Labour Force Survey (LFS) results were issued for Q2 2018 on August 28 and the CSO has produced this note to assist users in the interpretation of trends measured as the LFS is one of the most important measures of the state of the economy in Ireland. The note is, in particular, intended to aid users in understanding why this Q2 2018 LFS release has resulted in a large revision to the previously estimated Monthly Unemployment forecast.
What does the Q2 2018 LFS tell us about employment, unemployment and labour force data for Ireland?
In Q2 2018, employment totalled 2,255,000, up 3.4% or 74,100 from the same quarter in 2017. When adjusted for seasonal factors, employment increased by 0.8% or 17,300 from Q1 2018.
There were 144,300 people unemployed in Q2 2018 and this was down 10.1% or 16,200 from Q2 2017. When seasonally adjusted, the unemployment rate fell from 5.9% to 5.8% over the quarter, while the seasonally adjusted number of persons unemployed was 139,300.
The total number of persons in the labour force is now 2,399,300, up 2.5% or 57,900 from Q2 2017.
Is this the highest level of employment that has been recorded in Ireland?
On a seasonally adjusted basis, the number of persons in employment (2,256,500) in Q2 2018 is the highest on record. Excluding Q1 2018, the previous high was 2,237,300 which was recorded in Q4 2007.
How do current unemployment rates compare to previous rates?
Over the period 2000 to 2007, the annual average unemployment rate (seasonally adjusted) ranged between a low of 4.2% in 2001 and a high of 5.0% in 2007. The average rate then increased to 6.8% in 2008 and continued to increase and peaked at 15.4% in 2011 and 15.5% in 2012.The seasonally adjusted figure of 5.8% for Q2 2018 is therefore not as low as these historic figures but it is the lowest rate of unemployment recorded in Ireland since Q1 2008 when the comparable rate was 5.4%.
Why does the Labour Force Survey figure for unemployment differ so much from the Monthly Unemployment forecasts released last month by the CSO?
For months where the quarterly unemployment data is not yet available, the ratio of the LFS monthly estimate to the Live Register monthly estimate is forecast forward in order to extrapolate a monthly LFS estimate. The monthly series is revised each quarter when the latest quarterly data becomes available. This ‘hybrid’ approach adheres to agreed international practice but can result in revisions to forecasted data.
This forecasting approach was used to generate the previously published 5.1% for July and it is the availability of new Q2 2018 benchmarks from the LFS which have revised the rate upward to 5.9%. Revisions will tend to be larger when these benchmarks are different from what the forecasting system had predicted.
The original Monthly Unemployment forecast figures published for May 2018 which were revised as part of the Q1 2018 LFS release are included in Table 1 below. Original Monthly Unemployment forecast figures published for July 2018 which are now revised as part of the Q2 2018 LFS release are also included:
| Table 1: Recent adjustments to Monthly Unemployment volumes and rates | ||||||
| Original estimate | Revised estimate | Original estimate | Revised estimate | |||
| Monthly Unemployment | Labour Force Survey | Monthly Unemployment | Labour Force Survey | |||
| May-18 | Q1 2018 | Jul-18 | Q2 2018 | |||
| March 2018 | Unemployment volume | 143,700 | 132,200 | 132,000 | 137,900 | |
| Unemployment rate | 6.0% | 5.6% | 5.6% | 5.8% | ||
| April 2018 | Unemployment volume | 140,800 | 126,600 | 126,400 | 138,500 | |
| Unemployment rate | 5.9% | 5.4% | 5.4% | 5.8% | ||
| May 2018 | Unemployment volume | 139,000 | 123,700 | 123,300 | 139,600 | |
| Unemployment rate | 5.8% | 5.3% | 5.2% | 5.9% | ||
| June 2018 | Unemployment volume | : | : | 120,700 | 139,900 | |
| Unemployment rate | : | : | 5.1% | 5.9% | ||
| July 2018 | Unemployment volume | : | : | 120,500 | 140,800 | |
| Unemployment rate | : | : | 5.1% | 5.9% | ||
How does the absolute level of unemployment impact on the calculation of monthly estimates?
Unemployment in Ireland has been declining on an annual basis since Q3 2012 and in absolute terms was estimated to have been 144,300 in Q2 2018. With relatively low levels of unemployment now being seen, movements in the Live Register may not fully reflect relatively small absolute changes in unemployment levels from the LFS and this can lead to larger revisions than previously. It can also be noted that the measure of statistical variability will increase as the absolute measure of unemployment becomes lower.
Could the trend in the rate of decline of unemployment be contributing to the discrepancy between the forecasted monthly unemployment rate and the revised monthly unemployment rate?
Figures published in the Q2 2018 LFS release indicate that while employment continues to show strong annual increases, the 10.1% decline in unemployment in the year to Q2 2018 is at its lowest level since an annual decline of 6.4% recorded in the year to Q2 2013. This reduction in the annual decline in unemployment is to be expected however, as unemployment moves towards a natural stability. This transition would be expected to have a greater impact on the level of revisions seen as the Live Register figures used to forecast the monthly rate would not necessarily show the same trend as the these LFS unemployment benchmarks.
Could increasing employment cause overall participation in the Labour Force to increase?
As employment continues to increase, the expectation of becoming employed will also increase. This increasing expectation of employment would be expected to result in a higher rate of participation in the Labour Force. This increase in participation may further impact the scale of revisions to the forecasted measure of unemployment as persons entering the Labour Force may not be counted on the Live Register.
Will the Population and Migration estimates for April 2018 have an impact on Labour Market estimates?
The Population and Migration estimates for April 2018 show inward migration of just over 90,000, of whom 28,400 are returning Irish nationals. Overall, there was net inward migration in the year to April 2018 of 34,000. This increasing population, who may not be captured as part of the Live Register, may also impact the scale of revisions to the forecasted measure of unemployment.
Principal Economic Status (PES) estimates are also available for these migration results and show that of the 79,800 inward migrants aged 15 or more, 48,700 (61.0%) were self-classified as ‘At work’, 9,300 (11.7%) were self-classified as ‘Student’ and 9,000 (11.3%) were self-classified as ‘Unemployed’.
Why does the CSO forecast Monthly Unemployment estimates if they are subject to revision?
The LFS is collected and published on a quarterly basis. LFS results are generally not available until 8 weeks after the end of the quarter so results for the first quarter are published towards the end of May, for the second quarter towards the end of August, for the third quarter towards the end of November and for the fourth quarter towards the end of February of the following year.
The CSO therefore introduced Monthly Unemployment forecasts in 2015 in an effort to provide more timely Labour Market data for users. While these estimates are, and have previously been subject to revision, the CSO believes that there continues to be a need for a short-term, more timely measure for Labour Market estimates even if subject to revision.
What will the CSO do to improve monthly forecasting?
In light of the scale of recent revisions to the monthly unemployment rate, the CSO is preparing to review the existing methodology used to forecast these estimates. This review will, in particular, take into account changes required as the economy moves towards a state of full employment. This review will be carried out by CSO methodology experts and it is also intended to work with experts from other National Statistical Institutes as part of this process.
Jim Dalton (+353) 21 453 5623
Martina O'Callaghan (+353) 21 453 5491
Learn about our data and confidentiality safeguards, and the steps we take to produce statistics that can be trusted by all.