Report on Eurostat Quality Visit to the Portuguese Statistical Office

Report on Eurostat Quality Visit to the Portuguese Statistical Office

TABLE OF CONTENTS

1.Purpose of the visit

2.Institution

3.Survey

4.Relevance

5.Definition of the target population

6.Sampling Frame

7.Sampling design

8.Product nomenclature

9.Method of data collection

10.Confidentiality

11.Measurement errors

12.Processing errors

13.Non-response errors

14.Data Processing and Quality Assurance

14.1.Data Processing

14.2.Quality Assurance

14.3.Missing data

15.Dissemination

1.Purpose of the visit

The Eurostat PRODCOM team is conducting a number of "Quality Visits" to the PRODCOM teams in the NSIs, in order to discuss the methodology used, detect potential problems and identify best practice that can be shared with other countries.

After each visit, Eurostat will prepare a report on the discussions and, after checking it with the team visited, will load it to the PRODCOM CIRCA website, together with any material provided by the team in the NSI.

2.Institution

The principles rules and structures of the National Statistical System (NSS) have been enforced by law (N° 22/2008 of 13 May 2008).

The Portuguese Statistical System is governed by a Statistical Council which includes data producers and data users. It approves the strategic directions for NSS and fixes the annual work plan.

A Decree-law N°16/2012 (2 July 2012) specified the organisation of Statistics Portugal. It is divided in 6 departments. There are 3 horizontal departments: Administration and management; Methodology and Information systems; Data collection. 3 statistical departments: Demographic and Social statistics; Economic statistics; National accounts.

3.Survey

The IAPI (Inquérito anual à produção industrial – Annual survey on industrial production) is an annual survey conducted by Statistics Portugal in order to collect PRODCOM data.

At NACE level, the enterprise provides the following variables:

–Total value of sold production (VSP)

–VSP in the domestic market

–VSP in other EU countries

–VSP in the third countries

–Industrial services

–Part of the industrial services that result from sub contracting.

For each product manufactured the enterprise provides the following variables:

–Total production

–Sold production

–Self consumption production

–Value of Sold production

The enterprises are required to provide information on industrial consumption of raw materials by NACE (every 5 years from 2011 except for a restricted number of NACE that will be still collected every year):

–Total costs

–Costs from the domestic market

–Costs from other EU countries

–Costs from the third countries

The enterprises are required based on a pre-filled list of raw materials (PRODCOM + 4 digits) to provide:

–Purchases

–Consumption

–Costs

Finally, enterprises are required based on a list of industrial consumption of Energy (PRODCOM + 4 digits) to provide

–Consumption

–Costs

4.Relevance

The main users of PRODCOM data can be presented in different groups, namely:

5.Definition of the target population

All the enterprises that produce industrial goods and industrial services on the Portuguese territory during the reference year are targeted by the survey.

6.Sampling Frame

The Business Register is used to select enterprises.

The business register is based on administrative sources (Ministry of Justice - National Registry of Legal Persons; Ministry of Finance - Corporate Tax- Income Tax; Ministry of Labour and Solidarity - Personnel Files).

Each year a Structural population frame is built:

–All active enterprises are included

–All enterprises with activity cessation in reference period are included

–All Institutional sectors are included.

The sampling frame will be obtained from this Structural population frame.

The PRODCOM frame population includes all enterprises with their primary or secondary activities in sections B, C and D (only class 3511, 3521, 3530, 3831 and 3832) of NACE Rev.2.

7.Sampling design

There are a lot of small and medium-sizedenterprises in the Portuguese economy. 98% of enterprises in the Structural population frame are SMEs (which means that 1% of the turnover value of a particular NACE could include several hundreds of enterprises). The rule that only enterprises of 20 employees or more should be surveyed, laid down in the PRODCOM regulation, is not applicable to Portugal.

Statistics Portugal has setup a number of rules to build their sample:

-To minimise the confidentiality issue the sample is using at least 5 enterprises by NACE (whenever possible).

–Enterprises that have reported product a year are automatically inserted in next year sample.

–Enterprises are sorted by descending order of turnover. They are included in the sample until the 90% of turnover is achieved.

The size of the sample is about 16 000 enterprises:

8.Product nomenclature

The PRODCOM nomenclature has been extended to cover needs from National accounts. The annual survey on industrial production also collects data onindustrial consumption of energy, to cover needs from SBS (V20110 Purchases of energy products).

The National accounts are using a classification of 433 Products. They are derived from PRODCOM but 65 national breakdowns have been added to the PRODCOM classification. The areas where breakdown have been added are:

–Wines, wood, wood pulp – EAA and EAF (ie Red and white wines, Port / Madeira wines, Protected designation of Origin / Protected Geographical Indication).

–Oil products (Not surveyed in PRODCOM normally- Jet fuel, Gas oils, Fuels oils, Propane and butane liquefied)

–Pesticides (Agricultural and non-agricultural uses)

–Wiring (used in vehicles / others)

As the PRODCOM survey is used as a vehicle to collect other information, some additional codes are part of the National PRODCOM list. 38 codes are included to collect information on raw materials use. 25 codes (types of energy products) are also used for SBS survey purposes.

When data are collected the users get a predefined list of products but have the possibility to specify a product that is not in the list. Then the Data Collection Team intervenes to help with the classification process. As his task is quite time consuming efforts are made to reduce it.

9.Method of data collection

The data are collected via a web interface (WebInq). If an enterprise is part of the sample, several questionnaires have to be filled out, one by each NACE.

1. Form at NACE level for information on Sold Values and the area of sales (domestic, EU, third country) at Product level (real European PRODCOM requirement);
2. For each NACE, the form also includes data concerning industrial raw materials and industrial consumption of Energy.

On the forms, enterprises have access to a tab by Kind of Activity Unit. The list of all products from this KAU is presented.

Enterprises are encouraged to use the Web questionnaires. If an enterprise has difficulties, guidance is providedby the data collection team. Only if the guidance is not successful or not accepted, paper forms can be downloaded by the enterprise and filled out manually.

The identification and authorisation of the respondents is based on the provision of master passwords to the enterprises by registered mail. Through these passwords, the representatives of the enterprises can manage the delegations of respondent privileges to the accounts of their choice. Accounts can be primary or secondary respondents. Primary respondents receive alerts when secondary respondents submit data. This system minimises the management load.

10.Confidentiality

The confidentiality of data is protected by a national law.

Portugal is the only Member State to use a dominance percentage of 100%.

In 2011, Statistics Portugal reported 2240 active products, of which 426 were signalled as confidential.

11.Measurement errors

Statistics Portugal has identified following measurement errors:

• Wrong industrial product code;
• Classification of new products;
• Wrong volume unit;
• Making a clear distinction between production and industrial services;
• Making a clear distinction between production and trade;
• Making a clear distinction between units produced and units sold;

The errors are mitigated by direct and close contact with respondents. Statistics Portugal is providing guidance and support to respondents.

The webforms are including more and more validation rules and alerts to guide the respondents.

12.Processing errors

An integrated system is used to collect data, which enables cross checking data credibility. During the credibility analysis the following checks are performed:

• Code validation
• Units of Measurement
• Comparison between indicators
• Comparison between reporting units
• Link with other surveys (consistency analysis)

Examples of credibility checks that are implemented:

–Product value is +/- 30% and +/- 150.000 € compared to previous return

–Unit value is +/- 25% compared to previous return

–Unit value is +/- 25% away from aggregate unit value

13.Non-response errors

Statistics Portugal is applying an automatic non response adjustment for the first reporting (June Year +1). Non response is replaced by information from previous years.

For data provided in December Year+1 and December Year+2 non responses are adjusted only in case data might have significant impact on national figures. In such cases the sources of information for adjustment are:

–PRODCOM data from previous year

–Imputation based on average values by NACE, for the reference year

–STS data

–Simplified Business Information data

–International Trade Statistics data

14.Data Processing and Quality Assurance

14.1.Data Processing

The data processing system implements an integrated and consistent approach to the data collection processing system.

The integration, harmonisation and reuse of production processes provide efficiency gains. It increases the comparability by using the same population for all business surveys. It aims to reduce the statistical burden for respondents by requesting common information needed in various surveys only once.

The same system is used directly by respondents over internet (Electronic form) and inside Statistics Portugal (for paper responses and editing of provided answers).

14.2.Quality Assurance

The system implements a version stack of answers. When data editing is done a new response version is created. All versions of the responses are saved and can be audited.

Only the last version, when micro data have been validated, is transferred to the data warehouse environment.

14.3.Missing data

From April to October the respondents are reminded by email (at least three). When needed they are called by phone. Some enterprises (without email – or of specific interest for Statistics Portugal) are contacted by letter or even visited.

The response rate is monitored based on number of enterprises from the sample and percentage of Value of Sold Production. The results are shown in the chart below:

15.Dissemination

Statistics Portugal is producing an annual publication on Industrial production. The publication presents the main findings of the Annual Survey on Industrial Production (PRODCOM). Results are released for a large set of products produced (expressed in volume), sold products (in volume and value) and industrial services provided. The latest publication concerned 2010 data and was released in January 2012:

IndustrialProduction Statistics - 2010 .

The data are accessible through the Statistics Portugal website as statistical indicators:

Sales of products (€) by industry by Product type (By CAE Rev. 3); Annual

1