This publication is part of a series of projects that the CSO has established in collaboration with Irish public sector bodies to examine learner outcomes. The CSO has developed a statistical framework known as the ‘Educational Longitudinal Database’ (ELD) to act as the basis for these projects. The ELD is produced by matching datasets on learners that have completed courses or programmes to other datasets which describe their outcomes in subsequent years. The data sources used to describe learner outcomes include employment and self-employment datasets from Revenue, benefits data from the Department of Social Protection, and data on educational participation from the Department of Education and Youth and several state agencies, including the Higher Education Authority (HEA), Quality and Qualifications Ireland (QQI) and SOLAS.
The ELD can be used to analyse outcomes for learners from a wide variety of educational programmes, ranging from post-primary level to adult education. The integrated approach of the ELD ensures that analysis datasets are built in a consistent and highly efficient manner. The CSO hopes that the ELD will provide the basis for a series of partnerships with other public sector bodies leading to policy-relevant insight in outcomes analysis and impact evaluation. These projects will be mutually beneficial, with high transferability not only of technical processes but also of understanding and expertise.
The present project on higher education outcomes uses as its primary data source an annual dataset on graduations from Irish higher education institutions, which is provided by the HEA. In line with CSO data protocols, all identifiable information from each of the data sources is removed, such as name, date of birth and addresses. The PPSN is replaced with a ‘protected identifier key’ (PIK) and it is this PIK which is used to link person-based data. The resulting data is then said to be ‘pseudonymised’ and this is what is used for all analysis.
For further information on the data sources and linking procedures, please refer to the primary ELD documentation page.
The CSO is committed to broadening the range of high-quality information it provides on societal and economic change. The large increase in the volume and nature of secondary data in recent years poses a variety of challenges and opportunities for institutes of national statistics. Joining secondary data sources in a safe manner across public service bodies, while adhering to statistical and data protection legislation, can provide new analysis to support decision-making and accountability in a way that is not possible using discrete datasets. Furthermore, a coordinated approach to data integration can lead to cost savings, greater efficiency and a reduction in duplication.
The CSO has a formal role in coordinating the integration of statistical and administrative data across public service bodies that together make up the Irish Statistical System (ISS). Underpinning this integration is the development of a National Data Infrastructure – a platform for linking data across the administrative system using unique identifiers for individuals, businesses and locations. The data linking for statistical purposes is carried out by the CSO on pseudonymised datasets using only those variables which are relevant to the research being undertaken. A strong focus on data integration, which requires the collection and storage of identifiers such as PPSN and Eircodes, is a priority of the ISS in its goal of improving the analytical capacity of the system.
Data protection is a core principle of the CSO and is central to the development of the NDI. As well as the strict legal protections set out in the Statistics Act, 1993, and other existing regulations, we are committed to ensuring compliance with all data protection requirements. These include the Data Sharing and Governance Act (2019) and the General Data Protection Regulation (GDPR, EU 2016/679).
This report on higher education outcomes, using administrative data from the HEA, is a good example of the type of partnership approach the CSO can adopt with a public agency using the National Data Infrastructure (NDI). The CSO is hopeful that this joint project between the CSO and the HEA, as well as the innovative methodologies used in the report, will become a template for further collaborations with other government departments and agencies.
All graduate counts have been rounded to the nearest ten. Median earnings have been rounded to the nearest €5 and average earnings have not been calculated for cases where the average referred to fewer than five graduates.
A threshold age is defined for each type of award, and a graduate must be of an age equal or younger than this at the time of graduation to be classified as ‘Young’. The threshold ages for each award type are detailed on the ELD methodology page.
For example, if a graduate was aged 26 at the time of finishing an ordinary degree they would be classified as ‘mature’ since they are older than 23 years of age. However, a person who is aged 27 and graduating with a master’s degree would be classified as ‘young’ since they are not older than the threshold age for this type of award which is 27.
The Irish National Framework for Qualifications (NFQ) is a framework which classifies learning achievement based on the level of knowledge, skill and competence. 'Award type' here refers to names that are commonly given to different types of qualifications, such as certificate, bachelor’s degree, master’s degree, etc. For the most part, NFQ level 6 awards are advanced certificates or higher certificates, level 7 awards are primarily ordinary bachelor’s degrees and level 8 awards are primarily honours bachelor’s degrees. Level 9 awards include master’s degrees and postgraduate diplomas. Level 10 awards are doctoral degrees (PhD, including higher doctorates). The relationship between award type and NFQ level is not precisely one-to-one. Where discrepancies arise, course information is examined and the NFQ level is adjusted to reflect the award type as appropriate.
The year of graduation refers to the latter of the two calendar years spanned by the final academic year. For example, where a graduate’s final year was in 2012/2013, the graduation year is taken as 2013. The first year after graduation then refers to the calendar year following the graduation year (2014 in this case).
The fields of study referred to in this report are based on the International Standard Classification of Education (ISCED) broad fields. Due to a change in the ISCED classification framework in 2013, some mapping was used to assign equivalent broad field classifications to courses from years prior to 2014. This mapping is described in the overall ELD methodology documentation.
As noted above, mature graduates were excluded from all analyses presented here. Mature graduates are likely to represent a heterogeneous cohort and their outcomes may be strongly influenced by working experience prior to enrolling in a higher education course. ‘Mature’ graduates were defined according to the threshold ages for young and mature graduates delineated in the above section on graduate terminology.
Graduates were also excluded if they were recorded as being an “overseas student”. These are students who attend a campus which is associated with an Irish institution but located in a different country.
Graduates from non-HEA institutions do not feature in these analyses.
Graduates from certain programmes were also excluded, according to the following criteria:
Graduates recorded as being enrolled in an upskilling programme (Springboard, Labour Market Activation, and ICT Skills Conversion)
Graduates from General/Generic courses (broad ISCED field 0)
Graduates from Access/Foundation courses, FETAC Certificates, Professional Training Qualifications, and occasional courses.
A number of graduation records have a missing or invalid PPSN. As a pseudonymised version of PPSN was the basis for the linking of records, any individuals who lack a PPSN cannot be matched to any other administrative data sources. These graduation records are included in the statistics on graduation numbers shown in the Chapter on Background Statistics but they are excluded from all other chapters on graduate outcomes, NACE sectors of employment, and average earnings. The rates of missing and invalid PPSN across a number of parameters are given in the Overview of Graduates chapter. This is given as a guide to the user and to inform interpretation of the other chapters.
As noted in the chapter What Graduates Earn, when presenting analysis of earnings, only income through the PAYE system was included in this analysis. Income from self-employment activities registered through the self-assessment system was therefore excluded.
Graduates were sometimes recorded as having received more than one higher education award in a single year. Typical examples include a course which was jointly hosted by more than one institute, or a graduate who received a diploma in education in combination with an award for completing a degree course. If the award types were different, it is possible for a graduate to be classified as young in the case of one award and mature in the case of another within the same year. In such cases the duplication would not have to be dealt with since the mature graduations were already excluded.
It is desirable for the purposes of data matching that there be only one graduation record per year per individual. In cases where an individual graduated from more than one course per year, one course was kept according to a hierarchy based on Award Type. The following order of preference from highest to lowest was applied: PhD, master’s, honour’s degree, ordinary degree, postgraduate qualification, certificate. In cases where a graduate had more than one course and the same Award Type then one course was kept according to NFQ level.
An individual is regarded as being in ‘substantial employment’ within a given calendar year if they fulfil either of the criteria below.
1. Substantial P35 Employment
They fulfil the following two requirements:
They have at least 12 weeks of insurable work within the calendar year across all employments. This can be supplemented by weeks of maternity and/or illness leave.
The average weekly earnings from their main employer is at least €100 per week;
2. Substantial Self-Employment
Their total turnover across all self-employment activities is at least €1,000 within the calendar year.
Employment data from Revenue includes one record for each occupation of every individual. Each record includes the number of weeks of insurable work and the gross pay received by the employee. It does not include the hourly wage or the number of hours worked. The Main Employer for each individual is the one which contributes the single largest pay to that individual over the course of the year. The Weekly Earnings for each individual is found using data from the main employer only, and is calculated as the gross pay divided by the number of weeks of insurable work.
For years 2011-2018 the employee tax data used in the ELD came from employer end-of-year returns, P35, submitted to Revenue. The P35 was an annual return that was completed by all registered employers after the tax year end, up to 2018. Since 1 January 2019, Revenue have operated real-time reporting of payroll known as “PAYE Modernisation” (PMOD). Employers are required to report their employees’ pay and deductions in real-time to Revenue each time they operate payroll. Information is provided to Revenue at individual payslip level. The ELD analysis for 2019 and subsequent years is based on the employee tax data provided from Revenue’s PMOD system.
Re-enrolment of graduates in higher education was analysed using enrolments data provided by the HEA, which includes a record for each academic year that an individual is enrolled. Since the academic year spans two calendar years, for the purposes of the outcomes analysis, a graduate was considered to be re-enrolled in both of the calendar years covered by an academic year. For example, an individual enrolled in 2013/2014 was categorised as being in education in both 2013 and 2014. Certain types of course excluded from the graduation dataset, such as FETAC courses, professional training qualifications and courses in the General/Generic field (ISCED code 0) were not excluded from the enrolment dataset. For example, a graduate from a FETAC course would not have been included in the original graduate cohort, but a person who graduates with an honours bachelor’s degree and subsequently begins a FETAC course is considered to be in education from the perspective of assigning a graduate outcome destination.
This category comprises graduates who are neither enrolled in higher education nor are involved in substantial employment within the year but have some record in an administrative data source for that year. These graduates may have some record of non-substantial employment (not meeting the criteria previously delineated), may have been in receipt of some social benefit during the year, or have an enrolment record in an education source that is not counted in the previous category. Data sources include payments made by the Department of Social Protection, payments made as part of the Pandemic Unemployment Payment scheme, enrolment activity in SOLAS’ PLSS database, and enrolment activity in post-primary databases which record enrolment in some PLC programmes
A graduate is assigned to the category of ‘not captured’ if there was no evidence of their having engaged in any of the above classified outcome activities. Most graduates categorised as ‘not captured’ are assumed to have emigrated or returned to their country of origin, but it is important to note that this does not represent a definitive measure of emigration. It is possible that a graduate remained in the country but nevertheless was not captured by any of the administrative data. Conversely, it is also possible that a graduate had in fact emigrated but had engaged in some activity which was captured by the administrative data, and was therefore categorised as being in ‘neither employment nor education’, rather than ‘not captured’.
This category does not include those graduates who lacked a valid PPSN as part of their graduation record. These individuals are not considered in any outcomes-related analysis because their record cannot be matched to any other data source.
In 2022 ‘National University of Ireland, Galway’ was renamed to ‘University of Galway’; for consistency, the updated name is used for all reference years, including those prior to this name change. More fundamentally, following the Technological Universities Act 2018, a number of pre-existing institutes of technology and colleges merged to form new Technological Universities, as is evident in Table 5.2 in the Overview of Graduates chapter. For further information, please consult the HEA website and the ELD methodology page.
The country of domicile refers to the permanent address of a graduate prior to their entry to the programme of study for three of the five years prior to initial enrolment. Thus if the student had been residing in Ireland for three of the five years prior to registering for their course of study, their domiciliary of origin would be classified as Ireland.
This variable has been newly included in various tables in this latest version of the release to provide clearer insight into international graduates, but it should be noted that PPSN coverage (which is required for data linkage and therefore assessment of outcomes) was higher for Irish-domiciled graduates.
Some minor revisions have been made to the methodology underpinning this publication since the previous release. These changes and their impact on the associated PxStat tables are detailed below.
As outcome categories are derived from a range of administrative data sources, routine updating of various data sources has occurred since the last publication of this release. The present analyses represent those using the latest available data at the time of writing for all years considered in the report.
At the time of the previous release, self-employment data was not available for outcome year 2020. This has since become available and is now included in the relevant analyses which can have affected the classification of a graduate’s outcome in this year.
A methodological adjustment to the way in which education outcomes [‘In Education Only’ and ‘In Employment and Education’] has been implemented since the previous publication. Previously, the presence of a graduation record in either academic year was used to retrospectively infer an enrolment in a respective outcome year. In the updated protocol, only an enrolment record is used to define subsequent enrolment and whether a subsequent graduation occurred is not considered. This is seen as preferable as it entails fewer assumptions, but has resulted in a number of altered outcome classifications (as detailed in Table 6.1 below).
It should be noted that changes to the earnings calculations described in the proceeding section may also have changed the outcome that a given graduate was assigned since the previous release. For example, if a graduate’s newly revised earnings were reclassified as substantial employment or, in the opposite case, if newly revised earnings fell below the required threshold where previously it has been classified as ‘substantial’.
The net effect of the methodological adjustments to enrolment status, earnings, and routine updating of data sources on graduates’ outcome categories is small overall but is summarised by graduation year in the table below. Values in the table represent the proportion of cases where an individual outcome record has been changed since the previous edition of this release. This is presented up until 2020 because this was the extent of the outcome year periods covered in the previous release.
As detailed in the overall methodological notes concerning the Educational Longitudinal Database (ELD), for ELD-derived publications released in 2025 and after, earnings values are no longer adjusted using the Consumer Price Index (CPI). Secondly, in line with the updated analysis protocol for earnings, additional employment records (at different PSRI classes) with the same employer were aggregated. This means that the earnings and insurable weeks associated with different PSRI classes were summed together for each employer creating one record of employment for each employer.
As the previous adjustment utilised 2016-based CPI values, the precise impact of this change varies by calendar year. Nevertheless, as a generalisation, given the rate of inflation over these years, updated (non-CPI-adjusted values) tended to be slightly lower for years up until around 2012 from which point they were more likely be increased. Outcome year 2020 is an exception to this as it was disproportionately affected by the change in how different PRSI classes were aggregated, as noted in the overall ELD methodology page, and so values of earnings in this calendar year generally saw a decrease. COVID-19 income support schemes active at this time may have had an outsized impact on determining the Average Weekly Earnings for each individual from their Main Employer in this year.
The net impact of both the above changes on average earnings is summarised for graduation years 2010-2019 in the table below, and overall is considered fairly small. This has been calculated for records of graduates who remained classified as being in substantial employment and whose averaged earnings have therefore subsequently been assessed and featured in the chapter concerning earnings.
To provide an overall sense of the scale of the changes in earnings, across all records of these graduates in which the difference was assessed, in around 63% of cases an individual’s median earnings changed by €10 or less.
Learn about our data and confidentiality safeguards, and the steps we take to produce statistics that can be trusted by all.