Date Released: August 24, 2007
REBASING OF THE MISSI INDICES FROM 1994 TO 2000 MONTHLY INTEGRATED SURVEY OF SELECTED INDUSTRIES FOR MANUFACTURING 1. INTRODUCTION The predecessor of the Monthly Integrated Survey of Selected Industries (MISSI) is the Survey of Key Establishments in Manufacturing or SKEM. Through the years of its operation, the SKEM and later on the MISSI continued to utilize the shuttle type questionnaire for data collection. The MISSI timeline is shown below:
2. FEATURES OF THE 2000 BASED MISSI SERIES E.O. No. 352 lists the MISSI as one of the designated statistical activities being undertaken by the NSO. The basic features of the 2000 based MISSI indices are presented as follows: 2.1 Base Year is 2000 Based year values of the variables Value of Production and Value of Net Sales are 2000 based. 2.2 Industry groupings The industry groupings used in the 2000 based series follow the same groupings utilized in the 1994 based MISSI series. There are 20 major sectors, and eight (8) of these are presented by sub-sectors. 2.3 Industry weights Weights of major sectors and sub-sectors were obtained from the results of the 2000 Census of Philippine Business and Industry (reference year 1999). 2.4 Selection of Sample Establishments Sample establishments for the 2000 rebased series were taken from the list of sample establishments covered in the 1994 based MISSI series. Additional sample establishments were included in later years. 2.5 2000 Based Indices The indices with 2000 as the base year include the Value of Production Index (VaPI) and Volume of Production Index (VoPI), as well the following new index series: Value of Net Sales Index and Volume of Net Sales Index. 2.6 Index Methodology Index computation in the 2000 based series follows the same methodology utilized in the 1994 based series. 2.7 Imputation Method Imputation is resorted to for missing data. Missing data in the MISSI are mainly due to late reporting by sample establishments. In a few cases, imputation is also done for problem data that need verification from the sample establishments. The imputation method used was historical imputation without trend adjustment, that is, the latest available data of the sample establishment was used to replace the missing monthly data. However, for the 2000 based MISSI series, there are very few cases of imputed data at the industry level as the response rates are rather high, at nearly 95 percent. 3. STEPS IN REBASING OF MISSI INDICES The following outlines the procedures in the rebasing of the MISSI indicators: 3.1 Identification of Base Year The rebased MISSI index series will have the year 2000 as the new base year. This means that base year values of the variables are 2000-based. 3.2 Determination of Indices for Rebasing The indicators to be generated in the 2000 based MISSI series include the following: New Index Series: The last two indices, Net Sales Value and Volume Indices, are new indicators in the 2000-based series. In the 1994 based series, only year-on-year and month-on-month growth rates were generated from the Net Sales value and volume data. 3.3 Determination of Sectors and Sub-sectors The sectors covered in the MISSI and PPI for the 2000 based series follow the same industry groupings utilized in the 1994 based MISSI, that is, there are 20 major sector groups which are in conformance with the National Income Accounts groupings. 3.4 Determination of Weights VaPI and VoPI Three options were initially considered in the determination of weights for the 2000 based VaPI and VoPI, as follows: data from the 2000 Census of Philippine Business and Industry (CPBI- reference year 1999), data from 2002 Annual Survey of Philippine Business and Industry (ASPBI- reference year 2001), and the average of the 2000 CPBI and 2002 ASPBI data.
The final choice was to utilize weights for the major sectors and sub-sectors from the 2000 CPBI (Option 1 above). Data on Value of Production for the manufacturing establishments with ATE 20 and over from the 2000 CPBI data (reference year 1999) were used in deriving the weights for the major sectors (3-digit PSIC). On the other hand, data on Total Revenue for manufacturing establishments with ATE 20 and over were used in deriving weights for the sub-sectors (4-digit PSIC). Since it is assumed that the structure do not change very much over a short period of time, the 2000 CPBI data was used for the derivation of the weights for the 2000 based indices. Value of Production is derived by adjusting the value of products sold for changes in inventories of products and work-in process (ending less beginning). This indicator differs from the Value of Output as other revenue items like industrial services and fixed assets produced on own account are not considered. Net sales volume and value indices Weights for these indices for the major sectors (3-digit PSIC) were taken from the 2000 CPBI data on Value of Products Sold for manufacturing establishments with ATE 20 and over. Value of Products sold is one of the components of Total Revenue. On the other hand, data on Total Revenue for manufacturing establishments with ATE 20 and over were used in deriving weights for the sub-sectors (4-digit PSIC). Refer to Appendix 1 for list of major sectors and sub-sectors and their corresponding weights for VaPI, VoPI and Net Sales value and volume indices, respectively. 3.5 Selection of Sample Establishments Sample establishments for the 2000 based series were taken from the list of sample establishments of the 1994 based MISSI series. This was done to maintain consistency in the composition of samples for the two base periods. The criteria for the selection of samples are as follows: New sample establishments are those that were included as new samples during the particular year, regardless of the year these establishments first started operation. Excluded as samples in the 2000 based series are the following: Below is the comparative distribution of the number of sample establishments for the 1994 and 2000 base years. Number of Sample Establishments by Year
Refer to Appendix 2 for the distribution of number of sample establishments by major sector and sub-sector for the years from 2001 to 2006. 3.6 Computation of Base Year Values Variables for the determination of average base year values are Value of Production and Value of Net Sales. Average base year values of these variables for the year 2000 were computed for each of the sample establishments by aggregating all the monthly values of these variables and divided by 12. The 2000 monthly base year values consist of the actual data reported by the sample establishments and were used in the 1994 based series. Data of late reports received after the final release of the 1994 based series were also used in the computation of the 2000 average base year values. 3.7 Methodology for Index Computation The monthly indices on production and sales are presented by selected industry groups following the 3-digit (sectoral) and 4-digit (sub-sectoral) groupings of the 1994 PSIC. Of the 20 major sectors in the MISSI, eight (8) industry sectors, namely, food, textile, wood and wood products, chemicals, petroleum products, non-metallic mineral products, basic metals and electrical machinery have corresponding sub-sectors while the rest have none.. The procedures used in the computation of the indices are as follows: Sub-sectoral Level (4-digit PSIC)
The formula above was modified to a chained index formula to allow for changes in the composition of the sample establishments under each sub-sector, as follow:
Sectoral Level with no corresponding sub-sectors (3-digit PSIC)
The formula was modified to a chained index formula to allow for changes in the composition of the sample establishments under each sector, as follow:
Sectoral Level with corresponding sub-sectors (Food, Textile, Wood and Wood Products, Chemicals, Petroleum Products, Non-Metallic Mineral Products, Basic Metals, Electrical Machinery)
Total Manufacturing
4. DISSEMINATION OF RESULTS Results of the 2000 based MISSI series are disseminated as Special Release at the NSO website covering the period 2001 to 2006. Succeeding releases of the 1994 based MISSI series will include the 2000 based series for the year 2007. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
REBASING OF THE PPI DATA SERIES FROM 1994 TO 2000 PRODUCER PRICE INDEX 1. INTRODUCTION The National Statistics Office (NSO) generates the monthly Producer Price Index (PPI) for the Manufacturing ahead of the indicators generated from the MISSI such as the Value of Production Index (VaPI), Volume of Production Index (VoPI), capacity utilization rate, among others, to better measure the performance of the industry. 2. HISTORICAL BACKGROUND PPI timeline is presented below:
3. FEATURES OF THE 2000 BASED PPI 3.1 New Base Year is 2000 The new PPI series will have the year 2000 as the new base year. This means that base year values of the producer prices of commodities are for the year 2000. 3.2 Industry groupings The industry groupings in the 2000 based PPI follow the groupings used in the 1994 based series, that is, the groupings correspond to industry groups (3-digit) and selected industry classes (4-digit) of the 1994 PSIC. 3.3 Industry Weights Since
the PPI is a Paasche type index, ideally, it should use current year
weights instead of base year weights. Due to the time lag in the
availability of data that will serve as current year weights, data
on Value of Production from the latest available annual
survey or census of manufactures were used to derive the weights for
the PPI. Value of Production is derived by adjusting the
value of products sold for changes in inventories of finished
products and work-in-process (ending less
beginning).
The 2000 based PPI used the same weights utilized in the yearly series of the 1994 based PPI, as follows: year 2006 : weights were obtained from 2003 ASPBI2005 : weights were obtained from 2001 ASPBI 2004 : weights were obtained from 1999 CPBI 2003 : weights were obtained from 1998 ASPBI 2002 : weights were obtained from 1998 ASPBI 2001 : weights were obtained from 1997 ASPBI Refer to Appendix 3 for list of major sectors and subsectors and their corresponding weights by year. 3.4 Selection of Commodities and Sample Establishments Sample commodities included in the 2000 based PPI basket were taken from the list of sample commodities of the 1994 based series. This was done to maintain consistency in the composition of the sample commodities for the two base periods. Additional sample commodities were, however, included in selected industries such as in cement, fabricated metals, leather products, microcircuits, petroleum products. The list of commodities in the 1994 based series were updated to include the following sample commodities for the 2000 based series: Generally, all sample commodities in the 1994 based series with reported data for all or most months in 2000 to the present were selected as samples. Commodities in the 1994 based series with reported data (from sample establishments) for all or most months of 2000 and 2001 were also taken as samples, regardless of whether these commodities ceased to be produced after 2001. New sample commodities in the 1994 based series that were included starting 2001 (or later) were also taken as samples, provided these commodities have reported data for at least two consecutive years. New sample commodities are those that were included as new commodities in the PPI basket during the particular year, regardless of the year these commodities were first introduced in the market. Sample commodities which were excluded are the following: Sample commodities with reported data for only one year. Sample commodities which were produced by establishments and thereafter have closed, ceased operation, on strike, and cannot be located. Below are the comparative distributions of the number of sample establishments for the 1994 and 2000 base years. Number of Sample Commodities by Year
Number of Sample Establishments by Year
Refer to Appendices 4 and 5 for the distribution of sample commodities and sample establishments by major sector and sub-sector, respectively. 3.5 Data Source Producer prices of commodities for the PPI are obtained from the Producer Price Survey (PPS) conducted monthly by the NSO. The respondents for the survey are selected manufacturing establishments that produced the selected commodities included in the basket. Period of data collection starts on the 15th day of the reference month and ends on the 30th day of the month after the reference month. The PPS uses a shuttle type questionnaire that is distributed by the NSO field staff at the beginning of the year to all the sample respondents. This type of questionnaire provides the respondents, as well as NSO staff, with a running account of the previous month's data for one year. 3.6 Imputation Method Imputation is resorted to for missing data. In many cases, missing data is due to late reporting by the respondents. For practical purposes, the imputed producer price of a commodity is based on the latest available monthly producer price data provided by the respondent, that is, with no trend adjustment. This method is appropriate for the reason that the prices of a number of commodities do not change very much over a period of time, except for volatile products, e.g. petroleum products. For the 2000 based PPI, there are only minimal cases of imputed values as the PPS had rather high response rate, at more than 95 percent as of the final release date of the PPI. 4. COMPUTATION OF BASE YEAR PRODUCER PRICE Average base year producer price is computed for each of the sample commodities by aggregating all the monthly producer prices for 2000, and divided by 12.
The 2000 monthly producer prices of a commodity consist of the actual price data reported by the sample establishments, as well as the imputed data for the non-responses, which were used in index computation in the 1994 based PPI series. If the actual price data for the imputed data are already available (that is, these were received later after the release of the final data of the 1994 based PPI), these were used in the computation of the 2000 base year PPI. Most, if not all, of the base year values in the 2000 based PPI are actual data of producer prices of commodities that were reported by the sample respondents. 5. METHODOLOGY IN THE COMPUTATION OF THE PPI The computation of the 2000 based PPI follows the same methodology used in the 1994 based series. imputed data for the non-responses, which were used in index computation in the 1994 based PPI series. If the actual price data for the imputed data are already available (that is, these were received later after the release of the final data of the 1994 based PPI), these were used in the computation of the 2000 base year PPI. Most, if not all, of the base year values in the 2000 based PPI are actual data of producer prices of commodities that were reported by the sample respondents. The industry class level (4-digit PSIC) The PPI at the 4-digit PSIC is simply the unweighted arithmetic mean of the price relatives of the commodities representing the industry class and multiplied by 100.
The formula for PPI ij at the 4-digit PSIC was modified to a chained index formula, that is, the average monthly price relative of the commodities in the industry class is multiplied by the previous month PPI at the 4-digit PSIC to obtain the current month PPI. That is,
The chained index method is especially useful when new commodities enter into market basket as replacements or additions. The base year prices of these commodities need not be obtained as the alternative then is to make available the previous month producer prices of the new commodities to allow the computation of their corresponding price relatives, Phm/Phm-1. The industry group and division levels (3-digit and 1-digit PSIC) The Producer Price Index for manufacturing at the 3- and 1-digit PSIC is obtained as the weighted harmonic mean of the component price indices.
6. DISSEMINATION OF RESULTS Results of the 2000 based PPI series are disseminated as special Release at the NSO website covering the period 2001 to 2006. Succeeding releases of the 1994 based PPI series will include the 2000 based series for the year 2007. Source: National Statistics Office Manila, Philippines Page last revised: August 24, 2007 |
