VIEWS: 18 PAGES: 16 POSTED ON: 5/4/2010
Strengthening Regional Capacities for Statistical Development in Southeast Asia Project Sponsored by UNSD, ESCAP and ASEAN Secretariat Bangkok, 6-10 August 2001 Compilation of Quarterly GDP: Methods, Problems, and Solutions The Case of the PHILIPPINES QUARTERLY PHILIPPINE SYSTEM OF NATIONAL ACCOUNTS (PSNA) Raymundo J. Talento WORKSHOP ON THE DEVELOPMENT OF ASEAN SHORT-TERM INDICATORS Compilation of Quarterly GDP: Methods, Problems, and Solutions The Case of the PHILIPPINES QUARTERLY PHILIPPINE SYSTEM OF NATIONAL ACCOUNTS (PSNA)1 Raymundo J. Talento2 I. BACKGROUND: The initial work on the Philippine Quarterly National Accounts started in 1983. It was conceived in response to the needs of planners and policy-makers for a more timely set of indicators regarding the performance of the economy. The first set of estimates covered the period 1981 and 1982 and basically follows the 1968 version of the System of National Accounts (SNA). The task of developing the quarterly national accounts was facilitated by the presence of quarterly data which some agencies have begun collecting as early as the mid 1970’s. The National Statistics Office (NSO) began their Quarterly Survey of Establishment (QSE) in 1974. Other agencies particularly, those with monitoring and/or regulating functions have started to adopt a quarterly reporting system. However, despite the availability of these quarterly data still, they were not sufficient to establish a consistent set of accounts. Moreover, the survey design of QSE was revised in 1981 to be more meaningful for national accounts purposes. From hereon, the National Statistical Coordination Board (NSCB) has been actively coordinating the data collection activities of the Philippine Statistical System (PSS), advocating and strengthening the data support for the quarterly national accounts. As the quarterly database improved, some of the estimation methodologies have to be refined or modified to reflect these improvements. 1 Paper presented in the ASEAN Workshop on the Development of Short-Term Indicators, held at ESCAP Bangkok, Thailand, 06-10 August 2001. 2 Division Chief for the Economic Indicators & Consolidation Division of National Statistical Coordination Board (NSCB) of the Philippines. 2 The publication of the quarterly accounts shows the Gross Domestic Product (GDP) both by industrial origin and by item of expenditures covering a three-year series. Subsectoral disaggregations are also presented however, no income and outlay account, capital reconciliation account or institutional accounts are compiled for the quarterly accounts. Deseasonalization of the GDP and its major components are also prepared and included in the publication. This paper presents the present compilation system of the Philippine Quarterly National Accounts following the outline provided by the ASEAN Secretariat. Some sectors in national accounts are discussed in detail while the other sectors are presented in a summary table. The deseasonalization of estimates will not be tackled here. II. DATA SOURCES FOR THE QUARTERLY ACCOUNTS: Censuses & Special Studies – Since it is difficult to obtain data that is on a one-to-one correspondence to the national accounts, the quarterly national accounts still relies on censuses and special studies in developing/deriving parameters that are part of the estimation methodology. Among others these parameters include gross value added ratio, undercoverage ratio and unit cost relevant to the sector concerned. The censuses essentially provide “complete” coverage of the population while the special studies cover specific areas or information that cannot be covered in the censuses considering that the censuses are geared towards providing general-purpose statistics. Special studies are normally done by the concerned agency in a particular area of interest or in some cases it is a collaborative effort among agencies that have also a stake or interest in that topic/area. Sample Surveys – Some agencies within the Philippine Statistical System (PSS) have started quarterly data compilation long before the quarterly account was initiated. The same is true for some private business associations. Majority of the sectors under the production side of the accounts rely on the Quarterly Survey of Philippine Business and Industry (QSPBI) However, these quarterly or even monthly surveys were in most instances not originally designed for national accounts purposes hence, the variables gathered served as indicators regarding the performance of a sector. In the absence of a census, the Annual Survey of Philippine Business and 3 Industries (ASPBI) is also utilized to update the various parameters used in the quarterly estimation. Administrative Data – This type of data came about out of a need to address the mandate of regulating and/or monitoring agencies. In most cases, they have complete coverage for their area of concern and most of the sectors under the expenditure side of the accounts utilize administrative-based data for their quarterly estimates. For monthly or quarterly monitoring, they cover a sample of the population that represents the performance of the sector. Some of the administrative based data can be used directly for quarterly accounts while others are used as indicators. III. ESTIMATION METHODS: The estimation methods among sectors in the quarterly PSNA vary depending on the available data for the quarterly accounts and to what extent they are applicable in addressing the national accounts concepts. For the production side of the accounts there are two main techniques used in the quarterly estimation of the gross value added (GVA), namely: Direct Estimates – Some sectors in the PSNA utilize this approach as their estimation method since the data available correspond to the national accounts requirements. For some sectors, quarterly production estimates are available as in the case of agriculture, fishery and forestry; mining and quarrying; and the utilities sector. However, for some sectors production estimates need further elaboration to incorporate the undercoverage of the indicators used for the quarterly estimates. This is done by applying an undercoverage ratio (UCR) suitable to the sector. For lack of information on a quarterly basis, the GVA for most of the sectors are derived by using Gross Value Added Ratio (GVAR) appropriate to each sector. These GVAR’s and UCR’s are computed from the latest Census of Philippine Business and Industries (CPBI) or ASPBI. Derived Estimates (Extrapolation) – Estimates for the manufacturing sector and sectors under Services are derived based on quarterly indicators. These sectors rely on the trend of the quarterly survey 4 of NSO to extrapolate either the output of the sector or its GVA. If the output of the sector is extrapolated, an appropriate GVAR is applied to derive its GVA. IV. COMPILATION OF THE GROSS DOMESTIC PRODUCT: The quarterly PSNA compiles the GDP both by kind of activity and by expenditure items. GDP by kind of activity (production side) is the summation of the gross value added of the different industry groups namely: a) agriculture, fishery and forestry; b) industry; and c) services. Under these main aggregates are the specific sectors so that, the first aggregate includes agricultural crops, livestock and poultry, fishery and forestry. Industry includes mining and quarrying, manufacturing, construction and utilities. Services cover wholesale and retail trade, transportation, communication and storage, finance, housing, and other private and government services. Each sectors has specified sub-sectors that are deemed useful for the users of the national accounts. The expenditure side of the quarterly PSNA is independently estimated and includes the following expenditure items: 1) personal consumption expenditure of households and private non-profit institution, 2) general government consumption expenditure, 3) gross domestic capital formation and, 4) exports and imports from the rest of the world. All estimates are expressed both at current and constant prices. The PSNA still uses the fixed base year method in arriving at the constant price estimates. The GDP level is determined on the production side of the accounts as data support on production is more reliable than on the expenditure side. Hence, a statistical discrepancy is a residual item of the expenditure on GDP. IV.1 GROSS DOMESTIC PRODUCT BY KIND OF ACTIVITY AGRICULTURE SECTOR Data Sources: Quarterly estimates for the agriculture sector rely heavily on both surveys, censuses, administrative-based data and special studies as well. Based on surveys, the Bureau of Agricultural Statistics (BAS) provides quarterly data on volume of production agricultural crops; inventory of livestock and poultry; number of slaughtered animals; milk and egg production; agricultural prices received by farmers; 5 and volume and value of commercial and municipal fishing, aquaculture, and other fishery products. The Philippine Coconut Authority (PCA) provides data on the volume of production of copra and other coconut products including its prices. The Sugar Regulatory Authority, on the other hand, is the source of data for the production, consumption, inventory, and prices of centrifugal sugar and molasses. Data on the production, sales and prices of fiber are sourced from the Fiber Development Authority (FDA). All of the above agencies conduct their respective cost of production studies, which are the sources of information in deriving the gross value added ratio (GVAR) for the different agricultural commodities. Estimation Method: Gross value added (GVA) estimates for agriculture, fishery and forestry are computed similarly both for current and constant price estimates by subsector. Values of production at current prices by subsector are determined based on the reports of respective data-source agencies. The respective GVARs by subsector are then multiplied to the production estimates to arrive at the current price estimates Estimates of GVA at constant prices follow the above approach however, value of production is expressed at base year price using base year unit price of corresponding commodities. MANUFACTURING SECTOR Data Sources: Aside from the censuses and annual surveys, which serve as the basis for the benchmark estimates, the manufacturing sector also utilizes the NSO Quarterly Survey of Philippine Business and Industry (QSPBI), complemented by the Monthly Integrated Survey of Selected Industries (MISSI) and the Wholesale Price Index (WPI). The Integrated Survey of households (ISH) conducted by NSO provides data on employment. Estimation Method: The manufacturing sector has twenty subsectors. The estimation procedure for manufacturing consists of two components, the organized and unorganized sectors. The organized component refers to the large and small establishments covered by the NSO establishment surveys. Benchmark estimate of the gross output 6 of this component is based on the CPBI/ASPBI. GVA at current prices is derived using the production approach. The unorganized component, on the other hand, refers to manufacturing units, which do not possess the attributes of establishment hence, are not covered in the establishment inquiries of NSO. Benchmark GVA estimate is computed using the employment approach. The difference between the employment report from the CPBI/annual establishment survey and the ISH total employment for manufacturing is the estimated employment for the unorganized sector. GVA is estimated by applying the GVA per worker of small establishments based on the results of the latest CE/ASPBI. The total GVA at current prices is simply the total of the GVA of the organized and unorganized sectors. The benchmark annual GVA estimates are disaggregated into quarters by applying the index of production derived from the QSPBI. Preliminary current price quarterly estimates are derived by extrapolation based on production trends of the QSPBI and MISSI. Deflating GVA at current prices by the corresponding WPI derives GVA at constant prices. TRADE SECTOR Data Sources: The primary data source for the trade sector is the NSO. The CPBI and ASPBI provide information on receipts, employment, compensation, costs, inventories and fixed assets of trade establishments. The ISH provides information on employment while the WPI and the Retail Price Index (RPI) serve as the deflators. Estimation Method: Estimation of GVA at current prices utilizes the production approach and similarly follows the method used in the manufacturing sector wherein the organized and unorganized sectors are estimated separately. Preliminary quarterly estimates at current prices are derived by applying the trend of gross revenues from the quarterly survey to the previous quarter’s GVA level. To arrive at the GVA at constant prices, current price estimates are deflated using as the deflator the implicit price index (IPIN) of the sector, which is updated 7 based on the corresponding price indices. The IPIN is simply equal to the GVA at current prices divided by the GVA at constant prices multiplied by one hundred. OWNERSHIP OF DWELLINGS & REAL ESTATE SECTOR Data Sources: The real estate sector relies on the results of the CPBI, ASPBI, QSPBI and the CPI as provided by the NSO. The financial reports of the Commission on Audit (COA) provide information on the rental activities of government corporations. Ownership of dwellings, on the other hand, utilizes the Census of Population and Housing (CPH), Family Income and Expenditure Survey (FIES), Building Permits and the CPI as compiled by NSO and likewise the reports on damages by the National Disaster Coordinating Council (NDCC). Estimation Method: For real estate, annual benchmark estimate for output at current prices is based on the results of the CPBI/ASPBI and rental income of households is based on the FIES. The production approach is utilized to derive the GVA at current prices. Benchmark GVA is disaggregated into quarters using the gross revenue index computed from the QSPBI. Preliminary GVA estimates is derived by extrapolation using the trend of the QSPBI. For ownership of dwellings, benchmark quarterly GVA at current prices is estimated by multiplying the gross output, which is the imputed rent, by the GVAR for owner-occupied dwellings. Imputed rent of owner-occupied dwellings are derived by applying the average rental expenditure per dwelling unit to the estimated number of existing dwellings adjusted by the quarterly price index (CPI). The FIES provides the benchmark estimates of average rental expenditure of families and the GVAR, while the CPH provides the benchmark estimates of owner-occupied dwellings. For preliminary current price estimates, quarterly GVA for real estate is derived by extrapolation based on the trends of gross revenue from the quarterly survey. For ownership of dwellings, preliminary current price GVA is extrapolated based on the trend of the stock of residential structures. Stock data is updated using the trend of the Building Permits and reports on damages. In estimating GVA in real terms for both the real estate and ownership of dwellings, the current values are deflated using the CPI as deflator 8 OTHER SECTORS The estimation methods for the other sectors of the production accounts are summarized in a table as part of the attachment. IV.2 EXPENDITURES ON GROSS DOMESTIC PRODUCT Most of the expenditure items on GDP rely on administrative data for the quarterly estimates. General government consumption expenditure and capital formation on public construction depend on government statistics from the Department of Budget and Management. Merchandise exports and imports employ the direct estimation method based on the tabulation of the NSO Foreign Trade Statistics whose basic data come from the Bureau of Customs. Likewise, non-factor exports and imports use the Balance of Payment (BOP) data except for travel. PERSONAL CONSUMPTION EXPENDITURE (PCE) Data Sources: Data for PCE comes from a variety of sources since there is no quarterly household expenditure survey. Data for production comes from the quarterly survey of NSO and likewise from the surveys and administrative reports of data source agencies for the production side of the accounts. Other data from the NSO includes the Foreign Trade Statistics (FTS) and the Family Income and Expenditure Survey (FIES). Estimation Method: The Philippine national income accounts series presents personal consumption expenditure by purpose such as food; beverages; tobacco; clothing and footwear; fuel, light and water; household furnishings; household operations; transportation and communication; and miscellaneous expenditures. Generally, personal consumption expenditures are recorded using the commodity flow approach, where the disposition of the produced goods and services to intermediate consumption and other final users are subtracted and PCE then is derived as a residual. Exports, which include declared purchases made by tourists are likewise considered in the use flow. This is of course undervalued by the amount of undeclared purchases made. On direct purchases made by Filipino residents abroad, 9 necessary adjustments on final expenditures are not being made for lack of information. Benchmark estimates and expenditure patterns are derived on based on the FIES. Constant price estimates are derived by deflating the current expenditures by the appropriate CPI component. GENERAL GOVERNMENT CONSUMPTION EXPENDITURES (GGCE) Data Sources: GGCE relies heavily on administrative data with reports coming from the Department of Budget and Management, the Commission on Audit, the Government Service Insurance System and the Social Security System. Estimation Method: General Government Consumption Expenditure on goods and services is equivalent to the value of the goods and services, which they produce. It includes compensation of employees and purchases of goods and services less their sales to households and industries. Compensation of employees accounts for the remuneration of general government employees including social security contributions of the government funds (or in lieu, the pensions actually paid). The remuneration of members of the armed forces, besides their regular salaries, includes allowances for food and standard clothing and housing. Purchases by general government comprise: (a) purchases for military purposes, which include expenditures on arms, ammunitions, aircraft, road vehicles, ships, new buildings for military use and military service, and (b) purchases for civil administration, for governing purposes. The latter includes the cost of operating services such as schools, hospitals, social service, police forces, roads and other transport installations including gross rent of buildings for government use. Rental expenditures are also imputed on buildings owned and occupied by government. Imputed rent is only estimated in respect of buildings such as office premises, schools and hospitals but not in respect of historical buildings, museums, etc. Provisions for the depreciation of government buildings are included in the imputed gross rents. No imputed interest or depreciation is charged in respect of other government construction such as roads and traffic installations because of practical difficulties. 10 In determining the scope of general government consumption expenditure, it is necessary to decide which purchases are to be treated as consumption expenditure as distinct from capital formation. In this connection, expenditure for defense purposes, excluding civil defense is treated as consumption expenditure, whereas all expenditure on capital formation for civil defense is included in gross domestic capital formation. For constant price estimates, compensation of employees is deflated using an average quarterly compensation index derived from the data of the DBM. The maintenance and other operating expenses use the implicit price index (IPIN) of construction, private services and freight and transportation services. The IPIN is simply derived by dividing the current price estimates by the constant price estimates. EXPORTS AND IMPORTS OF GOODS AND SERVICES Data Sources: Data for merchandise exports and imports come from the Foreign Trade Statistics being compiled by NSO. For non-factor exports and imports the primary data source is the Balance of Payments being compiled by the Bangko Sentral ng Pilipinas (Central Bank of the Philippines) and the travel data from the Department of Tourism (DOT). The compilation of the Philippine BOP has adopted the Balance of Payment Manual 5 (BPM5) starting CY2000. However, there are still some items in our BOP that can not be readily use for national accounting due to the limitations of its source of data. Estimation Methods: Exports and imports of goods and services consist mainly of transactions between the residents of a given country and the rest of the world. These transactions are categorized into the following: Commodity Export/Import consists of shipment of goods plus the monetization or demonetization of gold; and Non-factor services, which include: o transport, communication and insurance services; o direct purchases of foreign government and extra-territorial bodies such as foreign embassies, international organizations or foreign armed 11 forces in the domestic market of a country are included in its exports of commodities; o travel items consisting of direct purchases in the domestic market by non-resident households and direct purchases abroad by resident households (i.e. tourist expenditures); and o miscellaneous expenditures including (1) reimbursement of the cost of home office services of parent company from foreign branches and subsidies; (2) other services such those related to construction; (3) subscription and cable charges; and (4) film and real estate rentals. Estimates for non-factor exports and imports are based on the BOP except for travel, which is based on the travel data of DOT. In principle, all transactions should be recorded at the moment when ownership of, or legal title to, the goods passed between buyer and seller. However, for purposes of estimation, exports and imports of merchandise are based on Foreign Trade Statistics (FTS), which record the physical movement of goods across boundaries. Adjustment is made to exclude special transactions (gifts, samples, returned goods, leased equipment) from FTS to conform with the national accounting concept. Merchandise exports and imports are expressed in real terms using the base year unit prices of commodities. Non-factor exports are deflated by applying the appropriate domestic price index (WPI or CPI). On the other hand non-factor imports are deflated using the weighted average CPI of selected trading partners of the Philippines adjusted for exchange rates. OTHER SECTORS The estimation methods for the other sectors of the expenditure accounts are summarized in table as part of the attachment. V. Reconciliation of the Quarterly and Annual National Accounts Quarterly estimates are always made consistent with the annual estimates. For the preliminary estimates no adjustments are made since the annual estimates are simply the summation of the four quarters. However, separate annual estimates are 12 prepared once a more reliable data set is available. The quarterly estimates are always reconciled if every time the latter is revised. Two methods are adopted in reconciling the two sets of accounts based on the availability of data: a) quarterly estimates are adjusted based on the quarterly revisions of the basic data and; b) applying a simple pro rata adjustment based on the quarterly distribution of output or the quarterly distribution is based on the quarterly indicators. Other methods were also explored like the Bassie Method however, considering the revisions in the data and the relevance of the quarterly accounts the above two methods were deemed practical. VI. Revision of the Quarterly National Accounts The updating or revision of the quarterly national accounts depend on the availability of data requiring a large set of information from various sources, covering data generated from surveys as well as those compiled from administrative records. These sources of data have varying degrees of timeliness and follow different revision schedules. As such NSCB Resolution No. 8-97 approved the updating of the quarterly accounts for each quarterly estimation round to be limited to the immediately preceding quarter, and for the rest of the past quarters to be done only during the May round of estimates. VII. Dissemination of the quarterly accounts NSCB has adopted the policy of releasing in advance (before the year ends) the schedule of press release of the national accounts for the succeeding year. NSCB Resolution No. 9-97 stipulates that the quarterly national accounts be released 60 days after the reference period however, the advance annual and fourth quarter national accounts estimates shall be released 30 days after the reference period. Hence, prior to the start of the year, there are already specific dates for the release of the quarterly national accounts and data source agencies are made aware of these. To enhance the transparency of the quarterly national accounts, a press conference is held at the NSCB every time the quarterly accounts are released. Likewise, on this date the accounts are also made available through the NSCB website. Hard copies of the accounts are given free for those attending the press conference and NSCB also keeps a mailing list for the distribution of the accounts. 13 Copies for the accounts can be secured through our National Statistics Information Center (NSIC), a one-stop center for statistical information developed and maintained by NSCB. VIII. Problems/Issues in the Compilation of the quarterly accounts The quarterly national account has its own peculiar problems/issues some of which are briefly discussed below: Timeliness of data – Despite the coordinative effort of NSCB in ensuring the availability of quarterly data for the quarterly national accounts, the timeliness of some data still posed a problem from time to time. There are instances wherein response rate is low; there is incomplete quarterly data (one or two months of the quarter is not available) or in some cases, voluminous administrative based data are not compiled in time for the estimation period. The NSCB is presently strengthening its institutional linkages with the data source agency through memoranda of agreement or formal institutional letters. The NSO, on the other hand, has recently embarked on a project to improve its annual surveys. Parameters/Coefficients used – The parameters /coefficients used in the quarterly national accounts were based or derived from the results of annual surveys, censuses or special studies. As such, these parameters/coefficients need to be updated frequently to be more relevant. However, results of annual surveys and censuses are available some years from the reference year and special studies are not readily updated. Given this situation, there is a need to review the sectoral methodologies vis-à- vis the present availability of data. Other data sources will be tapped for the updating of the parameters. Allocation of outputs to quarters – For some sectors, output is easily defined. However, in the case of agriculture and construction these sectors have a long production cycle such that it is quite difficult to allocate their outputs into quarters. 14 IX. Future Directions: With the advent of the 1993 SNA, NSCB faces the challenge of carrying on the adoption of the new SNA to its quarterly accounts. Last year (CY 2000), NSCB has drawn up its Revision Program for the PSNA. This program of revision is comprehensive since it involves the improvements in the estimation methods and at the same time will incorporate the recommendations of the 1993 SNA to the extent possible. Phase one of the program is scheduled to start this September 2001. This involves revision of the 1991-1998 series of the national accounts and likewise the preparation of a set of volume measures using the Chain Volume Measure approach. With this revision, we also plan to develop a computerized compilation system for the quarterly PSNA that will make the accounts more transparent and will minimize if not eliminate estimation errors. We look forward to a Revised PSNA that is more relevant to the changing times 15 ATTACHMENT 16
"Philippine Quarterly National Accounts"