EUROPEAN COMMISSION EUROSTAT Directorate C: National Accounts, Prices and Key Indicators Unit C.3: Statistics for administrative purposes Luxembourg, 17 March 2014 Doc. A6465/14/15 Meeting of the Working Group on Article 64 and Article 65 of the Staff Regulations to be held in Luxembourg on Monday 24 th and Tuesday 25 th March 2014, in the Bech Building Room Ampère, starting at 9:30 a.m. Reviewing the approach to establish Intra-EU PPP for Rent in large duty station cities Test calculation to show impact of integrating detached houses when calculating PPP for Rent in large agglomerations (Madrid, Rome, London). Paper relating to Agenda item 13.4 This document can be found on the dedicated CIRCABC site for meeting delegates: https:/circabc.europa.eu/
Contents 1. Introduction... 2 1. Current Rent parities and A64 approach to PPP calculation... 2 2. Test calculations containing all dwelling types... 3 3. Conclusion... 4-1 -
1. Introduction The goal of the Article 64 exercise is to establish salary correction coefficients which maintain equivalence of purchasing power relative to Brussels. The Article 64 exercise is designed to generate robust results at aggregate level, but to the extent that problematic issues at sub-aggregate level exist and do not cancel each other out, the quality of the overall result is to some degree inevitably dependent on its component parts. In recognition of the uniqueness of accommodation costs and their importance within the consumption basket (typically 20%-25%) - and the consequent potential impact on the overall Correction Coefficient - since 1990 a specific methodology for calculation of rent parities has been endorsed by the Article 64 Working Group. This methodology involves a six-year moving average model to reflect typical occupancy periods, with a set of defined dwelling types for which price data is collected through annual estate agency rents surveys and updated using relevant price indices, and for which preferences are identified by periodic staff housing surveys (an average structure is applied for duty stations with low staff population response rates). For each duty station the dwelling type average rent prices are compared with the corresponding situation in Brussels. These ratios are then weighted using the preferences for Brussels (ie. Laspeyres-type) and the duty station (ie. Paasche-type) and geometric average gives the overall rent parity (Fisher-type). 2. Dwelling types for current and future rent parities Housing cost is covered by two basic headings: 20 (tenants rents) and 21 (imputed rents of owner-occupiers). The weights for heading 20 are obtained directly from the Survey of Household Expenditure (SHE), while those for heading 21 are obtained by imputing average rents by housing type, also from the SHE. The parities for heading 20 are calculated from ratios of rents as reported from the Estate Agency Rent Surveys. These parities are then imputed to heading 21. For the annual adjustment in June 2013, for the rent parity calculation, new dwelling types were introduced in 3 duty stations (Athens, Paris, The Hague). The main reason for this is the alignment with the actual dwelling types and locations people are living in. Table 1. Dwelling types introduced in June 2013 Place Non Detached House Detached House EL Athens 140-160 m 2 190-220 m 2 FR Paris 110-130 m 2 150-180 m 2 NL The Hague - 150-180 m 2-2 -
In order to smooth the impact, the introduction of these two dwelling types was phased-in for three other duty stations, with prices for non-detached houses integrated for the June 2013 CCs and prices for detached houses scheduled for introduction in the June 2014 exercise. Table 2. Dwelling types introduced in June 2013 and June 2014 Place Non Detached House (in 2013) Detached House (in 2014) ES Madrid 140-160 m 2 190-220 m 2 IT Rome 110-130 m 2 190-220 m 2 UK London 80-100 m 2 110-140 m 2 Regarding the actual data included, in the case of Rome, rent surveyors have reported through the years that detached houses for the rental market surveyed are not typically found within the city limits but in other residential districts (suburbs); and, furthermore the current price per square meter for the previously included typology (flats in central location) is significantly different from the price per square meter for the other type of accommodation (houses and flats in other locations). This conclusion can confidently be expected to also apply for other large conurbations. 3. Test calculations containing all dwelling types The following table shows the definitive values established and used for 2013 (including nondetached houses only). Table 3. PPP and CC for rents June 2013 Place PPP rent 2013 Actual CC rent 2013 Actual ES Madrid 1.176 117.58 IT Rome 1.334 133.42 UK London 2.658 311.58 In order to quantify the likely impact for 2014, test calculations were run for 2013 including also prices for the additional dwelling type (detached houses). These are shown in the following table. Table 4. Test results for rent PPP and CC as of June 2013 Difference (%) Place PPP rent 2013 Test CC rent 2013 Test PPP rent CC rent ES Madrid 1.167 116.66-0.78% -0.78% IT Rome 1.241 124.14-6.95% -6.95% UK London 2.658 311.52-0.02% -0.02% - 3 -
It can be seen that the rent parities in Rome would have been around 7% less, if all dwelling types were included as of June 2013. 4. Implication for global parities and CCs On the basis of the above test calculations, the global PPPs and CCs would vary noticeably only in IT-Rome, with -1.5% for staff and -0.9% for pensioners. The other two duty stations are mostly not affected by the new dwelling types being inserted. Table 5. Impact on PPPs and CCs for staff as of June 2013 Annual adjustment 2013 (12 month period : June 2012-2013 ) Purchasing power parities and correction coefficients relative to Brussels for staff June 2013 Mar 2014 01-Jul-12 Final 01-Jul-13 Calculated 01-Jul-13 Simulated Difference 2013 (%) PPP CC PPP CC PPP CC PPP CC ES Madrid 0.971 97.1 0.963 96.3 0.961 96.1-0.17% -0.21% IT Rome 1.042 104.2 1.044 104.4 1.028 102.8-1.53% -1.53% UK London 1.181 147.8 1.188 139.2 1.187 139.2-0.01% 0.00% Table 6. Impact on PPPs and CCs for pensioners as of June 2013 Annual adjustment 2013 (12 month period : June 2012-2013 ) Purchasing power parities and correction coefficients relative to Brussels for pensioners June 2013 Mar 2014 01-Jul-12 01-Jul-13 01-Jul-13 Difference 2013 Final Calculated Simulated (%) PPP CC PPP CC PPP CC PPP CC ES Madrid 0.909 90.9 0.913 91.3 0.912 91.2-0.10% -0.11% IT Rome 0.974 97.4 0.979 97.9 0.970 97.0-0.98% -0.92% UK London 0.950 119.0 0.968 113.5 0.968 113.5 0.00% 0.00% 5. Conclusion Delegates are invited to take note of/to comment on the information presented in this document. - 4 -