an accrual basis. Main shortcomings in most countries relate to coverage (exclusion of certain extrabudgetary operations and social security funds), classification, basis of recording (a mixed basis of cash and accrual), and discrepancies between the overall balance and financing data.

Monetary Statistics

Monetary statistics are reliable, but uneven coverage is an issue. Costa Rica, El Salvador, and Panama follow, to a large extent, the Monetary and Financial Statistics Manual (MFSM), though its adoption lags behind in Guatemala, Honduras, and Nicaragua. However, the latter countries are working with the IMF to improve their databases. In general, the main shortcomings are lack of information on offshore operations of bank conglomerates and investment funds (particularly important in Costa Rica), and difficulties in obtaining detailed sectorized deposit and credit data from banks. The latter hampers the reliability of the estimates of the public sector financing by levels of government. Costa Rica publishes a monetary survey comprising the monetary authorities and other depository corporations, as well as data on other financial corporations (such as cooperative development bank and mortgage banks). Monetary statistics in El Salvador also include data on nonmonetary financial corporations, whereas an expansion of coverage is needed in Panama to ensure inclusion of data on other nonbank institutions such as cooperatives.

Balance of Payments

Lack of sufficient source data hinders the reliability of the balance of payments in some countries. Costa Rica, El Salvador, and Panama compile the balance of payments in conformity with the methodology recommended in the fifth edition of the Balance of Payments Manual (BPM5), which is more demanding regarding the need for reliable source data. In contrast, compilation in Honduras and Guatemala is still based on the BPM4, although a move toward adoption of BPM5 is under way. Guatemala is in the transition between BPM4 and BPM5, and has adopted significant BPM5 recommendations, in particular for the current account. Costa Rica and Panama compile and disseminate quarterly balance of payments and annual International Investment Position (IIP) statistics, whereas El Salvador disseminates annual and quarterly balance of payments and IIP data. In spite of substantial progress in coverage and timeliness of balance of payments data by these countries, certain deficiencies remain as a result of the lack of reliable source data on services and financial account transactions. Sizable data revisions may suggest that there is room for improvement in quality control, even though this may also reflect improvements in coverage such as in Panama.

Data Suitability for Analytical Purposes

Despite these shortcomings, the quality of statistical data is broadly adequate for analytical purposes, particularly in the case of Panama and Costa Rica. Data are somewhat less suitable in Guatemala and Honduras, while they have improved considerably in Nicaragua. Until recently, El Salvador enjoyed a reputation for having a good statistical system in relation to regional practices. However, since dollarization in 2001, the ensuing streamlining of the functions of the central bank—the main provider of economic statistics—and an erosion of financial and human resources deployed to statistics have had some adverse effects, including the discontinuation of several key surveys. This has affected, in particular, the national accounts and the balance of payments (see Table 7.A2).

Dominican Republic: Data Issues

Main Data Issues and Their Suitability for Analytical Purposes

The Dominican Republic has made some progress in improving the quality of its macroeconomic statistical framework and, along with Honduras and Nicaragua, is in the process of participating in the GDDS initiative. Nonetheless, several shortcomings remain that hamper the use of the GDDS for analytical purposes.

A revision of the national accounts is being undertaken in line with 1993 SNA and an updated base year (1991). The quarterly series have, however, outdated conceptual and methodological frameworks that make them unreliable for analysis. Price statistics have new consumer price indices based on updated and more accurate consumption baskets, with weights based on the 1997–98 national household survey.

Government finance statistics are collected from various institutions but the lack of interinstitutional coordination creates serious problems for matching above-the-line data on expenditures and revenues to below-the-line financing from foreign and domestic sources. This gap is represented in some years as about 5 percent of GDP. Moreover, data on the operations of public enterprises are often partial and available with excessive delays.

Along with price statistics, the monetary statistics data set is one of the most reliable. It broadly follows the MFSM, although deficiencies in the data collection system, being addressed through the setup of a new database with automated data processing, will not be solved until the end of 2004.

Balance of payments statistics, despite being compiled in conformity with methodological standards of BPM5, have serious data sources constraints related mainly to coverage of imports, capital transfers, and income and financial accounts that generate large errors and omissions.

The debt data set has coverage and classification problems that have led to sizable underestimations of stocks (see Tables 7.A1 and 7.A2).

Statistics Technical Assistance

Since the inception of the data initiatives (1996), the Dominican Republic has received 16 STA technical assistance missions, which, compared with the Central American countries, place it as one (along with Costa Rica) of the major beneficiaries of STA technical assistance missions (vis-à-vis an average of 10 missions per Central American country during the same period). The bulk of the missions were on balance of payments and national accounts statistics. As a result, the compilation of balance of payments follows BPM5 and national account statistics have been revised. No mission has been fielded so far for government finance statistics. One multisector mission took place in 2002, and a data ROSC mission was requested by the authorities after a recent monetary and financial statistics mission, which also included a follow-up in other statistical areas (see Table 7.A3).

Monetary statistics are the most reliable and suitable for analysis and financial programming, although coverage and sectorization gaps continue to be a matter of concern. Closing these gaps would require substantial effort, as covering offshore operations and the needed modifications to the chart of accounts for financial institutions in some countries would entail resource costs and require further technical assistance (see also Section VI, on offshore financial transactions).

Shortcomings in the national accounts (the SNA) and the balance of payments complicate the analysis of sources of economic growth, national savings, output gap, and external sustainability. The national accounts are considered more reliable in Panama and Costa Rica. In Nicaragua, recent implementation of the 1993 SNA has updated the base year and introduced numerous surveys to broaden coverage, thus increasing data reliability. Although the balance of payments is considered more reliable than the national accounts, large and volatile errors and omissions introduce uncertainty to estimates and projections, particularly in Guatemala and Nicaragua.

Fiscal statistics are considered acceptable for analysis in most countries. However, the coverage of public enterprises is incomplete, and data are untimely in some countries, particularly in Nicaragua. Discrepancies between the overall balance and financing data, particularly in El Salvador and Nicaragua, have resulted at times in a gap of about 2 percent of GDP.

Cross-sectoral consistency is a major issue in the region. This affects especially the national accounts, balance of payments, and government finance statistics. For example, in Costa Rica, data on the operations of large nonfinancial enterprises have been affected by unreliable estimates of changes in inventories and imports. In Guatemala, operations of the maquila enterprises are included in the balance of payments, although reflected on a net basis, but excluded from the national accounts. In Nicaragua, the new series of national accounts include better estimates of capital and current expenditures than in the government finance statistics, for which overestimation of capital expenditures is a concern. Good progress to improve intersectoral consistency is being made in Panama, where reconciliation of data on capital flows between banking sector and balance of payments statistics is being conducted following a compilation method introduced by the Superintendency of Banks in 2002.

Analysts strongly support the implementation of up-to-date methodologies, but are concerned about possible breaks in the series. They consider that the implementation of more advanced methodologies improves data classification and sectorization, thereby facilitating the use of data for analytical purposes. They caution, however, that breaks in the series could complicate intertemporal analysis.

Moving Toward a Harmonized Statistical Framework

Deepening economic integration of the region should be supported by compilation of consolidated macroeconomic statistics for the region. Coordination of macroeconomic policies would, at the outset, require the exchange of statistical information among members so as to compare data sources and methodologies for classification and sectorization. Given the lack of an integrated statistical database for the Central American countries, there is a need to develop a coordinated, consolidated, and harmonized statistical system.

Some efforts in this direction have been undertaken by the Central American Monetary Council (CAMC), which has created working groups to deal with statistical issues. These groups are preparing working papers on estimating the co-circulation of domestic and foreign currency and analyzing operations of offshore banks. A working group on the national accounts is tasked with researching the determinants of economic growth in the region, and another group is working on a common definition of money in Central America.5 Also, the CAMC is working on a project to disseminate debt data of the central governments and central banks.

Some progress in harmonizing monetary statistics has been fostered by the policy needs of regional central banks. Monetary and financial policy issues common to the countries in the region are being discussed in the context of the CACM, with a focus on avoiding unexpected capital movements across the region. Accordingly, data sources, level of coverage, and data classification and sectorization have been assessed across countries to determine the degree of comparability. This project has been implemented with support from the IMF’s Statistics Department (STA) in the framework of the GDDS initiative.

The Partnership in Statistics for Development in the 21st Century (Paris 21) has also been promoting regional integration of Central American statistics.6 Workshops to strengthen the statistical capacity of Central American countries, organized by the Secretariat of Paris 21, took place in El Salvador (May 2003) and Panama (April 2004). The principal recommendations arising from these meetings include creation of a regional authority for statistical coordination and standardization of regional definitions and methodologies.

Methodological convergence is needed for a consolidated and a harmonized statistical system across the region. Regionwide implementation of the 1993 SNA would help in the preparation of consolidated accounts for Central America. This and further improvements to the national accounts would require better source data through a strengthening of the statistical institutions, the establishment and maintenance of statistical registers, and the regular conduct of economic surveys and censuses. Similarly, in the area of fiscal statistics, countries would be well advised to update the Government Finance Statistics Manual, budget, and accounting systems based on the GFSM 2001. Some countries should also improve fiscal data coverage, in particular by including extrabudgetary units, social security funds, and local governments. Guatemala, Honduras, and Nicaragua need to make further progress in the adoption of the MFSM and in compiling the balance of payments in conformity with the methodology recommended in BPM5.

Technical Assistance to the Region

Since the inception of the data standards initiatives (SDDS, GDDS, and data ROSCs), the IMF has fielded 88 missions to Central America and the Dominican Republic to assist in the improvement of the macroeconomic statistics. Costa Rica and the Dominican Republic received the most (15 missions each), followed Guatemala and Honduras (12 each), Panama (11), El Salvador (10), and Nicaragua (9) (Table 7.A3). Most missions were related to monetary statistics, followed by balance of payments and national accounts.

Data ROSC missions have visited Costa Rica, El Salvador, Honduras, Guatemala, and Nicaragua and have prepared detailed assessments of statistical practices based on international standards.7 In addition, several missions and seminars are taking place in 2004 to assist the Dominican Republic, Guatemala, Honduras, and Nicaragua in preparing to participate in the GDDS. These missions and seminars explain the modalities and benefits of using the GDDS as a framework to improve their national statistical systems and have assisted in (1) documenting compilation methods and sources (metadata), (2) identifying deficiencies in the compilation and dissemination process, (3) documenting the plans for improvement over the short and medium term, and (4) identifying technical assistance needs in the short and medium term. Metadata for the external and monetary sectors have already been finalized, whereas those for the real and fiscal sectors are being drafted.

Technical assistance has helped strengthen statistical systems in the region. A good example is Nicaragua, which has made considerable progress in updating the statistical methodologies. Nicaragua has significantly improved its monetary statistics (in particular central bank aggregates), publishes national accounts data that are compatible with the 1993 SNA, and presents fiscal data to the IMF in the format of the GFSM 2001.

Technical assistance could play a role in the support of regional integration. To this end, technical assistance will need to attach priority to strengthening methodological convergence and to making statistics comparable across countries. The IMF could provide technical assistance to regional institutions for creating coordinated and harmonized statistical systems. Given the high level of trade integration in the region, future technical assistances should also focus on harmonizing balance of payment statistics.

Conclusions

Deepening economic integration and increased macroeconomic policy coordination should be mirrored in a strengthening and methodological convergence of statistical systems in Central America. In the medium term, this could be supplemented through the development of an integrated statistical database for the region. Convergence of statistics should take place at the most advanced level, guided by those countries that already have access to international capital markets and that subscribe to the SDDS.

Overall, Central America has a good record of providing data to the IMF and has made progress in improving statistical frameworks. Among the Central American countries, Costa Rica and Panama are at the forefront of data compilation and use of cutting-edge methodologies. Costa Rica and El Salvador subscribe to the SDDS, whereas Panama and Guatemala subscribe to the GDDS. The other countries are preparing to participate in the GDDS. Statistical deficiencies remain, however, with uneven data quality across sectors and countries, including the use of outdated methodologies.

Appendix. Statistical Issues, Suitability of Data for Analytical Purposes, and Technical Assistance Missions

The following tables provide detailed information on statistical issues for each country.

Table 7.A1.

Statistical Issues

article image
article image
article image
article image
Key: BOP = Balance of payments; CPI = Consumer price index; GDDS = General Data Dissemination System; GFSM = Manual on Government Finance Statistics; GFSY = Government Finance Statistics Yearbook; IFS = International Financial Statistics; IIP = International Investment Position; IPRI = Industrial price index; MoF = Ministry of Finance; PPI = Producer price index; SDDS = Special Data Dissemination System; SNA = System of National Accounts; TA = technical assistance; CAMC = Central American Monetary Council; * = forthcoming missions.
Table 7.A2.

Suitability of Data for Analytical Purposes

article image
article image
article image
Table 7.A3.

Statistics Technical Assistance Missions, 1995–end-January 20051

article image

An asterisk represents “forthcoming missions.”

1

Although geographically the Dominican Republic is not part of Central America, it has joined a number of regional initiatives, including the Central American Free Trade Agreement (CAFTA-DR). Statistical issues for the Dominican Republic are summarized in Box 7.2 and noted as memorandum items in tables.

2

Costa Rica, the Dominican Republic, El Salvador, Guatemala, and Panama have issued sovereign bonds in international capital markets. Although Honduras and Nicaragua have not issued sovereign bonds, a number of private residents of these countries have issued securities internationally.

3

The SDDS was established in 1996 to guide countries that have, or that might seek, access to international capital markets in the dissemination of data to the public. Subscribers must fully meet the requirements of the SDDS at the time of subscription.

4

The GDDS was established in 1997 to guide countries in providing the public with comprehensive, timely, accessible, and reliable data. The GDDS, which provides recommendations on good practices in compilation and dissemination, is less demanding than the SDDS because it requires progress in statistical improvement over the medium term.

5

For information on research papers and working groups, see www.secmca.org/documentos.asp?mnu=11.

6

Paris 21 was launched at a meeting of statisticians and policymakers, which was held in Paris in November 1999. The meeting was an initiative of the United Nations, the Organization for Economic Coordination and Development (OECD), the IMF, the World Bank, and the European Commission in response to a UN Economic and Social Council resolution. This global consortium aims to help promote a culture of evidence-based policymaking and monitoring, especially in poor countries. The initial focus is on helping countries to prepare national Poverty Reduction Strategy Papers (PRSPs).

7

Starting in 2001, the data module of Reports on the Observance of Standards and Codes (data ROSCs) includes an assessment of data quality based on the IMF’s Data Quality Assessment Framework (DQAF).

Cited By