The objective of the Retail Trade Index is to provide information on the fundamental characteristics of the companies engaged in retail trade in Spain, which allow the evolution of the sector's activity to be measured over the short term.
This operation is framed within the scope of Regulation nº 1158/2005 of the European Parliament and Council of 6 July 2005, amending Regulation 1165/98 of the European Union Council of 19 May 1998 concerning short-term statistics.
This is a statistic with direct data collection. Results are presented as indices with the aim of measuring variations using the base year 2015 as a reference.
The RTI studies the population formed by companies whose main activity is registered in division 47 of section G (Trade) of the National Classification of Economic Activities (CNAE 2009).
The basic statistical unit is the company that carries out any of the activities included in the population scope as the main economic activity.
The company is also the respondent unit, as it is perfectly defined and located, and having the accounting and employment data available facilitates the response and provides homogeneous information.
The target population is made up of companies whose main activity is included within division 47 of the CNAE-2009 'Retail trade, except of motor vehicles and motorcycles'.
All statistical units located in the Spanish territory form the object of the research.
In 2018, the publication of RTIs using base 2015 began
Some of the main series that can be consulted are:
Original series of national turnover indices at current prices and at constant prices of:
The series similar to the previous ones but corrected for calendar effects and corrected for seasonal and calendar effects can be consulted from the same date as the original series.
Original series of turnover indices by Autonomous Communities and Cities of:
Employment series of:
In 2018, the indices begin to be published using the new base 2015, in order to comply with the requirements established in Eurostat's Regulation 1165/98 on short-term statistics.
The RTI uses chain-linked indices so that:
The unit of measure is the indices and their rates of variation (monthly, annual and year-to-date average).
The turnover reference period is the calendar month.
The employment reference period is the employed persons on the last day of the month.
Data referred to the period: Mensual A: 2019 MES: 01
The compilation and dissemination of the data are governed by the Statistical Law No. 12/1989 "Public Statistical Function" of May 9, 1989, and Law No. 4/1990 of June 29 on “National Budget of State for the year 1990" amended by Law No. 13/1996 "Fiscal, administrative and social measures" of December 30, 1996, makes compulsory all statistics included in the National Statistics Plan. The National Statistical Plan 2009-2012 was approved by the Royal Decree 1663/2008. It contains the statistics that must be developed in the four year period by the State General Administration's services or any other entity dependent on it. All statistics included in the National Statistics Plan are statistics for state purposes and are obligatory. The National Statistics Plan 2017-2020, approved by Royal Decree 410/2016, of 31 October, is the Plan currently implemented. This statistical operation has governmental purposes, and it is included in the National Statistics Plan 2017-2020. (Statistics of the State Administration).
This operation is intended to comply with Regulation No 1158/2005 of the European Parliament and Council of 6 July 2005 amending Regulation No 1165/98 of the European Union Council of 19 May 1998 concerning short-term statistics, although the latter is still in force.
The aim of these Regulations is to create a common framework for the production of EU statistics on the short-term evolution of supply, demand, production factors and prices.
Through the collaboration agreements signed with the regional Statistics Institutes, a monthly exchange of information takes place, complying with the established requirements to preserve statistical confidentiality.
All statistical personnel have the obligation to preserve statistical confidentiality (art. 17.1 of the Law on the Public Statistics Function - LFEP). In the specific case of this survey, only aggregated information is published.
All information related to this index is published on the INE website. The following publications can be found on the websit:
All these publications can be consulted at the following link:
INEbase is the system the INE uses to store statistical information on the Internet. It contains all the information the INE produces in electronic formats. The primary organisation of the information follows the theme-based classification of the Inventory of Statistical Operations of the State General Administration . The basic unit of INEbase is the statistical operation, defined as the set of activities that lead to obtaining statistical results on a determined sector or subject based on the individually collected data. Also included in the scope of this definition are synthesis preparation.
Within the Retail Trade Index, indices and different rates of variation of sales and employment can be found for different levels of disaggregation by group, mode of distribution and geography. A distinction is made between index series of sales at current prices, at constant prices, adjusted for calendar effects and corrected for seasonal and calendar effects, as well as employment series. In addition to these series, the following link shows the weightings used in the calculation of the aggregate indices:
During 2018, the number of accesses to the INE website for consulting the tables containing information related to the Retail Trade Index statistical operation were:
AC1= 187,604 acceses
And the number of accesses for consulting metadata have been:
AC2= 1,348 acceses
In the case of the Retail Trade Index, an anonymous microdata file is not published, these are only provided to the Regional Statistics Institutes with which a collaboration agreement has been signed for the submission of said information, by virtue of the collaborative Framework Agreement between the INE and the Central Statistics Offices of the Autonomous Communities (OCECAS).
It is possible to request tailored information from INE's Information area. Limitations on confidentiality or accuracy are taken into account when processing such requests.
Documentation with the methodology can be consulted at:
The rate of completeness of metadata" is the ratio of the number of metadata elements provided to the total number of metadata elements applicable.
Metadata completeness rate, AC3=100% as all fields of the methodological file are filled in.
The statistics are processed in accordance with the EU guidelines, complying with the Code of Practice implemented by Eurostat.
Sections 10.6 to 17.2 of this document are the user-oriented quality report for this operation.
Furthermore, whenever Eurostat requests any kind of additional information they are supplied as soon as possible.
Quality assurance framework for the INE statistics is based on the ESSCoP, the European Statistics Code of Practice made by EUROSTAT. The ESSCoP is made up of 16 principles, gathered in three areas: Institutional Environment, Processes and Products. Each principle is associated with some indicators which make possible to measure it. In order to evaluate quality, EUROSTAT provides different tools: the indicators mentioned above, Self-assessment based on the DESAP model, peer review, user satisfaction surveys and other proceedings for evaluation.
The RTI production process is designed to guarantee the quality of the information used, as well as that of the results obtained from it. For this, different quality controls have been designed in all phases of the process, such as the detection of invalid values, estimation of non-response, supervision and inspection of fieldwork and other permanent controls on the flow of information.
This statistical operation is carried out in accordance with Council Regulation 1165/98 of 19 May 1998 on short-term statistics and follows EU guidelines, complying with the Code of Practice and taking the necessary measures to ensure that the published indicator meets all the necessary quality requirements.
The RTI is a high quality product. Its sample size offers information up to the Regional level within reasonable sampling errors, the harmonised methodology used allows comparisons to be offered at the international level with a high level of solvency and gives an extremely valuable measure on the retail sector, which no other statistical source provides with the same level of detail.
As far as the limitations of the survey are concerned, it is necessary to point out those inherent to statistical operations by sampling, such as non-response and the aforementioned sampling errors or coefficients of variation of the estimates. In both cases, they remain within a reasonable limit.
Regulation No 1158/2005 of the European Parliament and Council of 6 July 2005 (amending Regulation No 1165/98 of the European Union Council of 19 May 1998 concerning short-term statistics, although the latter is still in force) establishes that the quality of the variables should be checked periodically by comparing them with other statistical information. This comparison will be carried out by each of the Member States and by the Commission (Eurostat). Furthermore, internal consistency will be checked.
The quality criteria listed in Article 12(1) of Regulation (EC) No 223/2009 of the European Parliament and Council, namely relevance, accuracy, timeliness, punctuality, accessibility, clarity, comparability and coherence, shall also be met.
The RTI covers the statistical information needs of different user groups:
The survey also responds to the needs of foreign users, both institutional and private. The former include Eurostat, OECD, IMF, United Nations, etc.
Some specific examples of the uses to which our information is destined are:
- The European aggregate calculation of RTI in Eurostat.
- The use of RTI in Spanish National Accounts for the calculation of GDP.
- Companies use it to know the evolution of the sector and their relative position within it.
The INE has carried out general user satisfaction surveys in 2007, 2010, 2013 and 2016 and it plans to continue doing so every three years. The purpose of these surveys is to find out what users think about the quality of the information of the INE statistics and the extent to which their needs of information are covered. In addition, additional surveys are carried out in order to acknowledge better other fields such as dissemination of the information, quality of some publications...
On the INE website, in its section Methods and Projects / Quality and Code of Practice / INE quality management / User surveys are available surveys conducted to date.(Click next link)
Those responsible for the survey are in permanent contact with the main users, in order to meet any specific need for information.
Furthermore, in view of the possibility that there may be requests that have not been met due to their complexity, these are evaluated at each base change, as well as any suggestions made by the main users. Most of these demands have been satisfied.
The survey responds to all information requirements, both national and international.
The RTI is governed by Regulation No 1158/2005 of the European Parliament and Council of 6 July 2005 (amending Regulation 1165/98 of the European Union Council of 19 May 1998 on short-term statistics, although the latter is still in force). The requirements contained in each of them are satisfied by this indicator.
The "Rate of available statistics", R1, is the ratio of the amount of data produced in the statistical operation subject to a regulation, to the total amount of data required by that regulation.
Rate of available statistical results, R1=100%
The sample design attempts to minimise sampling errors and the different survey processes are aimed at eliminating or reducing, as far as possible, survey errors, both in the collection phase (monitoring the response rate and controlling the filtering of errors) and in subsequent imputing of non-response and calculation of aggregate indices, which allows a high degree of reliability of the statistical operation to be obtained.
The sampling error, A1, is the coefficient of variation (or standard relative error) of a parameter estimator.
In order to obtain a quality measurement of the indices, an approximate relative sampling error is calculated for the year-on-year variation rates for the turnover and total employed persons variables. The general expression of the estimated relative error, assuming negligible bias, is given by:
is the ratio estimator, given by the quotient of the total estimate of variable Y (turnover or total employed persons) in month m of year t, , and the estimate of Y obtained in the same month m of year t-1, .
The method used to calculate the variance estimator is Taylor linearisation.
The mean sampling error, A1, for 2018 (January - December) of the General Retail Trade Index is:
Non-sampling errors are checked throughout the entire statistical process. Specific information on the non-response rate is also available.
A continuous quantitative assessment is carried out of non-sampling errors and risks of bias associated with:
- Lack of coverage, excess coverage and duplicates in the framework.
- Collection tools.
- Total non-response, including causes of non-response and measures to reduce it.
- Partial non-response in the main statistical operation variables.
- Editing, coding and imputing of data.
- Model used in the estimation.
The results of this measurement are analysed in order to carry out relevant actions to reduce the different error types.
An objective measurement is carried out, based on a quantitative indicator (Non-response rate) that is calculated as the percentage assumed by the number of units without information or with unusable information with respect to the total of initially selected sampling units (i.e., discounting framework errors).
Although a reserve sample is available, the rate is only calculated based on collaboration or otherwise of the titular units of the sample, i. e. excluding the units that make up the reserve sample.
A2: Tasa de sobrecobertura
La tasa de sobrecobertura mide la proporción de unidades elegibles a través del marco que no pertenecen a al población objetivo (están fuera del ámbito).
Para el año 2018 la tas de sobrecobertura es A2 = 15.91%
A4: Tasa de no respuesta
La tasa media de no respuesta A4, se calculan con la Tasa de No Respuesta por unidad, que es la razón de unidades que no responden respecto al total de unidades de la muestra:
- Tasa Media de No Respuesta, A4 de la primera versión de los doce meses de 2018: A4 = 9,47%
- Tasa Media de No Respuesta, A4 de la versión definitiva de los meses comprendidos entre enero y octubre de 2018: A4 = 4,86%
A5 Tasa de no respuesta por ítem
Esta tasa se calcula como el cociente entre las unidades de la población objetivo encuestadas que no han respondido a la pregunta de cifra de negocios o empleo, sobre las unidades que han cumplimentado el cuestionario total o parcialmente.
La tasa media de no respuesta en el año 2018 es:
- Para cifra de negocios: A5 = 0.15%
- Para empleo: A5 = 0.01%
A7 Tasa de imputación
La tasa de imputación mide el porcentaje de valores asignados a las empresas en las que faltan estos datos, son inválidos o inconsistentes.
La tasa media de imputación para el año 2018 es:
- Para cifra de negocios: A7 = 7.02%
- Para empleo: A7 = 6.66%
The RTI is usually published in the last week of the month following the reference month, thus meeting the deadlines established by Eurostat.
The Opportunity, T1, is the number of days between the last day of the reference period of the data and the day of the first publication.
Mean of T1 Indicator. Year 2018: T1 = 30 days
The RTI is published according to the calendar established by the INE at the beginning of each year.
The survey is designed homogeneously throughout the geographical area, in terms of methodology, design, information collection process, and data filtering. This ensures the comparability of the results between the different Autonomous Communities and Cities.
On the other hand, in the European area, comparability between countries is ensured thanks to the Council Regulation 1165/98 of 19 May 1998 on short-term statistics, which makes the statistics uniform with the rest of the countries in the European Union that publish the same information.
The RTI was first published in 1995. The base changes carried out since then (base 2001, base 2005, base 2010 and base 2015) have modified some methodological aspects of the survey, which has meant gaps in the series that have been suitably treated to offer a linked series from the first year of publication of the survey to the present.
Comparability over time, CC2, is the number of comparable elements in a time series since its last gap.
The number of comparable data, CC2, as of December 2018 was:
When data is validated, information from other statistics is taken into account in order to guarantee consistency between them, such as
- INE's Annual Services Survey (ASS).
- INE's Labour Force Survey (LFS).
- INE' s Services Sector Activity Indicators (SSAI).
- Household Food Consumption (HFC) of the Ministry of Agriculture and Fisheries, Food and the Environment.
The consistency between variables is compared at all stages of the statistical process.
Furthermore, the results are checked to ensure that they are consistent with the economic reality of the market and, in particular, with its development in the retail trade sector.
The questionnaire has been designed to include all the necessary information and at the same time to be as simple as possible in order to facilitate its completion by the companies making up the sample. This will help reduce, as far as possible, excessive use of certain respondent units arising from the completion of multiple surveys in the same year.
The response burden endured by the respondent units is continuously monitored, measured as the average time to complete the questionnaire.
To prevent respondent fatigue and sample ageing, allowing the selection of newly created companies, an annual rotation of 20-25% is carried out, i. e. between 20% and 25% of sample units are replaced by new companies.
The budgetary appropriation necessary to finance this statistic, provided for in the 2019 Annual Programme, is 2,191.12 thousand euros.
The INE of Spain has a policy which regulates the basic aspects of statistical data revision, seeking to ensure process transparency and product quality. This policy is laid out in the document approved by the INE board of directors on 13 March of 2015, which is available on the INE website, in the section "Methods and projects/Quality and Code of Practice/INE’s Quality management/INE’s Revision policy" (link).
This general policy sets the criteria that the different type of revisions should follow: routine revision- it is the case of statistics whose production process includes regular revisions-; more extensive revision- when methodological or basic reference source changes take place-; and exceptional revision- for instance, when an error appears in a published statistic-.
At each publication date, the data from the last three months of the historical series are provisional, with those from the preceding months being definitive.
However, the series adjusted for calendar effects and seasonal and calendar effects are reviewed from 2000 in their entirety every month.
The first time the index of a given month is published, it is deemed provisional, and does not become definitive until three months later. In this time frame, this index is revised by applying any changes in the data that may have occurred since the first publication date, for example due to the substitution of the estimated data (by management issues) with the real data filled in by the company.
For data revision purposes, the indicator A6 is used, Mean absolute revision, where revision is defined as the difference between the last estimator and the preliminary one. Two estimators are calculated, over a series of reference periods: MAR (Mean Absolute Revision) and RMAR (Relative Mean Absolute Revision with respect to the last estimator).
The MAR, RMAR values of the General Retail Trade Index collected from information for the last 36 months (the last month being considered as December 2018) are:
A6_MAR turnover at constant prices (December 2018): A6_MAR=0.08
A6_RMAR turnover at constant prices (December 2018): A6_RMAR=3.12%
A6_MR turnover at constant prices (December 2018): A6_MR=0.02
A6_MAR calendar adjusted turnover at constant prices (December 2018): A6_MAR=0.09
A6_RMAR calendar adjusted turnover at constant prices (December 2018): A6_RMAR=2.86%
A6_MR calendar adjusted turnover at constant prices (December 2018): A6_MR=0.02
A6_MAR Employment (December 2018): A6_MAR=0.06
A6_RMAR Employment (December 2018): A6_RMAR=4.11%
A6_MR Employment (December 2018): A6_MR=0.01
To obtain survey data, questionnaires are used for companies whose main activity is registered in Division 47 of Section G of the National Classification of Economic Activities (CNAE-2009).
The sample size has been calculated in order to give indicators of the representative turnover and employment variables. The sample consists of approximately 12,000 respondent units.
Annual rotations of approximately 20-25% of the sample are carried out, where all new companies from the comprehensive strata are incorporated, where those that have been low, unreachable, erroneously included, merged are eliminated and where small companies for which the completion of the questionnaires requires a great effort are replaced.
.The annual rotation process consists of making a new selection of sample units that represents commercial distribution in Spain and a re-weighting of these units so that the sample data is elevated to the population data.
The rotation of the sample for the index calculation takes place in January of each year, at which time the elevation factors and weightings are also updated.
Data collection is carried out monthly.
The respondent companies send their data via the Internet; by ordinary mail (questionnaires completed on paper); or by fax or telephone.
In the questionnaire completion phase, a system is used to detect possible inconsistencies whereby the respondent is warned and is required to confirm or modify the information provided.
When this information is recorded by INE's own staff, there is a recording control system that warns of possible inconsistencies that may arise. In this process, for any data where there are discrepancies, the staff responsible for collecting the information will contact the respondent to confirm or modify the information provided.
Once all the information has been compiled, a coverage control is carried out in order to guarantee the accuracy of the recorded data, as well as to detect duplicates and coverage errors, non-response, etc.
Prior to starting the process for obtaining the indices that will be published regarding this statistical operation, the lack of response is imputed with a figure corresponding to those units that have not provided the information corresponding to the reference month.
Subsequently, with the information received at the INE collection centres, the data filtering process begins. The process consists of two phases: in the first, a microdata filtering system is carried out, while the second consists of an analysis for different levels of aggregation, or macro-filtering.
The macro-filtering consists of checking the validity of data from each company, taking into account the historical series provided by the company itself. The analysis establishes month-on-month and year-on-year comparisons, so that the data is validated without re-contact when its evolution is repeated periodically.
The second part of the filtering is performed once the data is processed together. A centralised or macro-filtering process is carried out with the aggregated information. Unlike micro-filtering, in this phase the weightings of each activity form part of the calculation, taking into account the variation rates for different levels of aggregation, etc. In cases where information is available from other sources, a consistency process is carried out between these and the results of the survey.
As a result of the change of base to the new base of 2015, the weightings have been updated, when the samples for the years 2015, 2016 and 2017 have been updated, the weightings have been recalculated with data from December (t-1) according to these new updated samples, and both the elementary indices and the aggregate indices of these three years have been recalculated.
Calendar effect: the impact produced in the time series of a variable, due to the different structure that the months present in the different years (in both length and composition), if the remaining factors influencing said variable remain constant. The RTI indicator is very sensitive to the calendar effect, since one more holiday in a particular month means a proportional decrease in the sales for that month.
To correct this problem and obtain a series that compares production over time in a homogeneous way, the series is adjusted to take into account the following calendar factors: working days, Easter and leap year.
Once the calendar effects are removed, seasonal effects are removed accordingly. Seasonal fluctuations are movements that occur with similar intensity in each month and are expected to continue.
The seasonally adjusted series, i.e., corrected for seasonal and calendar effects, provides an estimate of what is "new" in a series (changes in the trend, cycle, and irregular component).
The method is based on regression models with intervention variables of the three mentioned effects.
The adjusted series is published together with the non-corrected series for the sake of a complete analysis of the situationmetadatosicm_base2015_en.xls