ANALYSIS:THE RECENT Irish Times/Ipsos MRBI survey of Irish emigrants provided the first real picture of what life is like for this new generation of Irish emigrants.
Most fascinating of all was the way in which it challenged our perceptions of the emigrant mindset and experience. The survey revealed that many have been forced to leave, but it also showed more left in search of a better lifestyle and greater opportunities abroad.
The methodology used for the survey has also been commented upon by readers of this newspaper. Scrutiny of how surveys are constructed is to be welcomed.
In our note on methodology accompanying the published findings, Ipsos MRBI acknowledged the use of a purposive sampling approach, where emigrants were selected at random through our national network of interviewers, providing they met the agreed recruitment criteria.
This is an entirely acceptable sampling technique, even if it is not as robust as probability or quota sampling techniques used in most other published surveys, such as for the Irish Times/Ipsos MRBI series of political polls. Ideally we would have used a probability or quota sampling approach, but we do not have an official register of Irish emigrants from which to draw a sample, nor do we have an accurate census of Irish emigrants with which to quota-control a sample.
Purposive samples are, as was highlighted in The Irish Times, open to distortion. To minimise the potential for distortion, we introduced as many controls as possible to the recent survey of Irish emigrants.
For example, our full panel of interviewers was employed to provide the widest geographic spread of interviews. Interviewers were given a maximum number of interviews to achieve, to minimise any interviewer effect. A telephone methodology was chosen because it provides nearly 100 per cent coverage of the emigrant population. Strict recruitment criteria were applied to avoid clustering and to filter out those who had not emigrated but were simply travelling or living abroad for a short period.
Each year the Central Statistics Office (CSO) produces data on migration, but this is not suitable for the purposes of anchoring or controlling our sample of Irish emigrants for a number of reasons. Firstly, the CSO provides a measure of people “now living abroad”, which means those travelling, studying or working on short-term work or student visas are included in the statistics. Essentially, the CSO data attempts to provide a measure of “outward migration”, and not “emigration” as we might typically understand the term to mean.
The focus of our survey on Irish emigrants was on exploring the background and experiences of those who had made a permanent move abroad, and in this context the CSO definition of emigration would not have been appropriate for sampling purposes.
Secondly, with the exception of gender, the CSO migration data is not broken out separately for Irish migrants and migrants of other nationalities. So even if the CSO definition of an emigrant were aligned with our survey needs, the data published is not detailed enough on the Irish emigrant to be used for sampling (or weighting) purposes.
And thirdly, CSO data on outward migration is collected during the Quarterly National Household Survey (QNHS). During the QNHS interview, householders are asked whether a previous occupant had left and is now living abroad. Thus, the CSO outward migration data relies on the knowledge and recall of the householder as to why a previous occupant left and where they went, assuming the property has been reoccupied.
While this may be the only approach available to the CSO, and acknowledging the extreme robustness of the QNHS, as a methodology for measuring outward migration it may be vulnerable to respondent error and, possibly, sampling error. Further investigation and clarification would be needed on the models and methods used by the CSO to counter these potential biases before the data could be considered robust enough to be used for sample control purposes.
In summary, it is extremely difficult to measure and profile the Irish emigrant population. Bar conducting a global census, every technique has significant limitations.
The Irish Times/Ipsos MRBI survey of emigrants was conducted in full knowledge of these limitations, and was designed to manage these limitations to the fullest extent.
The lack of information on our emigrant population with which to control or from which to draw a sample is frustrating for researchers, but this should not deter us from attempting to reflect as accurately and as objectively as possible the views of Irish emigrants who might otherwise have no voice at all.
Damian Loscher is managing director of Ipsos MRBI