Post on 18-Jan-2018
description
United Nations Statistics Division
Dissemination of IIP data
Dissemination of IIP data Dissemination principles Recommended content for publication
of IIP Data revisions International reporting
Dissemination principles General dissemination principles
apply to releases of IIP data Statistical confidentiality Equality of access Objectivity
Also described in IRIS 2008
Statistical confidentiality Included in the United Nations
Fundamental Principles of Official Statistics “Individual data collected by statistical
agencies for statistical compilation, whether or not they refer to natural or legal persons, are to be strictly confidential and used exclusively for statistical purposes.” (UNSC 1994)
Statistical confidentiality Sometimes it is possible to deduce
information about an individual economic unit from aggregated information especially when the contribution of one (or two) economic unit dominates the total
Statistical confidentiality Accepted rules in business statistics:
A tabulation cell should comprise at least three units
For cells with largest numbers, the three units with the largest values should together not dominate, i.e. not account for more than 70 per cent of the cell value
Statistical confidentiality Practices to protect disclosure of
confidential data Aggregation Suppression
May require secondary suppression Other:
Controlled rounding Perturbation
Equality of access Statistics compiled by national
statistical offices are collective goods no users are privileged and everyone can
take note of statistical data under equal terms
no new data are supplied to anyone before these are officially released internal release e.g. to key ministries may
be necessary
Equality of access In most cases press release is the first
publication Serves dual purpose:
making the data officially public signal to the data users that additional
data on the subject can be obtained from the NSO
Advance release calendar should be publicly available
Timeliness Timeliness of release of annual and
infra-annual statistics varies Trade-off between timeliness and
quality (accuracy)
Recommendation for IIP: Monthly IIP released within 45 days after
the end of the reference month
Objectivity Released data should not be
accompanied by judgments or recommendations The independent and objective position
of the national statistical office does not permit subjective interpretations
Note: Methodological explanations and advice should be included
Recommended content for publication Both ‘original’ data series and seasonally adjusted
data series should be published A reference period needs to be determined and
convention is that this period is set to a value of one hundred (100.0)
Present changes between month-to-month and change over same month of previous year
Present the main contributors to change, i.e. those product groups or industries that are primarily responsible for the monthly movement in the IIP
Additional content for publication Descriptions of the methodologies used in the
compilation of the index (choice of index formula, strategy for constructing the index series etc.)
Definitions of the underlying economic concepts the indices intend to measure
Weighting system used, weight update practices and frequency of weight update
Treatment of changes in the composition of commodities in the market as well as changes in quality
Limitations in the use or application of the indices
Data revisions Data revisions are inherent in the way
estimates are compiled and released, reflected in Preliminary data
Based mainly on trends and statistical techniques
Provisional data Based on limited amount of data
Final data Based on comprehensive data
Data revisions Result of the trade-off between need
for timeliness of published data and their reliability, accuracy and comprehensiveness
May be perceived as negative Unavoidable for data like IIP, where
high frequency and fast availability is important
Data revisions Reasons:
“normal” statistical procedures (new information available, change in data source, change in base year)
Correction of errors
Also: changes in underlying definitions, classifications etc. Less frequent, not strictly “revisions”, but
often occurring at the same time Explanation should be provided to avoid
misinterpretation of reasons for data changes
Data revisions Revision policy for normal statistical
data revisions should be developed, featuring: predetermine schedule reasonable stability from year to year openness advance notice of reasons and effects easy access of users to sufficiently long
time series of revised data as well as adequate documentation of revisions
International reporting IIP has been collected by UNSD since
the 1950’s Current level of collection
Monthly data at section level (1-digit) of ISIC
Quarterly data at division level (2-digit) of ISIC
Change from ISIC Rev.3 to ISIC Rev.4