HELLENIC REPUBLIC

Piraeus, 9 August 2018

PRESS RELEASE
LABOUR FORCE SURVEY: May 2018

The Hellenic Statistical Authority (ELSTAT) announces the seasonally adjusted unemployment rate for May
2018.
 The seasonally adjusted unemployment rate in May 2018 was 19.5% compared to the upward
revised 21.7% in May 2017 and the downward revised 20.0% in April 2018. The number of
employed in May 2018 amounted to 3,824,393 persons. The number of unemployed
amounted to 924,941 while the number of inactive to 3,230,269. The corresponding figures for
May 2013 to 2018 are presented in Table 1.
 The number of employed persons increased by 70,052 persons compared with May 2017 (a
1.9% rate of increase) and by 48,884 persons compared with April 2018 (a 1.3% rate of
increase).
 The number of unemployed persons decreased by 113,718 persons compared with May 2017
(a 10.9% rate of decrease) and by 18,029 persons compared with April 2018 (a 1.9% rate of
decrease).
 The number of inactive persons, i.e., persons that neither work nor look for a job, increased by
7,263 persons compared with May 2017 (a 0.2% rate of increase) and decreased by 34,293
persons compared with April 2018 (a 1.1% rate of decrease).

Seasonal adjustment is a statistical technique that removes the seasonal component of a time series,
making more visible the underlying trend in the change of a characteristic. Users should take into account
the fact that seasonal adjustment procedure requires data for many months in order to remove the
seasonal component from a time series. As a result, several “observations” (that is a sufficient number of
monthly results) are necessary so that the time series reflect a significant change in the trend of
employment and unemployment.

The monthly estimates for the number of employed, unemployed and unemployment rate can be subject
to revisions in the following months caused by updates to the seasonally adjusted series whenever new
monthly data are added, inclusion of the most recent quarterly Labour Force Survey (LFS) data and update
of seasonal adjustment model with complete annual data.

Users should also take into account that the first estimates of the most recent monthly unemployment
rates are likely to be revised as they are produced with the collected and processed at the time of the press
release survey data, which do not coincide with the finally collected and processed sample of the survey.
For that reason monthly estimates are revised when the final quarterly estimates are known.

Information:
Population and Labour Market Statistics Division
Labour Force Statistics Section
Dimitris Vatikiotis
Tel: +30 213 135 2173
Fax: +30 213 135 2948
E-mail : [email protected]
 Table 1: Employed, unemployed, economically non-active persons and unemployment rate,
May 2013 – 2018

May
2013 2014 2015 2016 2017 2018

Employed 3,529,835 3,519,318 3,620,695 3,664,784 3,754,341 3,824,393

Unemployed 1,345,706 1,302,684 1,197,989 1,131,999 1,038,659 924,941

Inactive 3,313,643 3,319,423 3,273,638 3,254,824 3,223,366 3,230,629

Unemployment 27.6 27.0 24.9 23.6 21.7 19.5

Graph 1: Unemployment rate by month, May 2004 – 2018

30.0

27.6 27.0
25.0
24.9
24.0 23.6
20.0 21.7
19.5
15.0 17.0

10.0 12.3
10.6 10.0
9.1 9.2
8.3
5.0 7.3

0.0
Jan-05

Jan-07

Jan-09

Jan-11

Jan-13

Jan-15

Jan-17
Jan-04

Jan-06

Jan-08

Jan-10

Jan-12

Jan-14

Jan-16

Jan-18
Jul-05

Jul-07

Jul-09

Jul-11

Jul-13

Jul-15

Jul-17
Jul-04

Jul-06

Jul-08

Jul-10

Jul-12

Jul-14

Jul-16

The values of the curve refer to May of every year

Tables 2 and 3 illustrate unemployment rate by gender and age groups from May 2013 to 2018. Table 4
presents the evolution of unemployment rate during last 15 months by Decentralized Administrations. The
complete time series for employed, unemployed, inactive and unemployment rate are available on
ELSTAT’s website.

Table 2: Unemployment rate by gender, May 2013-2018

May
Gender
2013 2014 2015 2016 2017 2018
Males 24.6 23.9 22.2 19.8 18.2 15.8
Females 31.6 30.9 28.1 28.3 26.0 24.1
Total 27.6 27.0 24.9 23.6 21.7 19.5

2
 Table 3: Unemployment rate by age groups1, May 2013-2018

May
Age Group
2013 2014 2015 2016 2017 2018
15-24 59.7 51.6 49.8 49.4 44.0 39.7

25-34 35.9 34.2 30.4 30.4 27.4 24.0

35-44 24.2 23.4 22.6 20.0 19.0 18.1

45-54 20.9 22.3 20.1 19.5 17.1 16.3

55-64 15.8 17.8 17.4 18.9 18.7 15.3

65-74 10.2 11.6 11.4 12.8 12.0 10.6

Total 27.6 27.0 24.9 23.6 21.7 19.5

Table 4: Unemployment rate during March 2017 - May 2018, by Decentralized

th st nd
Decentralized 3d 4th 5th 6th 7th 8th 9th 10th 11th 12th 1st 2nd 3d 4th 5th
Administration 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2018 2018 2018 2018 2018

Macedonia-Thrace 22.6 22.4 22.5 22.3 21.9 21.6 21.6 21.5 21.0 20.8 20.4 20.4 20.0 20.1 20.0

Epirus-Western
27.8 27.4 27.6 27.1 27.0 26.1 26.0 25.9 26.1 26.0 25.3 25.4 24.5 25.0 24.0
Macedonia

Thessaly – Sterea Ellas 21.8 21.3 21.3 20.4 20.4 19.8 19.5 19.6 19.3 19.2 18.7 18.8 18.8 18.6 18.4

Peloponnese, Western
22.4 21.5 21.4 21.1 20.7 21.1 21.8 21.8 21.7 21.4 21.0 20.7 19.8 19.6 19.0
Greece & Ionian Islands

Attica 21.7 21.7 21.6 21.3 21.3 21.3 21.4 21.5 21.4 21.4 21.2 21.4 21.0 20.8 19.8

Aegean Islands 19.2 18.4 17.2 17.8 16.5 17.5 17.2 18.3 21.7 20.4 19.2 19.4 18.6 19.5 19.2

Crete 21.3 16.0 16.9 17.2 16.5 15.9 15.8 15.2 18.5 16.7 16.7 16.5 16.7 11.5 13.5

Greece, Total 22.1 21.7 21.7 21.3 20.9 20.8 20.8 21.0 21.1 20.8 20.7 20.7 20.0 20.0 19.5

1
Estimates of “rare” characteristics, that is estimates of characteristics that refer to 10,000 persons or less are
accompanied by large sampling errors, as for example in the case of unemployment rate in the age group of 65
– 74 years old
2
Decentralized Administrations are the lowest geographical areas for which the Labour Force Survey (LFS)
publishes estimates. LFS results are not published at lower level (NUTS 2 or NUTS 3) because, due to small
population and sample size, estimates in these areas have large sampling errors. It should be noted that the same
problem of small populations and sample sizes exist also in certain Decentralized Administrations and as a result,
estimates in these areas have large sampling errors
3
 The effect of seasonal adjustment on the estimates of Labour Force Survey
The characteristics surveyed by Labour Force Survey (number of employed, unemployed, etc.) have large
seasonal variation. For example, in Greece, employment increases during summer months because of
tourism (if there are no other countervailing factors).

Table 5 presents the change in monthly estimates of employed and unemployed because of seasonal
adjustment for the period March 2017 – May 2018 while Graphs 2, 3 and 4 present adjusted and
unadjusted time series for employed persons, unemployed persons and the rate of unemployment for the
same period.

Table 5: Change in monthly estimates of employed and unemployed persons due to seasonal adjustment

Estimated Estimated
Change due Change due
number of number of
to seasonal to seasonal
employed Change unemployed Change
adjustment adjustment
without seasonal (%) without seasonal (%)
(in (in
adjustment adjustment (in
thousands) thousands)
(in thousands) thousands)
March 2017 3,669.2 49.9 1.4 1,079.7 -22.1 -2.0
April 2017 3,761.5 3.6 0.1 1,027.1 16.3 1.6
May 2017 3,754.6 -0.3 0.0 1,036.6 2.1 0.2
June 2017 3,834.8 -51.0 -1.3 961.0 60.3 6.3
July 2017 3,843.2 -54.1 -1.4 966.0 36.1 3.7
August 2017 3,786.4 -13.9 -0.4 977.3 15.1 1.5
September 2017 3,848.8 -56.3 -1.5 947.4 50.9 5.4
October 2017 3,818.3 -59.8 -1.6 976.5 22.7 2.3
November 2017 3,675.3 53.7 1.5 1,026.8 -30.8 -3.0
December 2017 3,731.3 33.1 0.9 1,003.2 -12.8 -1.3
January 2018 3,756.6 -13.4 -0.4 993.4 -16.7 -1.7
February 2018 3,655.2 64.2 1.8 1,037.0 -64.0 -6.2
March 2018 3,768.6 44.9 1.2 973.6 -19.7 -2.0
April 2018 3,801.0 -25.5 -0.7 941.3 1.7 0.2
May 2018 3,840.1 -15.7 -0.4 913.1 11.8 1.3
Graph 2: Unemployment Rate - Adjusted and Non-adjusted Estimates

23.0

22.5

22.0
WITHOUT
21.5 seasonal
adjustment
21.0

20.5
WITH
20.0 seasonal
adjustment
19.5

19.0
May-18
May-17

Mar-18
Mar-17

Aug-17

Apr-18
Apr-17

Feb-18
Sep-17
Jun-17

Nov-17
Dec-17
Oct-17
Jul-17

Jan-18

4
 Graph 3: Number of Employed Persons - Adjusted and Non-adjusted Estimates

3,900.0

3,850.0

3,800.0

3,750.0 WITHOUT
seasonal
3,700.0 adjustment

3,650.0
WITH
3,600.0 seasonal
adjustment
3,550.0
3,500.0

May-18
May-17
Mar-17

Mar-18
Apr-17

Apr-18
Aug-17
Sep-17

Feb-18
Jun-17

Nov-17
Dec-17
Oct-17
Jul-17

Jan-18

Graph 4: Number of Unemployed Persons - Adjusted and Non-adjusted Estimates

1,100.0
1,080.0
1,060.0
1,040.0
WITHOUT
1,020.0 seasonal
1,000.0 adjustment
980.0
960.0 WITH
seasonal
940.0
adjustment
920.0
900.0
May-17

May-18
Mar-17

Mar-18
Apr-17

Apr-18
Aug-17
Sep-17

Feb-18
Jun-17

Nov-17
Dec-17
Oct-17
Jul-17

Jan-18

in thousands

5
 Revisions of monthly estimates

The monthly estimates for the number of employed, unemployed and unemployment rate can be subject
to revisions in the following months caused by updates to the seasonally adjusted series whenever new
monthly data are added, inclusion of the most recent quarterly Labour Force Survey data and update of
seasonal adjustment model with complete annual data. In the currenPt press release the estimation of
unemployment rate for the period March 2017 – April 2018 has been revised (compared to the estimation
published in the previous press-release) as follows:

Table 6: Comparison of seasonally adjusted estimates

Estimations published in July Estimations published in the

March 2017 22.2 22.1

6
 EXPLANATORY NOTES

Labour Force Survey Labour Force Survey produces estimates since 1981 (second quarter of the year). From 1998 onwards
it is a continuous quarterly survey. The main statistical objectives of the Labour Force Survey is to
divide the population of working age (15 years and over) into three mutually exclusive and exhaustive
groups - persons in employment, unemployed persons and inactive persons. In addition, the Labour
Force Survey collects information on demographic characteristics, on main job characteristics, on the
existence and characteristics of a second job, on educational attainment, on participation in
education, on previous working experience and on search of job.

Legislation The current survey is completely harmonized with European legislation. The principal legal act is the
Council Regulation (EC) No. 577/98 that stipulates the provisions on design, survey characteristics
and decision-making processes.

Reference Period The sample of Labour Force Survey is equally allocated to the 4 (or 5) weeks of the month. Every
selected household is assigned to a specific week, the reference week, running from Monday to
Sunday.

Coverage For the monthly estimates, a sub-sample of the quarterly Labour Force survey’s sample was used.

Definitions Employed are persons aged 15 years or older, who during the reference week worked, even for just
one hour, for pay or profit or they were working in the family business, or they were not at work but
had a job or business from which they were temporarily absent.
Unemployed are persons aged 15-74 who were without work during the reference week (they were
not classified as employed), were currently available for work and were either actively seeking work
in the past four weeks or had already found a job to start within the next three months.
Inactive are those persons who are neither classified as employed nor as unemployed.
Economically active population (labour force) are persons either employed or unemployed.
Unemployment Rate is the ratio of unemployed divided by total labour force.

Seasonal adjustment Seasonal adjustment is a statistical technique that removes the seasonal component of a time series,
making more visible the underlying trend in the change of a characteristic. Hellenic Statistical
Authority is using Demetra 2.0 for seasonal adjustment. Seasonally adjusted series are produced by
TRAMO&SEATS algorithm. It should be noted that due to seasonal adjustment the whole series with
monthly estimates is recalculated every time a new month is added in time series. As a result,
estimations for the previous months are often revised.

Sampling errors The monthly results of Labour Force Survey are estimations that are based in a relatively small
sample size and have large sampling errors. An indication of the magnitude of survey’s sampling
errors is that estimations of characteristics that refer to 25,000 persons at the total country are
accompanied by a coefficient of variation of at least 15%. More accurate estimates and detailed
analysis of the changes in employment can be based on the quarterly results of the survey.

Methodology Labour Force Survey’s estimates are produced by a suitable unbiased estimator which takes in to
account a) the probability of selection of every sampled household, b) the response rate in every
primary sampling unit, c) the estimated population, for the corresponding month, allocated by NUTS
II areas, gender and age group.

References Analytical description of the Labour Force Survey’s methodology and definitions can be found at
http://www.statistics.gr/en/statistics/-/publication/SJO02/-