Your feedback can help us improve and enhance our services to the public. Tell us what matters to you in our online Customer Satisfaction Survey.
Wednesday 26th June 2019
11.15am – 2.15pm
Failte Ireland, 88 - 95 Amiens Street, Dublin 1.
Chairman: Donal Kelly, CSO
Agenda
11.15 – 11.20: Introduction, Welcome, Agenda & Minutes of 2018 Meeting
11.20 – 11.35: Update from NISRA
11.35 – 12.00: Review of Household Travel Survey 2018 – CSO
12.00 – 12.15: Tourism Satellite Accounts – CSO
12.15 – 12.30: Update on EDC and Tourism & Travel Release - CSO
12.30 – 12.45: Update on Overseas Travel and Household Travel Survey – CSO
12:45 – 1.15: Lunch
1.15 – 1.30: Origin – Destination visualisation of Domestic Holiday Trips – Fáilte Ireland
1.30 – 1.45: Tourism Action Plan – DTTAS
1.45 – 2.00: Any other business
Minutes
Chairperson: Donal Kelly, CSO
Attendees: Lou Ellen (CSO), Donal Kelly (CSO), Gregg Patrick (CSO), Maria Melia (DTTAS), Colm O’Connor (DTTAS), Jacqui Todoro (Department of Economy, Northern Ireland), Bernie Morris (Fáilte Ireland), Caeman Wall (Fáilte Ireland), Jim McColgan (NISRA), Brian Green (NISRA), Stephen Dunne (NISRA), Sarah McAuley (NISRA), Ian Stanley (NISRA), John Magill (NISRA), Olivia Mulhern (Tourism Ireland), John McGouran (Tourism Northern Ireland).
The chairperson welcomed members of the Liaison Group to the meeting. Apologies were forwarded for Joe Treacy (CSO) and Patsy King (CSO).
Stephen presented the 2018 annual tourist results for Northern Ireland, which were published on 6th June. Overnight trips were up 3% compared with 2017 and spend was up 6%. Cross border visits from the Republic were up 23% and expenditure by Republic of Ireland residents was up 20%. The number of rooms available for accommodation increased 19% in 2018, driven by increases in hotel rooms, guesthouses/B&Bs and particularly in self-catering. Due to the increase in supply room occupancy rates declined slightly for hotels and self-catering. However, the total number of rooms occupied increased.
Stephen gave an update on accommodation survey response rates. The response rates for hotels were generally good, always above 50%. The response rates from Guesthouses and B&B’s were historically much poorer, between 20-30% over the course of the previous two years. However, an effort was made to improve this, pushing the response rate up to 45.2% in April 2019.
Stephen also outlined other current projects ongoing in NISRA. Preliminary work is being undertaken on Tourism Satellite Accounts (the last time such an account was created for Northern Ireland was 2007). The Northern Ireland Tourism Investment Research Project was continuing, with a view to linking investment and marketing to headline tourism statistics. A Causeway Coastal Route area study was also underway. This study was seeking to assess the benefits of tourism accruing to communities actually living in this area.
Plans for 2019 and 2020 were also presented, including new statistics on day trips and enhancing statistics on the external trips of Northern Ireland residents. Methodological issues were also to be addressed, including a review of the domestic tourism multiplier and the calculation of confidence intervals. Challenges were also mentioned, including the need for a bigger sample size to provide more detailed breakdowns, the ongoing issue of survey response rates, the potentially perturbing effect of Northern Ireland hosting The Open golf tournament in July 2019 and the unknowns surrounding Brexit.
DTTAS were particularly interested in the work being done on the Tourism Satellite Accounts. NISRA promised to keep the group informed and noted that the 2007 report was available online.
The CSO were interested in finding out more about the Causeway Coastal Area study. Stephen explained that there are downsides as well as benefits to tourism so the net position was tricky to measure for a locality, but the consultants involved were working on this. Issues to be addressed were how much of the benefits actually accrued locally, in additional infrastructure required, etc.
Both Fáilte Ireland and the CSO were interested in the accommodation survey. Are the accommodation establishments obliged to register, is the survey voluntary, how was the response rate improved? NISRA responded that in Northern Ireland all establishments are obliged to register while it was noted that in the rest of Ireland only hotels needed to register, the survey was voluntary, but for hotels at least there was an incentive as the survey provides free benchmarking. Response rates were improved for guesthouses by employing a call centre to systematically follow up on non-response.
Tourism Northern Ireland were interested in whether some hotel groups had a policy of not responding. NISRA was not aware of this and would have to examine this further.
Donal combined the review of the 2018 Household Travel Survey as well as updates on the next Household Travel Survey and the Overseas Travel survey together.
Beginning with the 2018 Household Travel Survey, Donal acknowledged the frustrations caused to users by the delay in the results and outlined the reasons for the delay. The Quarter 2 results were cancelled as they appeared implausible, showing an unrealistically high number of overnight trips.
To investigate the matter, a comprehensive review of the estimates was launched, covering both the sampling and survey processes and the grossing methodology. A number of issues potentially had a bearing on the matter, including changes to the questionnaire (inclusion of same-day visit questions), removal of instructions to return the form even if there were no trips, and a fairly length GDPR insert. The sampling frame also changed mid survey. All these issues combined affected the response rate and led to a response bias, underestimating the number of people who did not go on trips.
There was no auxiliary data available to adjust for the non-response bias. Instead, response patterns observed in the quarterly data over the previous six years were used to statistically adjust the survey weights and produce more accurate estimates. Results for all four quarters of 2018 were published on 19th of June, including marginally revised estimates for the first quarter.
Unfortunately, the resources required to investigate the HTS response bias issue precluded any additional work being done on same day visits data, that has been collected on the HTS since January 2018. However, some results on these should be available by the end of 2019.
Donal explained that that the Household Travel Survey is currently undergoing a detailed methodological review. This review is expected to make recommendations for a new sample design and sampling frame, to be introduced in Q1 2020.
Regarding the monthly Overseas Travel release, the latest published estimates were for May 2019. These were published on 26th June.
NISRA complimented the CSO on their very thorough review of the Household Travel Survey but expressed some dissatisfaction with the frequency of communications throughout the process. NISRA felt that more regular updates from CSO were necessary in order to keep key users appraised of developments. The CSO acknowledged NISRA’s views and agreed to improve communications around issues such as these in the future.
NISRA commented that administrative data will feature in its 2021 Continuous Household Survey, as a means of adjusted for non-response, and wanted to know if the CSO’s publication schedule for the release was now back to normal. Having just published the revised 2018 HTS data the CSO was not yet in a position to confirm its schedule.
Lou Ellen gave a very comprehensive overview of the feasibility study she had made on the Tourism Satellite Accounts. Lou Ellen defined what a satellite account was, explained why Ireland needs such an account and illustrated the various tables of data that comprise the account.
On the demand side (tables 1-4), there are clear data gaps regarding domestic same day visits. There were also difficulties in breaking down total expenditure into the various sub-categories required, recording tourism expenditure and consumption separately, etc.
On the supply side (tables 5-7) expert knowledge will be required to complete the input/output tables, and expertise from National Accounts and Balance of Payments statisticians will also be required. The measurement of tourism sector employment poses particular challenges.
Data gaps are also prevalent with it comes to measuring tourism Grossed Fixed Capital Formation (tables 8-10). Complex methodological work will be required for all these tables.
Lou Ellen concluded with a series of recommendations, including establishing a mechanism for compiling domestic same day visitor data, looking at the possibility of surveying cruise ship same day visitors and collecting more extensive expenditure data in the existing tourism surveys. A specialist group should also be set up within the CSO to address the various methodological issues.
NISRA recognised many of the same problems in its endeavours to develop a Tourism Satellite Account and promised to keep the CSO informed of its progress.
DTTAS very much welcomed the work done to date given the national importance of this information and will brief the minister on the matter. It expressed the hope that a TSA would be produced on foot of the feasibility study whilst acknowledging the difficulties involved.
Fáilte Ireland also welcomed the work done to date and commented on the difficulties of surveying cruise passengers.
NISRA wondered how much information other countries published. The CSO responded by confirming that 19 countries sent data to Eurostat in 2016. This data was sent on a voluntary basis and the CSO confirmed that some countries such as the UK did not submit results even though they do publish TSA data. It was also highlighted that almost all the EU countries compile some of the tables but no country compiles all of the tables.
Gregg first updated the group on the Tourism & Travel release. The latest results, for the first quarter of 2019, were published in 12th June. These showed a significant decline in bednights and expenditure for foreign visitors to Ireland, despite a rise in the number of trips. The publication dates for the remaining quarters of 2019 were presented. No changes are planned for the content or format of these releases.
Next Gregg commenced an update on the Electronic Data Capture (EDC) project by recapping the original objectives. These included replacing the paper questionnaires with a mobile app, meeting new data requirements, being more flexible to changing data needs and increasing survey processing efficiency.
The project was reviewed at senior management level in the CSO in September 2018. The objectives of the review were to bring greater clarity to the project, revaluate the statistical needs, look for new possibilities of meeting needs and identify the various options and alternatives. As part of the review, the existing survey methodology was investigated by the CSO’s Methodology Division. A subsequent report recommended that a two-stage cluster sampling approach be adopted; a random sample of flights followed by a systematic sample of passengers on those flights. However, the report also raised some important questions about the stratification and size of the sample that could only be decided after a period of testing at the airport. Senior management decided to commence testing at Dublin Airport as soon as practicable.
Gregg outlined the first phase of field testing that took place from December to January. This trialled both the mobile phone devices and the two electronic questionnaires; the Passenger Card Inquiry (PCI) and the Country of Residence Survey (CRS). These proved successful and were well received by both enumerators and surveyed passengers. A systematic sampling regime was also trialled, and whilst this worked well under ideal conditions, more development was required to reflect operational circumstances on the ground.
Progress since the end of the first phase of testing was then outlined. The preliminary test results were carefully analysed and documented. The electronic questionnaires were adjusted, specifically to capture more information on non-response. The systematic sampling regime was refined. A random flight generator was built. A memorandum of understanding was agreed with daa plc on sharing flight-specific data, and an enumerator recruitment campaign was launched.
It was decided that the next phase of field testing will focus on departing flights only and all departing foreign visitors will be asked the full PCI questionnaire. The next phase of testing will be on a bigger scale, involving more enumerators. The goal is to fully test all aspects of the proposed two-stage cluster sampling methodology. Dynamic feedback loops are to be created, so that the sampling can be monitored and directed on a daily basis. The objective of this next phase is to inform key decision making, such as how much resources were required to conduct a full-scale parallel run and how many quarters such a run should last. It is hoped to commence this second phase of testing before the end of the summer.
Some members of the group expressed disappointment with the delays in the project to-date and the lack of regular updates. Concern was also expressed about the lack of a timeline for the project going forward.
CSO acknowledged the frustrations of the group with the rate of progress of the EDC project. However, CSO felt that the project was now making consistent incremental progress and that the project was on a clear path to completion. CSO explained that it was not possible to set an exact deadline for the completion of the project because the outcome of each stage of testing dictated the next steps required and this would have an impact on the timeline.
CSO asked the group for its views on the need for a full-scale parallel run spanning at least four quarters, once the current phase of EDC testing was completed. Members said that while data from the EDC project was needed as soon as possible, it was also important to maintain a coherent time series and explain differences between the old and new surveys. DTTAS said that without a parallel run, it would be more difficult to track the long-term trends and explain emerging differences, and that this was important from a policy perspective, in particular with regard to revised tourism policy targets which had recently been agreed. NISRA said that at present they cannot use the data on Northern Ireland visitors captured on the PCI and that data from the new survey was needed as soon as possible.
NISRA offered to assist the EDC project in any way they can. They queried if allocating more CSO resources would help the project, indicating that they would seek to influence the senior leadership team at CSO to deploy more resources to the work should that be the case. CSO responded that the main delays at present were caused by an office-wide change in mobile phone provider and by difficulties recruiting additional enumerators. More resources would not help these issues in the short-term.
NISRA requested access to the electronic questionnaire, to anticipate possible data gaps. CSO promised to provide this information to NISRA.
NISRA and Tourism Ireland commented that more regular communication around this project was needed. CSO stated that is would reflect this discussion back to senior management and promised to improve communications in future.
Bernie gave a demonstration of the use of a network graph to display origin-destination statistics for domestic holidays. The graph shows all 26 counties in the Republic, with interconnecting lines showing the flow of domestic holiday makers between counties. The greater the flow, the thicker the interconnecting arrows. Not surprisingly the greatest flows involved Dublin, particularly between Dublin and counties Cork, Kerry, Galway and Wexford. There were also very strong flows between Cork and Kerry. The data source was Fáilte Ireland’s Domestic Holiday tracking survey. The network graph was created using R.
Maria presented the Tourism Action Plan 2019-2021 to the group. The Action Plan falls under the government’s Tourism Policy Statement, People, Place and Policy – Growing Tourism to 2025. The importance of the tourism sector to the economy was underlined, with an estimated 260,000 employed in 2018.
Firstly, achievements under the Tourism Action Plan for 2016-2018 were highlighted. These included launching the Dublin marketing brand and developing Ireland as a food tourism destination.
The new action plan for 2019-2021 has 27 specific actions, developed from stakeholder workshops. It was pointed out that many of the original growth targets from the Tourism Policy Statement had already been achieved, including the target of 10 million visits to Ireland annually. The new plan looked at revising these targets.
The 2019-2021 action plan had a strong emphasis on sustainable tourism development. Under the plan a working group has been set up to review policy and best practise in sustainable tourism and to propose guiding principals for sustainable tourism development.
The action plan outlined how Ireland was to be marketed as a visitor destination, with a focus on regional growth, through initiatives such as the Wild Atlantic Way. The visitor experience was to be enhanced further, through measures such as greater accessibility, high quality accommodation and food tourism. Actions relating to research in the tourism sector aim to progress the development of a TSA, examine the competitiveness of the sector and also identify new sources of data relevant to tourism. There was also a focus on supporting local communities in tourism. For example, a capability programme for local authority tourism officers will be rolled out by the CCMA.
Finally, the role of the Tourism Action Plan 2019-2021 in the context of wider government policy was discussed. The plan was fully aligned with other policy statements such as Global Ireland: Ireland’s Global Footprint to 2025 and the National Development Plan 2018-2027.
NISRA asked in the tourism expenditure goals were in current or real prices. DTTAS confirmed that the prices were adjusted for inflation.
CSO commented that sustainable development has become a much more important consideration that was making itself felt across many sectors. However, it was not clear yet what the implications were for official statistics.
Any other business
No other business was raised. Therefore, the chairman thanked the participants for their attendance and the meeting concluded.