NATIONAL STATISTICS BOARD

Venue: Room 308, Government Buildings, D02 R583

Date: Wednesday 11 May, 2023 

Time: 10.00am to 15.00pm

Agenda

National Statistics Board

Minutes

The meeting took place in the Government Buildings and was a hybrid meeting.

Date:     Thursday 11^th May 2023

Members present: Ms. Anne Vaughan (Chairperson), Ms. Eithne Fitzgerald, Dr. Eimear Cotter, Mr. Gerard Brady, Mr. John Martin, Mr. John Shaw, Mr. John McCarthy and Mr. Pádraig Dalton*

Non-members in attendance from CSO:  Brian Cahill (Secretary), Richard McMahon and Tara Davis, Paul Morrin (attended for items 5, 6 & 7), Jennifer Banim* (attended for items 2, 3 & 4).

In attendance for Agenda Item 1: Labour Market Statistics: Current and Future Data Needs

SOLAS

Joan McNaboe, Nora Condon

Economic and Social Research Institute

Dr. Seamus McGuinness

Department of Children, Equality, Disability, Integration and Youth

Nicola Tickner

Department of Enterprise Trade and Employment

Marcus Breathnach, Dermot Coates

Department of Social Protection

Dermot Corcoran, Hugh Cronin

*attended virtually

Item 1: Labour Market Statistics, current and future data needs

The Board received the following presentations:

• Tara Davis and Richard McMahon of the Central Statistics Office outlined the breadth of current labour market outputs, new outputs in development, and some challenges faced and data gaps that exist.
• Joan McNaboe of SOLAS illustrated their main publications and highlighted the challenges faced in capturing and analysing data on labour market skills and identify emerging skills needs.
• Dr Seamus McGuinness of ESRI highlighted the main sources of data their Labour Market Research Unit use in their work, what analysis they undertake and outlined what they see as the key areas for data development in the future to ensure policy makers have the requisite evidence to react to an ever-changing economy and labour market.

The discussion between the Board, presenters, and policy colleagues from DETE, DSP and DCEDIY highlighted the strides that have been made in labour market statistics in the last number of years and illustrated the need for further developments in particular areas.

It was noted in the discussion that:

• The Programme for the International Assessment of Adult Competencies (PIAAC) is a hugely valuable survey and source of insight. Plans are in place in CSO, SOLAS and ESRI to utilise this data source once the next round of data becomes available.
• There is an opportunity for CSO and Revenue to collaborate on the provision of greater insight into the self-employment in Ireland.
• There is an opportunity for greater collaboration between North and South to produce All Island statistics. There is currently good collaboration between CSO and NISRA, especially in relation to tourism and trade statistics.
• There may be an opportunity for a Labour Market liaison group to continue this wide-ranging discussion on labour market data needs. Currently, the CSO have regular engagements and briefings with stakeholders on particular labour market items of interest. There may be an opportunity for an annual event that includes a larger number of stakeholders and users.
• There exists differences between LFS and Administrative based employment counts at a sectoral level. Due to the nature of data collection, there will always be differences between the sectoral classifications from surveys and administrative records, which are explained in relevant publications. There is scope to use administrative sector in survey-based outputs to increase coherence between the sources, but this would require breaks in series.
• Labour market employment statistics relating to persons with a disability could be expanded. There has not been a QNHS/LFS module on disabilities since 2002 and 2004. These types of statistics would inform disability employment strategy and policy. Modules on LFS are determined by IESS regulation and the use of Census data matched to administrative data is currently seen by CSO as the best option to produce statistics related to persons with disability.
• If care workers are specifically recognised in the Constitution more data and statistics may be required for this cohort.
• Real time earnings data to complement monthly administrative employment would be very beneficial. Similarly, Job Churn data from payroll data would be very beneficial. The UK Job Churn is based upon their LFS, while the CSO’s approach will be based upon administrative payroll data where flows and stocks are relatable.

Employee Hours

• There are a number of policy requirements for a source of hourly earnings and for employee hours and earnings which are relatable to a specific period. The earnings data is currently available via Revenue’s PMOD system. Employee hours from enterprises relating to the reported PMOD earnings is considered the most effective solution to fill the data gap.
• However, there are different measurements of employee hours that need to be considered, such as: actual hours worked, contracted hours, paid hours etc.
• The inclusion of an additional data collection through an administrative system, such as PMOD, would be an additional burden to the Revenue, payroll providers and enterprises who have to provide the data.
• The provision of employee hours by employers may be relatively straight forward for employees that have standard hours each working week, but when hours are more variable, providing hours data becomes more complex and burdensome.
• There would be a larger burden for smaller enterprise and those with non-standard employment arrangements and there would be cost implications for enterprises and payroll providers.
• There would be a need to work closely with employers and payroll providers to add hours to Revenue reporting requirements. It would be important to bring all stakeholders along in the process.
• Under the Gender Pay Gap (GPG) legislation there is an obligation for enterprises with 250+ employees to report their GPG, which requires employee level hourly earnings. This legislation will apply to enterprises with 50+ employees by 2026. The recording of employee hours by these enterprises will be a necessity to meet this legislative requirement.
• DCEDIY has responsibility for a number of groups in society. In order to assess equality outcomes, it is necessary to have hourly earnings statistics relating to these groups. Thus, the availability of employee hours on an administrative system is a priority to DCEDIY.

Job Vacancies – Web Scraping

• There is a need to better utilise available scraped data from online job sites. These data can provide timely insight on current and emerging skills demands; shortages in skills; and emerging technologies. Challenges remain including:
• Cost of utilising third party scraped data
• Non-standardised coding of third party scraped data
• If web scraping of job vacancies is undertaken in the public system, challenges in making the data meaningful arise including:
• De-duplication of the same job on multiple sites
• Coding and classifying – using machine learning to provide occupation coding from job descriptions.
• Scope and coverage of sectors
• This source provides flows of jobs but is not linked to employment stocks, which would provide greater insight.
• ESRI have used scraping to identify demand for Green Skills by searching for key words such as ‘energy’, ‘renewable energy’, ‘sustainability’ etc in the job description to classify Green Skills.
• These key words can be used to get number of vacancies advertised which require Green Skills

Skills

• There is significant demand for data on skills availability, skills needs and emerging skills. This demand will increase as new technologies and sectors emerge in importance such as Artificial Interagency, Green Skills, Health and Social-care skills.
• How are Green Skills and Green Jobs defined? The types of skills that are considered Green Skills have been identified and are used by SOLAS and other bodies. The challenge is getting data on the demand for and availability of these skills.

Conclusion

There has been significant progress in the development of labour market statistics in recent years which provides the public and decision makers with data and insight. There are further developments in train to help meet an ever-increasing demand. Challenges remain especially in the area of quantifying new and emerging skills; employee hours; and self-employed. There is an opportunity to continue the conversation on this topic through a Liaison Group organised by CSO.

Item 2: Minutes of NSB meeting March 29, 2023 (NSB 2023-3-2)

The minutes of the previous meeting were agreed with one minor amendment.

item 3: CSO Director General’s Report to NSB (NSB 2023-3-3)

&

item4: NSB Strategy Progress Update and review of progress on recommendations

The Board took the DG Report and NSB Strategy Progress Update together. The main novel outputs from CSO since the last meeting were first highlighted to the Board.

The Board noted that the main strategic challenge currently facing the CSO is recruitment and having sufficient staff resources to meet the growing demands for outputs and services, as well as providing the requisite resources to the IGSS. It was noted that recruitment challenges exist across the system and that CSO are competing for data skills in a very competitive labour market. The Board heard that CSO are evolving their approach to recruitment by moving to ongoing recruitment; engaging with universities; communicating the whole package of benefit; and endeavoring to be an employer of choice.

The Board congratulated CSO on significant number of new outputs since the last meeting. The Distributional National Accounts publication was very much welcomed, and the Board looks forward to the joint distribution of Income Consumption and Wealth which will be based upon a rolling HBS. The publication on Corporate Profits was noted as being very interesting and useful for tax forecasting.

The Board noted that there is an ever-increasing demand for county level data.

The Board agree to consider having a session/agenda item on a Shared Island data.

The Board enquired whether Eurostat products being produced by CSO or ISS can all be made available on PX Stat.

Flash estimates

The Board noted that preliminary estimates of GDP and other key indicators are adding to the available knowledge for decision makers and the public. There is a need to continue to explain what these estimates are, their use and their limitations. CSO must continue to communicate to expert users, non-experts and to national and international audiences. Reconciliation between preliminary and final estimates and reasons for difference would provide greater clarity to users. The Board indicated that there may be demand for flash estimates more focused on the domestic economy such as MDD and GNI*.

The Board also noted that as global tax changes come into effect these will affect the Current Accounts, and that the potential impact of these changes on national statistics could be communicated in advance to users.

Big data

The Board enquired as to the progress being made in the acquisition and use of Big Data to produce statistics. The Board heard that utilities data is being utilised for environmental statistics; aggregate data from mobile phone providers was used to measure mobility during COVID-19 restrictions; scanner data is being used for prices statistics; and port and shipping traffic data is also being utilised.

There are opportunities to use aggregated mobile phone data to produce tourism statistics and provide insights into commuting behaviour. There is also potential to use bank card data for consumption and retail sales statistics and to supplement the Household Budget Survey.

The Board heard that the EU regulation on privately held data will help progress the area of Big Data further. For privately held big data, CSO must have the highest standards of data governance, classification, and procedures. There will be significant scrutiny to ensure the systems are fit for purpose.

The Board enquired about the affects COVID-19 has had on the seasonality of data series. The board heard that the CSO methodology division worked closely with business areas throughout the COVID-19 period to ensure the appropriate use of seasonal adjustment.

Item 5: Future Vision for the Irish Government Statistical Service (IGSS) (NSB 2023-3-5)

The Board noted that the possibility (discussions at very early stages) of CSO having a recruitment remit for data skills for the whole public and civil service would be hugely time and resource intensive and urged caution. As IGSS grows, the relationship and reporting structures between IGSS staff and CSO needs to be clearly defined. As demand for data outputs grows, Department analysts may take on more analytical responsibilities, with support from CSO and IGSS.

The Board heard that IGSS staff within Government Departments are the link between Department and the CSO. The IGSS staff bring relevant questions to CSO, but CSO are removed from the policy process. Expanding Department’s Data and Analytics capabilities outside IGSS may weaken the connection between CSO and Departments.

The Board heard that there is a possibility to extend CSO’s analytical function to the broader civil and public service. But this would be a huge undertaking which would have many challenges. CSO intend to take a stepped approach in the context of their current recruitment constrained experience but will also engage with Departments and PAS to set standards for data skills recruitment across the service.

Item 6: Data Stewardship: Summary of CSO’s current operational

                                                                                       (NSB 2023-3-6, NSB 2023-3-7, NSB 2021-1-5)

(NSB 2023-3-6, NSB 2023-3-7, NSB 2021-1-5)

The Board welcomed the papers on Data Stewardship and enquired as to whether the CSO intend to provide a Data Science and Data Analytics function as part of their Data Stewardship role. The Board heard that CSO have the Data Science and Data Analytics skills and intend to work with public bodies to support the system in this regard.

The Board noted that there remain silos of administrative data throughout the public system and heard that CSO intend to systematically work with Departments to incorporate them into the expanded National Data Infrastructure (NDI+). In addition to the use of unique identifiers across public sector data holdings the NDI+ will include the use of standardised data classifications. The first standard classifications to be rolled out as part of the NDI+ are Nationality and Occupation classifications. This will be expanded to twenty additional standard classification from household and economic surveys.

The Board heard that ISSCoP certification required Departments to have an independent timetable of data publication and quality documentation supporting their outputs.

Item 7:  Any Other Business

The Board were updated on planning for a NSB seminar on the future of Census.

The Board were asked for their feedback on the Equality Data Strategy in writing by COB 19^th May 2023.

The Board were informed that process of appointing new NSB Board members is ongoing in the Department of the Taoiseach.

The Board thanked John Martin, Gerard Brady and Eimear Cotter, who are stepping down from the NSB after six years, for the work, commitment, and insight they have each given to the Board and wished them well in the future. Padraig Dalton also thanked them on his own behalf and that of the CSO. John, Gerard and Eimear each said they were delighted to serve on the Board and had enjoyed the experience.