Migration Estimates for Ireland from Administrative Data Sources 2014-2020

Migration between 2014 and 2019 was estimated using two different counting rules based on annual data – the ‘one dataset rule’ and the ‘two dataset rule’. The first of these rules is the simplest - it looks at absence, and presence in a minimum of one dataset, in consecutive years. The ‘two dataset rule’ is better able to deal with short-term migrants, as it needs presence in two datasets although there are exceptions, see Methodology.

The numbers of immigrants and emigrants are different for both approaches, but net migration is very similar, see Figure 2.1 and Table 1.1. People who enter and leave Ireland in the same calendar year are counted as both an immigrant and emigrant and therefore are not included in the net flows. Note that improvements to the coverage of the Protected Identifier Key (PIK), see Methodology for description of PIKs, of certain data sources from 2017 onwards leads to some variation in net migration between 2016 and 2017. Furthermore, self-employed data is not available for 2020, and is currently preliminary/incomplete in the data available to CSO for 2018 and 2019.

Total net migration between 2014 and 2019 was 199,100 using the ‘one dataset rule’, slightly higher than the ‘two dataset rule’ figure of 190,400.

Annual net migration more than doubled between 2014 and 2018, increasing from around 16,000 to around 42,000 persons for both counting rules. However, in 2019 net migration was higher under the 'one dataset rule' at 44,900, compared with 33,800 using the 'two dataset rule'. Note that the immigration figure for the 2019 'two dataset rule' is likely an underestimate. This is because a person could also be counted as an immigrant when present in 2019 and 2020 data, see Methodology, however as less datasets are available for 2020, not all of those persons will be included (in particular students and the self-employed are not currently in the data available to CSO). 

Both counting rules produce similar values for net migration for Irish and non-Irish nationals, except for Irish nationals in 2019, where net migration was -6,600 under the ‘one dataset rule’ and -13,400 under the ‘two dataset rule’. See Figure 2.2.  As noted above, the immigration figure for 2019 is an underestimate for the 'two dataset rule' which seems to impact Irish national figures in particular. 

Net migration of Irish nationals was negative or close to zero for all years. In contrast, net migration for non-Irish nationals each year was positive and increased from about 27,000 in 2014 to 49,000 in 2018. The ‘one dataset rule’ showed net migration of non-Irish nationals was 51,500 in 2019, compared to 47,100 using the ‘two dataset rule’. See Table 2.1a.

In 2014 the highest numbers of non-Irish immigrants were from Poland and the UK. By 2019 the highest numbers of non-Irish immigrants were from Brazil and Romania for both 'annual' methodologies. See Table 2.2. Note the nationality breakdown in this analysis does not capture any changes in citizenship status, for example, an immigrant acquiring Irish nationality. 

Net migration was highest in the age group 25-44 between 2014 and 2019 using both counting rules. Net migration rose from about 12,000 in 2014 for those aged 25-44 using both counting rules to 29,100 for the ‘one dataset rule’ and 25,600 for the ‘two dataset rule’ by 2019.

Note that the fluctuation in the age group 45-64 between 2016 and 2017 may be caused by improvements in coverage of PIKs.

Also note that 2019 figures are preliminary. The net outward migration in age group 45-64 for the 'two dataset rule' is partially attributable to self-employed data not being available yet for 2020 to the CSO. The fall in net migration in the age group 15-24 for both methodologies is also partially due to the educational datasets not being available yet for the 2019/2020 academic year, resulting in emigration being overestimated.

Net migration was very similar for males and females between 2014 and 2019, with 97,900 males and 101,200 females for the 'one dataset rule' and 94,400 males and 95,800 females for the 'two dataset rule'.

The highest net migration for both males and females was in the age group 25-44 for all years.

Net migration was high for people in substantial employment, that is substantial P35 employment with at least 12 weeks of insurable work within the calendar year or substantial self-employment with total turnover across all self-employment activities of at least €1,000 within the calendar year (see Methodology for more details on criteria).  However net migration was low or even negative for people in the unemployed and other categories. See Figure 2.5 and Table 2.4a for estimates up to 2019. Net migration for the substantially employed in 2019 was 28,700 under the 'one dataset rule' and 27,500 in the 'two dataset rule'. In comparison, net migration was -2,300 for the unemployed and 1,300 for pensioners using both counting rules in 2019. 

Unemployment and pensioner figures are identical for the ‘one dataset rule’ and the ‘two dataset rule’ as welfare recipients only require presence in one dataset and are treated the same in both methodologies.

Students are included in 'Substantial employment' or in the category 'Other'. Persons receiving welfare benefits other than a pension or unemployment benefit will be included in the category 'Other', as will persons in non-substantial employment. The fall in net migration for 'other' in 2019 for the 'two dataset rule' is most likely due to the missing education datasets for 2019/2020.

(Note that a person is classified as substantially employed in any given year if they are in substantial employment for at least 12 weeks and with significant wages. An immigrant who is looking for work immediately after arriving in Ireland but who subsequently gains substantial employment will therefore be counted as employed, and not unemployed. This employment can either be in the calendar year the person migrated to Ireland or the following year. This is because if a person migrated to Ireland in Q4 of a calendar year and started working immediately, they would not be counted as being in substantial employment (cannot work more than 12 weeks to the end of the calendar year) and hence allowance is made for substantial employment in the following year. This contrasts with the Official estimates, where the economic status is measured at a point in time. See Methodology for more details.)