Experimental statistics

Methodology for experimental statistics on early estimates of major indicators of gross and net social protection benefits

The European system of integrated social protection statistics (ESSPROS) was elaborated by Eurostat, and Commission regulations (Legislation - Eurostat (europa.eu)) include the bases and framework of and rules for data collection. By month t+21, Eurostat publishes the detailed database compiled as a result of the compulsory annual data supply by member countries.

In the Hungarian Central Statistical Office (HCSO), we estimate and publish major aggregate indicators of social protection benefits as experimental statistics, according to a methodology partly differing from Hungary’s official ESSPROS compilation, by month t+10 in order to have as soon as possible a picture of the development of expenditure. The methodological preparation was supported by the EU.

Early estimates are made of major groups of existence risks and livelihood needs, at the level of ‘social protection functions’: 1 sickness and health care; 2 disability; 3 old age; 4 survivors; 5 family and children; 6 unemployment; 7 housing; 8 other social exclusion, not elsewhere classified.

Concepts, definitions, coverage and principles of accounting meet the requirements of official ESSPROS statistics: European system of integrated social protection statistics ESSPROS — Manual and user guidelines — 2022 edition - Products Manuals and Guidelines - Eurostat (europa.eu).

The difference mostly affects the group of available data sources, and so the applied methodology. It was a methodological aim to ensure that early estimates of social protection expenditure should approximate as much as possible the preliminary estimates of relevant ESA2010 indicators in national accounts (social transfers produced and financed by general government and non-profit sectors), while observing requirements in the ESSPROS Regulation.

Early estimates are a new statistical data collection, the compilation of which can be broken down into two parts according to the available data sources:

1) Itemised compilation and grouping of benefits in cash according to the 8 functions of existence risks, using annual budget reports, detailed tables and textual explanations in chapters of the draft of the Act on Closing Accounts, and other specific adoptions of administrative data. The data sources are available from months t+7 and 8.

2) Estimation and breakdown of benefits in kind according to the 8 functions of social protection. National accounts (ESA2010 D.631 and D.632) estimates of social benefits in kind are adopted by month t+9, from estimates of actual final consumption of general government sector, starting with COFOG 07 and COFOG 10, and from estimates on the sector of non-profit organisations serving households, starting with TEÁOR (NACE) 86–87–88. For the breakdown by social protection category of existence, welfare statistical data collections, experts’ special estimates on housing statistics, and other adoptions of administrative data are used.

Estimates of net social benefits in cash are made from 2008 in the frame of the EU ESSPROS data supply by month t+24. The estimation of the net module of social protection benefits was based on the micro-simulation of income positions on SILC (statistics on income and living conditions) data. In compliance with requirements of an EU Regulation, it is the average rates for social security contribution and income tax payments on benefits in cash that are to be estimated at the elementary level of the basic module (at national level, there is no territorial breakdown), for all items subject to tax and/or contribution, taking into account tax and contribution allowances.

New methodological foundations are being elaborated, on an experimental basis, and it is a strategic goal of both the Office and Eurostat to make statistical use of administrative data as widely as possible in order to improve the accuracy and timeliness of data. With the gradual introduction of new administrative data sources, net data can also be published one year in advance, simultaneously with early estimates, for the basic functions of social protection:

1) average social security contribution rates in a reference year for the group of social benefits can be calculated from data in contribution return 08, taking into account contribution allowances;

2) data on payments of social security contribution and personal income tax, reduced with allowances, on jobseekers’ benefits in a reference year are available from direct data adoptions (employment offices’ expenditure on unemployment benefits).

To estimate personal income tax not available from administrative sources, average tax rates on the social benefits in cash concerned continue to be constructed using micro-simulation, based on SILC data for year t–2.