Italian National Institute of Statistics (ISTAT)

Integrated Labour, Education and Training Division

Social Statistics and Population Census Directorate

Statistics Production Department

Integrated Labour, Education and Training Division

Social Statistics and Population Census Directorate

Statistics Production Department

Istat - Italian National Institute of Statistics

Via Cesare Balbo, 16 - 00184 Rome - Italy

Indices of volumes of hours worked (STSCONS).

NACE Rev. 2

Activities covered: Nace Rev.2 section F.

Size classes covered: Enterprises with at least one employee.

Hours worked are defined in coherence with Commission Regulation (EC)  No 1503/2006.

Reporting and observation unit is enterprise

All enterprises with at least one employee which were active in the reference quarter in the STSCONS domain. In 2018, these enterprises were, on average over the four quarters, 158 thousand.

The whole country. Activities outside the geographical coverage are not included in the data.

The series start in 2000Q1.

2015

Indices with base year 2015=100.

2018. Quarterly data

Legal basis: all the three sources involved in the production of the indicators are among the compulsory surveys identified every year by a Decree of the President of the Republic (DPR).

Obligation on units to provide data: the two business surveys are mandatory. Penalties are foreseen for non-response of units with at least 250 employees. Furthermore, firms have an obligation to fill monthly the social contributions forms on which the administrative source is based.

Planned changes in legal basis, obligation to respond and frame used: none.

None. (Data are not sent to OECD, UN, etc. or used in reports sent to these institutions). 

Links to legal acts on the production of official statistics and confidentiality can be found on the Sistan - National Statistical System website.

Divisions ‘05’, ‘06’, ’07, ‘09’, ‘12’ are considered confidential because each of them includes very few firms. Therefore the series on these divisions are not transmitted to Eurostat. However, the data of firms belonging to these divisions are used to calculate higher level aggregates. 

At the moment the data are not released.

At the moment the data are not released.

Data are transmitted to Eurostat via Edamis.

Quarterly

At the moment the data are not released.

Not available

At the moment the data are not released.

At the moment the data are not released.

Data are trasmitted to Eurostat quarterly, within the 90 day regulation deadline, in SDMX format.

As indicated below in 18.1, volumes of hours worked in 1+ employee firms are calculated multiplying per capita hours worked on 10+ employee firms by the number of jobs in 1+ employee firms.

There are no methodological publication describing the method employed to calculate volumes of hours worked in 1+ employee firms.

However, the methods used to calculate per capita hours worked on 10+ employee firms are described in the following publication: Bellisai, D., Binci, S., Gigante, S., Libratore, A., Serbassi, L. and M. Sorrentino, “Verso la modernizzazione della produzione delle statistiche congiunturali sull’input di lavoro: il processo di integrazione tra la rilevazione mensile sulle grandi imprese e quella trimestrale sui posti vacanti e le ore lavorate”, Istat, Working Papers, n.11, 2013 (http://www.istat.it/it/files/2013/12/IWP_11_2013.pdf). 

Not available.

The data are produced in accordance with the European Statistics Code of Practice and with its Italian version (http://www.istat.it/it/files/2011/11/codice_statistica.pdf).

A specific assessment of data quality for each quality dimension is presented below. There are no relevant quality problems that deserve mentioning here.  

Eurostat is the main user. The produced data are coherent with the requests of the STS regulation. 

The data are considered satisfying the STS regulation requests by Eurostat. 

All STS requirements are fulfilled. 

As the HOWK series are calculated by multiplying survey based per capita hours worked by administrative based jobs, sampling errors affect HOWK via per capita hours worked. For an assessment of sampling errors on the numerator and dominator of per capita hours worked, see 13.2.

For what concerns non sampling errors, response rates can be found below in 13.3, while descriptions of models used in editing and imputation and grossing up can be found below in 18.5. Coverage errors affect the per capita hours worked because the surveys’ reference and study population differ due to the enterprises beginning and finishing their activities in the reference population between the SBR reference year and the surveys reference quarter. If these errors are measured via the difference between the SBR used by the survey and the OROS (see below, type of source, 18.5) list for the reference quarters, for 2017:

- 18.7% of firms in the surveys’ reference SBR were not in the OROS list for the reference quarters, and

- 29.3% of firms in the OROS list for the reference quarters were not in the surveys’ reference SBR.

As the HOWK series are calculated by multiplying survey based per capita hours worked by administrative based jobs, sampling errors affect HOWK via per capita hours worked.

For what concerns survey based per capita hours worked, in the four quarters of 2018 the coefficient of variation is estimated as 0.8% at the section level in the Nace sections covered by the STSCONS domain.  

Response rate for VELA and LES, considered together, for section F: 

Actions to speed up or increase the rate of response:

-          monthly reminders (by e-mail) and intensive follow-ups by phone are addressed to non responding LES units. Once a year a warning with penalty (registered letter with return receipt) is sent to firms that have not answered to LES during the previous year.

-          quarterly reminders by phone and e-mail are made to the VELA survey non responding units. Once a year a warning with penalty (registered letter with return receipt) is sent to firms that have not answered to VELA during the previous year and have at least 250 employees.

On coverage errors, see 13.1.

Timeliness: data for 2018 have been transmitted to Eurostat on average 73 days after the end of the reference quarter, that is 17 days before the regulation deadline.

Timetable of data collection: for LES, the deadline for responses is fixed at 18 days after the reference month; for VELA the deadline for responses is fixed at 37 days after the reference quarter.

Data for 2018 have been transmitted to Eurostat on average 73 days after the end of the reference quarter, that is 17 days before the regulation deadline.

Hours worked by self-employed are not included. On all other aspects, hours worked are defined in coherence with Commission Regulation (EC)  No 1503/2006. The data cover the entire national territory.

The 2000-2003 data are based on data collected through LES only (and not also through VELA) and appropriate estimation procedures that make the per capita hours worked that are multiplied by the OROS jobs (see below, type of source, 18.5) representative of all enterprises with 10+ employees.

HOWK data are coherent with those transmitted to Eurostat for STS variables EMPL and WAGE, because the series for all three variables are based on the OROS population of firms active in the reference quarter and measure of jobs.

Data at different aggregation levels are coherent.

For 2018:

-  for VELA, the cost of the phone reminders was equal to €165,500.

Furthermore, it can be estimated that around 5 person hours are needed each quarter to produce the HOWK indicators (from data collection, to data processing and validation).

Under the assumption that around one hour of work is needed to fill in the relevant parts of administrative forms and survey questionnaires by each responding firms each quarter, the burden on respondents can be estimated at around 1.173 million of hours of work per quarter for all STS domains. 

The data are disseminated provisionally at the time of the first transmission to Eurostat. A revision of the data will be carried out yearly with the dissemination of the indices of the first quarter for the period covering the previous 8 quarters.

If errors are identified during the year:

-          if they are considered relevant, the series are re-calculated and resent as soon as possible;

-          if they are considered less relevant, the series are re-calculated and resent together with the following regular revision, that is with the data for the first quarter of the year.

So far, comparisons with other sources has not led to revisions.

The revisions carried out once a year are due to late respondents in the survey and administrative sources.

A vintage database of all releases is available.

Istat's principles, rules and classifications of short-term statistics revisions can be found here: https://www4.istat.it/en/economic-trends/revisions

In June 2018 revisions for 2016 and 2017 have been transmitted together with the data for 2018Q1.

Type of source:  volumes of hours worked in 1+ employee firms are calculated multiplying per capita hours worked on 10+ employee firms by the number of jobs in 1+ employee firms. Per capita hours worked are based on two business surveys: VELA, a quarterly survey on 10-499 employee firms; and LES, a monthly survey on 500+ employee firms. The number of jobs in 1+ employee firms is derived from an administrative source (OROS) and is the same transmitted for the STS EMPL variable (detailed information on this source can be found in the metadata on the EMPL variable).

Frame on which the source is based: for both VELA and LES, the latest available release of the SBR.

Sample or census: VELA is sample based; LES is a census.

Criteria for stratification: 

for LES: none.

Economic activity, size and geographical area are used as stratification variables for the VELA sample. The considered stratification classes are the following:

- for economic activity: divisions from 05 to 36 and from 45 to 47; the aggregate of divisions 37, 38 and 39; sections F, and from H to N;

- for size (in terms of employees): 10-19, 20-99 and 100-499;

- for geographical area: North, Centre and South (this last class includes Sardinia and Sicily).

Threshold values and percentages: in 2018, the sample sizes for the STSCONS domain were the following:

- for LES, around 27;

- for VELA, around 1,000, representing 6.5% of the population of 10-499 employee firms in STSCONS domain in the latest release of the statistical business register

- for OROS, the population size is around 158 thousand.

Frequency of updating the sample: once a year (for the wave referring to the first quarter):

- the list of all the population firms with at least 500 employees is updated on the basis of the SBR most recently available version, and

- around one third of the 10-499 employee sample firms is rotated.

Frequency for per capita hours worked (business surveys): for 10-499 employee firms, quarterly.

Frequency for 500+ employee firms, monthly.

Frequency for jobs (administrative source): monthly.

Questionnaires used in the survey: See attached files.

Data collection media: Web, administrative data.

It is carried out through interactive assessment of influential observations, which are identified via both automated procedures and experts’ analyses on aggregate data. 

The files that are sent to Eurostat are produced from data stored in an Oracle database via a generalised Istat software. After their production, they are not checked with any further software or specialized tool. 

Estimates for non-response:

for what concerns LES, the methodology of estimating missing data (unit and item non responses) works on a deterministic basis. It uses both information of clusters defined in terms of economic activity and time series data of the enterprise itself.

For what concerns the VELA survey, only item non responses are imputed. For what concerns jobs (at the denominator of per capita hours worked), editing and imputation is based on an auxiliary source, the OROS survey (based on the social security database of the forms that all enterprises with at least one hour of remunerated work have to fill each month to pay the compulsory social contributions). For what concerns hours worked, editing and imputation of item non responses includes: identification of outliers and selective treatment of the most relevant ones; deterministic editing procedures; donor imputation procedures based wherever suitable on the same quarter of the previous year data for the same firm on which they are applied; identification and analysis of the records most influential for the aggregate data.

Estimates for grossing-up to population levels:

all responding firms and the unit non responses imputed records are assigned a weight via calibration. The only exception is a small list of extremely large firms which are assigned a unit weight.

The theoretical reference population for the calibration is made by all the enterprises with at least 10 employees in Nace Rev. 2 sections B to N which were active in the reference quarter. This population is well represented by the set of microdata of the OROS survey for the reference quarter. It is this set of microdata that is therefore used to derive the calibration constraints. The quarterly average of monthly data on jobs, as measured by OROS, is used as the auxiliary variable in the calibration.

The calibration classes are based on economic activity and enterprise size. The economic activity classes for calibration are based on Nace Rev. 2 divisions for sections from C to E, G and N, while they are based on sections for B, F, and from H to M. The considered size classes are based on employees and are thus defined: 10-19, 20-99, 100-499 and at least 500.

The initial calibration weights are based on both inclusion probabilities and response rates.

The calibration is carried out using a generalized software purposely built by Istat, ReGenesees, and within it, a truncated logarithmic distance function.

Type of index: value indices with a fixed base.

Method of weighting and chaining: none.

Planned changes in production methods: none.

See attached file.

Not available