Post on 30-Dec-2015
description
transcript
Better Information for Regional Government
Marie Cruddas, Minda Phillips & Pete Brodie, ONS.
Presented by Martin Brand, ONS Methodology Directorate
Outline of Presentation
• The Neighbourhood Statistics Programme– Background– Topics and data sources– Estimating for small areas
• Allsopp Review – Background– Recommendations– Developments
Neighbourhood Statistics - The Need
“…anyone can wander through some of these
[deprived] areas and know that something is very
badly wrong – but the government has never set
out to record or analyse the issues in a
comprehensive way.…”
PAT 18 Report on Better Information
What is Neighbourhood Statistics?
A service designed to meet the needs of the National Strategy for Neighbourhood Renewal.
“The absence of information about neighbourhoods has produced a series of failings at national, local and community level…policies can easily be misdesigned or mistargeted….and important trends have been missed”.
What does Neighbourhood Statistics offer?
Publicly available internet access to -
• Neighbourhood Profiles;• Thematically map any data;• Point location of services;• Library of datasets to view or download;• Pick and mix variables from different datasets;• Time series analysis; and• Create your own area.
http://neighbourhood.statistics.gov.uk
Topics and data sources
13 main topic areas relating to deprivation. Two examples are;
1. Work Deprivation: • Business and economic activity data, work-related benefits
claimants, and participation on government training programmes.
• Data sets include;– Benefit claimants (administative source)– Occupational group (Census)– Counts of enterprises by industry group (Business Register)– Employment rate (Survey)
Topics and data sources
2. Economic Deprivation: • Data relating to economic activity, poverty and the
provision of selected welfare benefits.
• Data sets include;– Income support claimants (administrative source)– Child benefit claimants (administrative source)– Household income (model based estimates)
Model-Based Estimation for Small Areas
• Census, survey and administrative sources do not cover all requirements
• Statistical techniques used to produce estimates for small areas when "standard" survey estimates for these areas are unreliable or cannot be calculated.
- Ghosh and Rao (1994) and Rao (1999).
• Use models to "borrow strength" over space, over time or from correlation with auxiliary information provided by administrative or Census sources
Small Area Estimates of Income
• Survey Data– Family Resource Survey (FRS)– Household total and net income – Sample size 21,000 households in 3,375 wards
• Covariate Data– 2001 Census– Department for Work and Pensions benefit claimant count
data– HM Land Registry dwelling price data– Council tax data– Regional indicators
• Estimation – Multi-level model– Model-based estimates for 9,275 wards– Published as experimental statistics
NeSS - Future Direction and Developments
• ‘First Stop Shop’;
• Increased use of administrative data;
• Improved analytical capacity;
• Ongoing improvements to usability; and
• Use of new technologies.
Allsopp Review – 2004
• Review of Statistics for Economic Policymaking
– To assess the demand for and provision of regional information, and
– examine whether official economic statistics adequately reflect changing UK economic structure
Allsopp Review – 2004
Recommendations relating to regional statistics
• Improvements to regional data– Good quality and timely estimates of annual Gross
Value Added for regions
• Make more use of administrative data
• Expand micro-economic and sub-regional data through the infrastructure used by NeSS
• Give greater access to the ONS business related administrative data
Developments
• Development of new Business Register Employment Survey (BRES)
– to inform the register– to provide the basis for the annual employment
estimates
• Access to Administrative Data for small businesses, eg
– Corporation Tax– VAT
Developments - BRES
Improved data– new questionnaire to improve data quality– employment size and turnover measure at LU level– Improved validation and imputation
• Improved stratification– complexity of business– FTE instead of headcount– “unusual” businesses defined using admin data
• Improved regional estimation– marginal level calibration– winsorisation
Developments - Regional Statisticians
ONS has established a presence in the 9 English regions. • work with regional partners on collaborative projects and provide advice
of the use of official statistics;
• gather information to improve the quality of ONS data and processes;
• quality assure final estimates of regional Gross Value Added (GVA);
• help improve the quality of the Inter-Departmental Business Register (IDBR) and business survey data;
• provide access to IDBR data; and
• provide a first point of ONS contact for key regional bodies.
Thank you